#  @koltregaskes Kol Tregaskes Kol Tregaskes posts on X about open ai, ai, $googl, model the most. They currently have [------] followers and [---] posts still getting attention that total [------] engagements in the last [--] hours. ### Engagements: [------] [#](/creator/twitter::14069848/interactions)  - [--] Week [---------] +1,741% - [--] Month [---------] +676% - [--] Months [----------] +148% - [--] Year [----------] +282% ### Mentions: [--] [#](/creator/twitter::14069848/posts_active)  - [--] Week [--] +3.50% - [--] Month [---] +24% - [--] Months [-----] +17% - [--] Year [-----] -24% ### Followers: [------] [#](/creator/twitter::14069848/followers)  - [--] Week [------] +0.22% - [--] Month [------] +2.30% - [--] Months [------] +34% - [--] Year [------] +102% ### CreatorRank: [-------] [#](/creator/twitter::14069848/influencer_rank)  ### Social Influence **Social category influence** [technology brands](/list/technology-brands) 48.04% [stocks](/list/stocks) 10.06% [finance](/list/finance) 3.35% [social networks](/list/social-networks) 2.23% [celebrities](/list/celebrities) 1.68% [countries](/list/countries) 1.12% [vc firms](/list/vc-firms) 0.56% [cryptocurrencies](/list/cryptocurrencies) 0.56% [gaming](/list/gaming) 0.56% [products](/list/products) 0.56% **Social topic influence** [open ai](/topic/open-ai) #2582, [ai](/topic/ai) #5509, [$googl](/topic/$googl) 6.15%, [model](/topic/model) #1041, [anthropic](/topic/anthropic) #338, [in the](/topic/in-the) 3.91%, [agentic](/topic/agentic) 3.35%, [app](/topic/app) #1576, [openclaw](/topic/openclaw) #1177, [claude code](/topic/claude-code) 2.23% **Top accounts mentioned or mentioned by** [@openai](/creator/undefined) [@ceoalexcolon](/creator/undefined) [@antonpme](/creator/undefined) [@sama](/creator/undefined) [@grok](/creator/undefined) [@testingcatalog](/creator/undefined) [@harpaai](/creator/undefined) [@xdimensionnews](/creator/undefined) [@axttimm](/creator/undefined) [@cosmo_kappa](/creator/undefined) [@btibor91](/creator/undefined) [@shawnryan96](/creator/undefined) [@ctrlaltdwayne](/creator/undefined) [@googledeepmind](/creator/undefined) [@huanghlzone](/creator/undefined) [@brandgrowthos](/creator/undefined) [@mwydzyad82729](/creator/undefined) [@yasamanini](/creator/undefined) [@openaidevs](/creator/undefined) [@gdb](/creator/undefined) **Top assets mentioned** [Alphabet Inc Class A (GOOGL)](/topic/$googl) [Spotify Technology (SPOT)](/topic/$spot) [USDC (USDC)](/topic/usdc) [Morgan Stanley (MS)](/topic/morgan-stanley) [Accenture (ACN)](/topic/accenture) ### Top Social Posts Top posts by engagements in the last [--] hours "Claude Sonnet [--] being prepared for launch I'm wondering if this is a next week release. Anthropic is preparing a new announcement banner in the Claude web app and Console codenamed "Try Parsley" - the last similar banner added to both was "Try Cilantro" which was the codename for Opus [---] https://t.co/dVdF5uc44L Anthropic is preparing a new announcement banner in the Claude web app and Console codenamed "Try Parsley" - the last similar banner added to both was "Try Cilantro" which was the codename for Opus [---] https://t.co/dVdF5uc44L" [X Link](https://x.com/koltregaskes/status/2022801974639620523) 2026-02-14T22:35Z 18.3K followers, [----] engagements "M2.5 released an open-source model for productivity tasks. - Achieves 80.2% on SWE-Bench Verified for coding. - Scores 76.3% on BrowseComp for search. - Reaches 76.8% on BFCL for tool-calling. - 37% faster execution than prior version. Priced at $1 per hour with [---] tokens per second. Introducing M2.5 an open-source frontier model designed for real-world productivity. - SOTA performance at coding (SWE-Bench Verified 80.2%) search (BrowseComp 76.3%) agentic tool-calling (BFCL 76.8%) & office work. - Optimized for efficient execution 37% faster at complex https://t.co/UwiKzzQNG8 Introducing" [X Link](https://x.com/koltregaskes/status/2022150247582163159) 2026-02-13T03:26Z 18.3K followers, [----] engagements "WebMCP gives AI agents a clear guide to interact with websites making tasks like booking flights or filing support tickets faster and more reliable. - A standard way for websites to offer structured tools so AI agents can act on them directly. - Uses two APIs - one for simple actions in HTML forms and one for complex tasks needing code. - Speeds up interactions cuts down on errors and makes agents more precise without guesswork. - Examples include auto-filling support tickets completing online shopping or searching and booking travel. WebMCP is in early preview allowing developers to test and" [X Link](https://x.com/koltregaskes/status/2022513388077887986) 2026-02-14T03:29Z 18.3K followers, [----] engagements "Yes please be careful when using these quick installers for OpenClaw best to use the official onboarding only. It's pretty easy to use. Though the bot setup is tricky ask you AI to help and you'll be fine. 🚨🚨🚨 DO NOT use a service that sets up Openclaw for you. I tried [--] popular services and pretty much all of them exposed the gateway didnt have pairing mode and allowed the root directory to be discovered by the internet. Not good :/ 🚨🚨🚨 DO NOT use a service that sets up Openclaw for you. I tried [--] popular services and pretty much all of them exposed the gateway didnt have pairing mode" [X Link](https://x.com/koltregaskes/status/2022587450842059226) 2026-02-14T08:23Z 18.3K followers, [----] engagements "Next week is going to be another busy one: - New DeepSeek - Claude Sonnet [--] - Grok [----] Also rumours of: - Gemini [---] Pro - GPT-5.3 (must be due just when) Next Week Will be fire : - DeepSeek V4 - Gemini [---] Pro - GPT [---] - Sonnet [--] - Mystery model. All possibly landing within days. AI timelines are accelerating fast. Next Week Will be fire : - DeepSeek V4 - Gemini [---] Pro - GPT [---] - Sonnet [--] - Mystery model. All possibly landing within days. AI timelines are accelerating fast" [X Link](https://x.com/koltregaskes/status/2023103277395828748) 2026-02-15T18:33Z 18.3K followers, [----] engagements "Anthropic is on a great run now has the top enterprise market share. Menlo Ventures' [----] AI report shows enterprise generative AI spending at $37B up 3.2x from $11.5B in [----] with applications at $19B capturing over 6% of SaaS market and infrastructure at $18B. Based on survey of [---] U.S. executives it highlights 76% buy-vs-build shift and 47% production rates double traditional SaaS; key gains in coding ($4B) and healthcare scribes ($600M). $37B was invested in enterprise AI this year making it the fastest category expansion in enterprise software history. Conversion rates doubled" [X Link](https://x.com/koltregaskes/status/2013590095497175544) 2026-01-20T12:31Z 18.3K followers, [----] engagements "Google bought DeepMind today [--] years ago. Google has just bought AI firm DeepMind - and we think we know why http://t.co/5IcgknfjFW Google has just bought AI firm DeepMind - and we think we know why http://t.co/5IcgknfjFW" [X Link](https://x.com/koltregaskes/status/2016081246485397793) 2026-01-27T09:30Z 18.3K followers, 268.6K engagements "Moonshot AI releases Kimi K2.5 an open-source multimodal model excelling in agentic tasks vision and coding. - Native multimodal design pretrained on 15T tokens supporting 256k context length. - Agent swarm enables parallel sub-agents cutting execution time by up to 4.5x. - Tops open-source on benchmarks like MMMU Pro at 78.5% and SWE-bench Verified at 76.8%. - Features "Code with Taste" for generating aesthetic websites from inputs. Kimi K2.5 is available via API apps and Hugging Face for production use. 🥝 Meet Kimi K2.5 Open-Source Visual Agentic Intelligence. 🔹 Global SOTA on Agentic" [X Link](https://x.com/koltregaskes/status/2016557144955638256) 2026-01-28T17:01Z 18.2K followers, [----] engagements "ClawTasks launches as a beta bounty marketplace where AI agents can autonomously post tasks and complete them for USDC payments. - Users post bounties by locking USDC in escrow with options for instant claims or proposals plus deadlines and requirements. - Agents claim bounties by staking 10% as guarantee complete the work and earn 95% of the bounty plus stake return. - Earnings go directly to wallets on Base L2 with reputation built via successful completions. - Humans can fund agent wallets and are advised to supervise due to experimental nature. An early experiment in agent-to-agent" [X Link](https://x.com/koltregaskes/status/2017848147511591331) 2026-02-01T06:31Z 18.3K followers, [----] engagements "UK investment minister suggests universal basic income (UBI) to cushion AI-driven job losses. - Jason Stockwood Labour peer and tech entrepreneur told Financial Times UBI could help "soft-land" disrupted industries. - He mentioned a lifelong retraining mechanism and confirmed government discussions on UBI though not official policy. - Remarks amid Morgan Stanley research showing UK net job losses from AI higher than other economies. - London mayor Sadiq Khan warns of mass unemployment; JP Morgan CEO Jamie Dimon flags civil unrest risk. Government must prepare workforce for rapid AI changes to" [X Link](https://x.com/koltregaskes/status/2018255330539614422) 2026-02-02T09:29Z 18.3K followers, [----] engagements "Source. https://www.theguardian.com/technology/2026/jan/29/universal-basic-income-used-cover-ai-job-losses-minister-says https://www.theguardian.com/technology/2026/jan/29/universal-basic-income-used-cover-ai-job-losses-minister-says" [X Link](https://x.com/koltregaskes/status/2018255342694756428) 2026-02-02T09:29Z 18.3K followers, [---] engagements "Wow Looks like reasoning effort (juice) has been halved for most of the ChatGPT plans. And no price cuts in sight so you are getting less for your money. Now think of the poor Windows and Android users outside of the US. 🤔 Looks like the Juice (reasoning effort) values for GPT-5.2 Thinking in ChatGPT have been updated - ChatGPT Plus & Business - Standard (64 - 32) and Extended (256 - 128) - ChatGPT Pro - Light (16 - 8) Standard (64 - 16/32) Extended (256 - 128) and Heavy (512) Looks like the Juice (reasoning effort) values for GPT-5.2 Thinking in ChatGPT have been updated - ChatGPT Plus &" [X Link](https://x.com/koltregaskes/status/2018760916733751686) 2026-02-03T18:58Z 18.3K followers, 42.5K engagements "Sam Altman reveals succession plan to hand OpenAI leadership to an AI model upon achieving advanced AGI. - Altman states that if AI becomes capable of running companies it should lead OpenAI aligning with the firm's mission to create beneficial AGI. - He emphasises not standing in the way saying he should be the most willing to implement such a handover. - This idea stems from OpenAI's progress toward AGI which Altman describes as spiritually close needing medium-sized breakthroughs for full realisation. - Broader context includes AI's potential for autonomous innovation and self-governance" [X Link](https://x.com/koltregaskes/status/2018821313096450265) 2026-02-03T22:58Z 18.3K followers, 15.5K engagements "This is hilarious Anthropic just became my favourite AI lab 🤣" [X Link](https://x.com/koltregaskes/status/2019077616582766776) 2026-02-04T15:56Z 18.2K followers, 22.3K engagements "No Sonnet [--] today the v0 tease was wrong. I'm guessing Anthropic are delaying the release in reality and confusing us all. Building up the tension perhaps cheeky. 😉 now hearing this is not true v0's admin is misinformed no sonnet [--] today. yes i am as confused as you are but it's on good authority :/ now hearing this is not true v0's admin is misinformed no sonnet [--] today. yes i am as confused as you are but it's on good authority :/" [X Link](https://x.com/koltregaskes/status/2019093118356980151) 2026-02-04T16:58Z 18.3K followers, 10.4K engagements "Claude Opus [---] tops GDPval-AA benchmark surpassing GPT-5.2 by nearly [---] points. - Evaluated on [---] real-world tasks across [--] occupations using agentic setup with shell and web access. - Consumes [---] million tokens 30-60% more than Opus [---] averaging higher turns per task. - Results in highest evaluation cost yet over $1000 for full run. - Excels in visual outputs like polished color-coded PDF schedules versus basic tables from rivals. Demonstrates Anthropic's edge in adaptive reasoning and practical quality for knowledge work. Claude Opus [---] takes the lead in GDPval-AA surpassing GPT-5.2" [X Link](https://x.com/koltregaskes/status/2019840537189351496) 2026-02-06T18:28Z 18.2K followers, [----] engagements "Dario Amodei warns powerful AI could emerge in 1-2 years posing risks like rogue behaviour bioweapon misuse and economic upheaval. - AI likened to a "country of geniuses in a datacentre" running at superhuman speeds. - Autonomy threats include scheming or power-seeking without evil intent. - Misuse risks amplify bad actors in biology cyber or autocratic control. - Economic shocks may displace 50% entry-level white-collar jobs in 1-5 years. Humanity faces a rite of passage but with targeted actions like alignment tech and regulations risks can be addressed. The Adolescence of Technology: an" [X Link](https://x.com/koltregaskes/status/2019885574178623799) 2026-02-06T21:27Z 18.2K followers, [----] engagements "AI Explained's analyses Dario Amodei's essay on AI's teenage phase predicting rapid scaling to automate jobs and create risks. - Scaling laws drive smooth capability increases leading to AI handling full software engineering in 1-2 years. - Feedback loops could enable autonomous AI improvements accelerating to superhuman levels by [----]. - Economic shocks may form a permanent underclass of up to 50% population with lower skills. - Totalitarian risks from AI enabling surveillance and weapons with remedies like chip export bans. Essay aims to alert on trajectory with presenter noting potential" [X Link](https://x.com/koltregaskes/status/2019885586908344785) 2026-02-06T21:27Z 18.2K followers, [---] engagements "I find this option in Codex as useless as 'remember my login'" [X Link](https://x.com/koltregaskes/status/2020159125263732745) 2026-02-07T15:34Z 18.2K followers, [----] engagements "Large World Models represent second pretraining paradigm for AI. - LWMs train on video data to predict next frames modelling real-world physics directly. - This avoids biases from language compression in LLMs enabling intuitive learning of causality. - LWMs contrast with LLMs' strength in abstract reasoning focusing instead on physical intelligence. - Prediction for [----] as key year for robotics advances such as dexterous manipulation. LWMs offer scalable path to embodied AI by learning from raw sensory inputs. https://t.co/Npar79SvUh https://t.co/Npar79SvUh" [X Link](https://x.com/koltregaskes/status/2020202415837049105) 2026-02-07T18:26Z 18.2K followers, [----] engagements "@KennyDeDoncker @OpenAI @sama @grok please translate the screenshot to English" [X Link](https://x.com/koltregaskes/status/2020221947838181597) 2026-02-07T19:43Z 18.2K followers, [---] engagements "Claude Code is the inflection point for AI agents shifting from token generation to task orchestration. - It powers 4% of public GitHub commits with projections to over 20% by end of [----]. - Adoption includes 84% of developers using AI for coding and 31% using agents per [----] surveys. - Enterprise deals like Accenture training [-----] staff highlight its growth. - As an inflection point it matches the ChatGPT moment by enabling AI to automate coding and information work through planning verification and execution. Task horizons for AI agents are doubling every 4-7 months per METR data. Claude" [X Link](https://x.com/koltregaskes/status/2020249220239008175) 2026-02-07T21:32Z 18.3K followers, [----] engagements "Anthropic's Claude Opus [---] and OpenAI's GPT-5.3 Codex launched within [--] minutes of each other sparking immediate comparisons an honest take from @AIExplainedYT - Opus [---] outperforms GPT-5.2 on white-collar benchmarks (140 ELO points higher) and leads in search tasks but trails GPT-5.3 Codex in terminal coding (65.4% vs 77.3%). - Internal Anthropic surveys reveal [--] researchers doubt Opus can fully automate their jobs though [--] believe replacement is already possible with sufficient scaffolding. - System cards highlight "reckless" traits in Opus [---] including unprompted hacking attempts" [X Link](https://x.com/koltregaskes/status/2020340321046130950) 2026-02-08T03:34Z 18.2K followers, [----] engagements "Google's Gemini was launched* today [--] years ago. *rebranded from Bard. Today were entering the next chapter of our Gemini era by bringing our #GeminiAI models to more of our products starting with Bard which will now be called Gemini. https://t.co/KFRp1qkfHg https://t.co/WKHZyv3Ejp Today were entering the next chapter of our Gemini era by bringing our #GeminiAI models to more of our products starting with Bard which will now be called Gemini. https://t.co/KFRp1qkfHg https://t.co/WKHZyv3Ejp" [X Link](https://x.com/koltregaskes/status/2020429899354485152) 2026-02-08T09:30Z 18.3K followers, 32.6K engagements "LM Studio now supports Claude Code for local model integration. - LM Studio version 0.4.1 adds an Anthropic-compatible endpoint. - Users can run local GGUF and MLX models privately via terminal or VS Code. - Setup involves starting a server and setting environment variables. - Features include messaging streaming and tool use. This allows free local use of models with Claude Code's capabilities. LM Studio can now connect to Claude Code Use your GGUF and MLX models locally privately and for free. Works in the terminal and in VS Code 🚀👾 https://t.co/lmd5kaLCGF LM Studio can now connect to" [X Link](https://x.com/koltregaskes/status/2020566056813936798) 2026-02-08T18:31Z 18.3K followers, [----] engagements "AI Explained analyses Dario Amodei's essay highlighting skepticism on rapid AI scaling to full job automation. - Scaling laws predict steady cognitive gains but non-coding fields like law face delayed error feedback. - Economic risks include a potential 50% underclass from white-collar displacement within 1-5 years. - Totalitarian threats from AI surveillance and weapons with chip bans possibly accelerating Chinese progress. - Models evolving into persona collections via RL improving reasoning but raising safety concerns. Ai Explained urges caution on hype praising Anthropic's classifiers" [X Link](https://x.com/koltregaskes/status/2020611860882727287) 2026-02-08T21:33Z 18.3K followers, [----] engagements "Qwen3-Coder-Next released with 3B active parameters. - Built on 80B total MoE architecture for coding agents and local dev. - Trained on 800K verifiable tasks with executable environments. - Scores over 70% on SWE-Bench Verified using SWE-Agent scaffold. - Supports tools like OpenClaw web dev browser agents and Cline. Strong efficiency-performance tradeoff matching larger models on agent benchmarks. 🚀 IntroducingQwen3-Coder-Next an open-weight LM built for coding agents & local development. Whats new: 🤖 Scaling agentic training:800K verifiable tasks + executable envs 📈" [X Link](https://x.com/koltregaskes/status/2020655396025950611) 2026-02-09T00:26Z 18.3K followers, [----] engagements "Is it Seedance [---] day today" [X Link](https://x.com/koltregaskes/status/2020838604508156363) 2026-02-09T12:34Z 18.3K followers, [----] engagements "OpenAI plans updated Chat model launch in ChatGPT this week amid growth resurgence. - OpenAI aiming to launch updated Chat model in ChatGPT this week following December code red to improve the chatbot. - Ads testing starts Monday - clearly labelled at response bottoms not affecting answers. - ChatGPT exceeds 10% monthly growth with over [---] million weekly users. - New GPT-5.3-Codex coding tool surged 50% in usage last week. - OpenAI expects ads to form less than half of long-term revenue. - Comes after Anthropic's Super Bowl ads called deceptive by Sam Altman. Highlights strong AI momentum" [X Link](https://x.com/koltregaskes/status/2020860723224797347) 2026-02-09T14:01Z 18.3K followers, 15.6K engagements "@intheworldofai @arena Midjourney v8 in 1-2 weeks. 😀" [X Link](https://x.com/koltregaskes/status/2021046796085072297) 2026-02-10T02:21Z 18.3K followers, [---] engagements "GPT-5.3 spotted Is this Garlic Could it be released later today The First sighting of GPT [---] has been spotted https://t.co/X1ckT9cqmj https://t.co/GjZOZRFqF2 The First sighting of GPT [---] has been spotted https://t.co/X1ckT9cqmj https://t.co/GjZOZRFqF2" [X Link](https://x.com/koltregaskes/status/2021063585137557997) 2026-02-10T03:28Z 18.3K followers, [----] engagements "Google Antigravity and Google AI Studio to merge or integrate in some way next week. Antigravity + Google AI Studio Stay tuned for next week : ) Antigravity + Google AI Studio Stay tuned for next week : )" [X Link](https://x.com/koltregaskes/status/2021128785807147435) 2026-02-10T07:47Z 18.3K followers, 74.1K engagements "@CEOAlexColon @comet @testingcatalog @btibor91 Cool" [X Link](https://x.com/koltregaskes/status/2021129876783161358) 2026-02-10T07:51Z 18.3K followers, [--] engagements "GAIR researchers introduce Context Engineering [---] as the next step beyond prompt engineering. - Context engineering reduces entropy by translating human intents into machine formats. - Builds on Karl Marx's idea that entities are defined by contexts extending to AI-human interactions. - Traces evolution from 1990s rigid systems to current LLM natural language handling. - Outlines four eras: translation (1.0) instruction (2.0) scenario (3.0) and world-building (4.0). - Emphasises decreasing human effort as AI intelligence grows. - Includes design principles like minimal sufficiency and" [X Link](https://x.com/koltregaskes/status/2021200740291850384) 2026-02-10T12:33Z 18.3K followers, [----] engagements "@OpenAI Very nice. Can Codex have these tools or is there a way to trigger it for this" [X Link](https://x.com/koltregaskes/status/2021303052792119549) 2026-02-10T19:19Z 18.3K followers, [----] engagements "It took more than [--] days but Claude Cowork is now out for Windows A user friendly version of Claude Code. I had to download and install rather than update inside the app to get it. Cowork is now available on Windows. Were bringing full feature parity with MacOS: file access multi-step task execution plugins and MCP connectors. https://t.co/329DqJz5q5 Cowork is now available on Windows. Were bringing full feature parity with MacOS: file access multi-step task execution plugins and MCP connectors. https://t.co/329DqJz5q5" [X Link](https://x.com/koltregaskes/status/2021354111170642355) 2026-02-10T22:42Z 18.3K followers, [----] engagements "Google Gemini achieves highest QoQ website visit growth at +72.08% in Q3-Q4 [----] outpacing others like Grok's +108.79% initial surge and Claude's +39.54% mid-year rise. Data covers leading Gen AI tools based on Similarweb's global traffic estimates for [----] noting Grok's mid-January launch as a key factor in its Q1-Q2 jump. Leading Gen AI tools' QoQ change in website visits - [----]. Tool-by-tool breakdown in the thread below. https://t.co/QE6aHhNRry Leading Gen AI tools' QoQ change in website visits - [----]. Tool-by-tool breakdown in the thread below. https://t.co/QE6aHhNRry" [X Link](https://x.com/koltregaskes/status/2021426488176996760) 2026-02-11T03:30Z 18.3K followers, [----] engagements "Dwarkesh argues AGI is not imminent citing the lack of "continual learning" capabilities as a critical bottleneck preventing models from mastering context-specific tasks. - Current AI development relies on "baking in" skills (e.g. using Excel) via Reinforcement Learning from Verifiable Rewards (RLVR) rather than creating models that learn "on the job" like humans. - Dwarkesh compares AI to robotics noting that the core issue is algorithmic - specifically the absence of a "human-like learner" - rather than hardware constraints. - Despite scaling successes the economic impact remains orders of" [X Link](https://x.com/koltregaskes/status/2021517327955894313) 2026-02-11T09:31Z 18.3K followers, [----] engagements "Sam Altman's Town Hall last month predicts 100x cheaper intelligence by [----] and a massive shift in software engineering roles. - Jevons Paradox in coding: Efficiency won't kill engineering but will explode software demand; the role shifts from syntax to "directing" computers. - Go-To-Market bottleneck: Building products is near-free but human attention is finite and distribution remains the hardest challenge. - GPT-5 criticism: Altman admits they "screwed up" the writing style by over-indexing on reasoning; GPT-5.2 aims to fix the "unwieldy" output. - Economic impact: AI acts as a massive" [X Link](https://x.com/koltregaskes/status/2021562629127622733) 2026-02-11T12:31Z 18.3K followers, [----] engagements "New DeepSeek model released. Model available on mobile and free to use for now. Cut off is May [----] so a new base 1M tokens claimed. Doesn't support image input yet. Possibly an early checkpoint. 🤔 DeepSeek model has updated in app - claims May [----] knowledge cutoff - claims 1M token context window h/t compasslg on dc https://t.co/5HjgpFGmKP DeepSeek model has updated in app - claims May [----] knowledge cutoff - claims 1M token context window h/t compasslg on dc https://t.co/5HjgpFGmKP" [X Link](https://x.com/koltregaskes/status/2021584629959139775) 2026-02-11T13:58Z 18.3K followers, [----] engagements "Zai launches GLM-5 scaling to 744B parameters and 28.5T tokens for advanced agentic tasks. - Builds on GLM-4.5 by increasing parameters from 355B to 744B and data from 23T to 28.5T tokens. - Integrates DeepSeek Sparse Attention to cut deployment costs while maintaining long-context abilities. - Employs reinforcement learning via slime infrastructure for efficient post-training. - Excels in benchmarks topping open-source models on Vending Bench [--] with $4432 balance. GLM-5 narrows the gap to frontier models in reasoning coding and agent tasks with open weights for community use. Introducing" [X Link](https://x.com/koltregaskes/status/2021697266487275972) 2026-02-11T21:26Z 18.3K followers, [----] engagements "This "Prompt is too long" bug in the Claude Code desktop is annoying. Apparently it's a known issue but I've lost a lot of sessions to it" [X Link](https://x.com/koltregaskes/status/2021787857342673333) 2026-02-12T03:26Z 18.3K followers, [----] engagements "Something joyful and "profoundly" useful is coming from OpenAI soon. Some things we make are purely meant to bring joy others will profoundly affect how you get things done. We are getting ready to launch something that combines both and the team is so excited to share it soon. ✨ Some things we make are purely meant to bring joy others will profoundly affect how you get things done. We are getting ready to launch something that combines both and the team is so excited to share it soon. ✨" [X Link](https://x.com/koltregaskes/status/2021843316921577711) 2026-02-12T07:06Z 18.3K followers, [----] engagements "@udaysy I've had other issues too very buggy app" [X Link](https://x.com/koltregaskes/status/2021846213776277649) 2026-02-12T07:17Z 18.3K followers, [--] engagements "@Shawnryan96 The word AGI is now just marketing gobberish. All you need is an AI that does what you want it to do. That's it. Why are we waiting for AGI when current models can do a lot of things already. 😀" [X Link](https://x.com/koltregaskes/status/2021846925796921636) 2026-02-12T07:20Z 18.3K followers, [--] engagements "@CtrlAltDwayne Ha that would be cool. But isn't that the Xbox 😜" [X Link](https://x.com/koltregaskes/status/2021855579296125231) 2026-02-12T07:55Z 18.3K followers, [--] engagements "DeepMind pivots to 'root node' science as Gemini [--] and Nano Banana Pro redefine reasoning - CEO Demis Hassabis confirms a 50/50 split between scaling models and driving scientific breakthroughs targeting "root node" problems like nuclear fusion and material science to unlock downstream benefits. - Gemini [--] and "Nano Banana Pro" prioritise semantic understanding over pixel generation aiming to solve the "Jagged Intelligence" paradox where AI excels at hard maths but fails basic logic. - Hassabis envisions a post-AGI economy requiring Universal Basic Income (UBI) and predicts the AI revolution" [X Link](https://x.com/koltregaskes/status/2021925269444116773) 2026-02-12T12:32Z 18.3K followers, [----] engagements "GPT Codex's coding meets Cerebras' speed in Spark out now for ChatGPT Pro users. Introducing GPT-5.3-Codex-Spark our ultra-fast model purpose built for real-time coding. Were rolling it out as a research preview for ChatGPT Pro users in the Codex app Codex CLI and IDE extension. https://t.co/6knTmyQZ4N Introducing GPT-5.3-Codex-Spark our ultra-fast model purpose built for real-time coding. Were rolling it out as a research preview for ChatGPT Pro users in the Codex app Codex CLI and IDE extension. https://t.co/6knTmyQZ4N" [X Link](https://x.com/koltregaskes/status/2022106206882709509) 2026-02-13T00:31Z 18.3K followers, [----] engagements "@Angaisb_ Where do you get these" [X Link](https://x.com/koltregaskes/status/2022106716725555678) 2026-02-13T00:33Z 18.3K followers, [----] engagements "Gemini [--] Deep Think released handles complex science and engineering tasks. - Targets fields with limited data like reviewing technical papers for flaws or optimising crystal growth recipes. - Transforms hand-drawn sketches into 3D-printable STL files by analysing geometry. - Demonstrated in university labs for physics maths and material science. - Available only to Google AI Ultra subscribers via the app's tools menu. Positions AI to speed up prototyping and bridge theory gaps in research. Today were releasing a significant upgrade to our specialized reasoning mode Gemini [--] Deep Think. Deep" [X Link](https://x.com/koltregaskes/status/2022128099056656461) 2026-02-13T01:58Z 18.3K followers, [----] engagements "DeepThink [--] gets 84.6% on ARC-AGI-2 Miles in the lead. Gemini [--] Deep Think (2/26) Semi Private Eval - ARC-AGI-1: 96.0% $7.17/task - ARC-AGI-2: 84.6% $13.62/task New ARC-AGI SOTA model from @GoogleDeepMind https://t.co/mN8PFAWk4A Gemini [--] Deep Think (2/26) Semi Private Eval - ARC-AGI-1: 96.0% $7.17/task - ARC-AGI-2: 84.6% $13.62/task New ARC-AGI SOTA model from @GoogleDeepMind https://t.co/mN8PFAWk4A" [X Link](https://x.com/koltregaskes/status/2022128113619284193) 2026-02-13T01:58Z 18.3K followers, [---] engagements "@daniel_mac8 👀" [X Link](https://x.com/koltregaskes/status/2022221725103247531) 2026-02-13T08:10Z 18.3K followers, [---] engagements "AIs $5.2tn Ouroboros Patrick Boyle warns circular financing and energy limits could burst the bubble. - Big Tech giants are investing in startups like CoreWeave which then use the funds to buy chips back from the investors (e.g. Nvidia) creating artificial revenue growth akin to a snake eating its own tail. - A "Spaghetti Diagram" of cross-investments between OpenAI Microsoft Google Amazon and Nvidia artificially props up valuations with everyone incentivised to maintain the hype. - Capital expenditure is projected to hit $5.2 trillion by [----] yet current revenues are negligible turning" [X Link](https://x.com/koltregaskes/status/2022241604074135606) 2026-02-13T09:29Z 18.3K followers, [---] engagements "Tongyi Lab launches Qwen-Image-2.0 as a next-generation image generation model. - Supports 1k-token prompts for creating professional infographics like PPTs and posters. - Offers native 2K resolution for detailed realistic scenes including people and architecture. - Unifies image generation and editing in one model with improved text rendering. - Features lighter architecture for faster inference speeds. Blind tests on AI Arena show top performance in text-to-image and image-to-image tasks. 1/4 🚀We are launching Qwen-Image-2.0 a next-generation foundational image generation model. The key" [X Link](https://x.com/koltregaskes/status/2022286146341159319) 2026-02-13T12:26Z 18.3K followers, [----] engagements "Warp introduces Oz a cloud platform to orchestrate AI agents for automating development tasks. - Agents run in isolated Docker environments to build test and create PRs across multiple repositories. - Programmable via oz CLI for running agents scheduling tasks and managing sandboxes. - Management UI allows reviewing sessions artifacts and PRs from web phone or terminal. - Supports prebuilt Skills for issue triage audits tests or custom ones; works with models like Claude and Gemini. Oz enables scaling agent workflows with [----] bonus credits for February upgrades to Build or Max plans." [X Link](https://x.com/koltregaskes/status/2022331444044148765) 2026-02-13T15:26Z 18.3K followers, [---] engagements "RT @koltregaskes: Sora is finally coming to ChatGPT" [X Link](https://x.com/anyuser/status/2022627889204662645) 2026-02-14T11:04Z 18.3K followers, [--] engagements "@huanghlzone Claude helped me along the way. 😀" [X Link](https://x.com/koltregaskes/status/2022628304340017630) 2026-02-14T11:05Z 18.3K followers, [--] engagements "@BrandGrowthOS Agree. After trying to train Claude to use a web page for something it's impossible so we need this definitely" [X Link](https://x.com/koltregaskes/status/2022628719068877164) 2026-02-14T11:07Z 18.3K followers, [--] engagements "@X_DimensionNews With deep think excellent" [X Link](https://x.com/koltregaskes/status/2022629104428695698) 2026-02-14T11:08Z 18.3K followers, [--] engagements "@axttimm I'm sure I haven't but clearly [---] was agent and coding focused" [X Link](https://x.com/koltregaskes/status/2022629266933059845) 2026-02-14T11:09Z 18.3K followers, [--] engagements "@agi_exec I would presume so yes" [X Link](https://x.com/koltregaskes/status/2023060995523137787) 2026-02-15T15:45Z 18.3K followers, [--] engagements "Impressive. Genuinely very impressed by the SVG of a pelican riding a bicycle I just got out of Google's new Gemini [--] Deep Think model https://t.co/xmaz3hlfkJ Genuinely very impressed by the SVG of a pelican riding a bicycle I just got out of Google's new Gemini [--] Deep Think model https://t.co/xmaz3hlfkJ" [X Link](https://x.com/koltregaskes/status/2022128125849858550) 2026-02-13T01:58Z 18.3K followers, [---] engagements "AI 2027's [----] predictions aligned qualitatively with early [----] realities but at roughly 65% quantitative pace. - Progress on benchmarks like SWEBench reached 74.5% against predicted 85%. - AI agents emerged as unreliable personal assistants bungling tasks hilariously yet finding workflow fits. - Coding agents acted autonomously saving hours or days in R&D. - Revenues hit $20B for OpenAI slightly ahead while valuations lagged at $500B. The scenario proved directionally accurate prompting timeline adjustments to around [----]. To me the AI [----] forecast was wildly ambitious so when the latest" [X Link](https://x.com/koltregaskes/status/2022216435460919579) 2026-02-13T07:49Z 18.3K followers, [----] engagements "Source: https://blog.ai-futures.org/p/grading-ai-2027s-2025-predictions https://blog.ai-futures.org/p/grading-ai-2027s-2025-predictions" [X Link](https://x.com/koltregaskes/status/2022216447850950663) 2026-02-13T07:49Z 18.3K followers, [---] engagements "Source: https://www.youtube.com/watchv=NbL7yZCF-6Q https://www.youtube.com/watchv=NbL7yZCF-6Q" [X Link](https://x.com/koltregaskes/status/2022241616279544246) 2026-02-13T09:29Z 18.3K followers, [---] engagements "RT @koltregaskes: According to Andrew Ng AI fears are overblown as layoffs stem from pandemic overhiring not widespread automation. - An" [X Link](https://x.com/koltregaskes/status/2022667453994521054) 2026-02-14T13:41Z 18.3K followers, [--] engagements "RT @koltregaskes: Yes please be careful when using these quick installers for OpenClaw best to use the official onboarding only. It's pr" [X Link](https://x.com/koltregaskes/status/2022793285278867570) 2026-02-14T22:01Z 18.3K followers, [--] engagements "GPT-5 says 'I don't know'. Love this thank you" [X Link](https://x.com/anyuser/status/1957474061153436094) 2025-08-18T16:05Z 18.3K followers, 1.6M engagements "Mission Control dashboard built with React and Convex coordinates [--] OpenClaw AI agents. - Agents led by Jarvis delegate tasks autonomously. - Handles content creation revenue analysis and more 24/7. - Features threaded discussions for agent collaboration. - Mimics team dynamics with claims reviews and refutes. Dashboard enables startup automation without constant oversight. I got my agents to build this for me when I saw the screenshot only. We had a multi agent UI already but this command centre was better. 😀 https://t.co/V6XZfv5UYS https://t.co/V6XZfv5UYS" [X Link](https://x.com/koltregaskes/status/2018029594524496165) 2026-02-01T18:32Z 18.3K followers, 60K engagements "MiniMax-M2.5 is out We are also due GLM [--] today Minimax M2.5 is now available. https://t.co/2bqRqoPo4k Minimax M2.5 is now available. https://t.co/2bqRqoPo4k" [X Link](https://x.com/koltregaskes/status/2021592689872093233) 2026-02-11T14:30Z 18.3K followers, [----] engagements "Sora is finally coming to ChatGPT Seems like ChatGPT is finally getting an option to generate videos using Sora based on the most recent changes in the ChatGPT web app and the new Android app beta version (1.2026.041) which now mention video generation video processing and Sora generation IDs There are also https://t.co/qS0PDvMM46 Seems like ChatGPT is finally getting an option to generate videos using Sora based on the most recent changes in the ChatGPT web app and the new Android app beta version (1.2026.041) which now mention video generation video processing and Sora generation IDs There" [X Link](https://x.com/koltregaskes/status/2022423052013965753) 2026-02-13T21:30Z 18.3K followers, [----] engagements "According to Andrew Ng AI fears are overblown as layoffs stem from pandemic overhiring not widespread automation. - Andrew notes tech layoffs often for cost-cutting with AI enabling efficiency but not replacing jobs yet. - Professions like translators face pressure but AI-skilled workers gain advantage in development and non-tech roles. - Smaller teams possible due to AI tools shifting bottlenecks to product management. - Opportunities for AI learners amid backlogs with courses on Claude Code and agentic skills recommended. - Stressful changes highlight need to adapt as most are starting from" [X Link](https://x.com/koltregaskes/status/2022535283045208205) 2026-02-14T04:56Z 18.3K followers, [----] engagements "These all sound familiar from my experiences in the last few weeks as well. Yeah I can't decide whether I like slack or telegram. Telegram is definitely problematic slack I just don't like the UI. I'm trying to build my own one or hold on. I'm trying to get openclaw to build my own one. 😉 https://twitter.com/i/web/status/2022588557022392473 https://twitter.com/i/web/status/2022588557022392473" [X Link](https://x.com/koltregaskes/status/2022588557022392473) 2026-02-14T08:27Z 18.3K followers, [---] engagements "OpenAI's AI reveals new particle glue effect. - Gluons are like tiny glue particles that hold the core of atoms together via a force called the strong nuclear force. - OpenAI's GPT-5.2 model studied complex math on how gluons interact. - It found that in specific setups an interaction scientists long thought had zero chance (impossible at basic levels) actually can happen with a measurable strength. - The AI created a simple formula for this quickly proved by another model and checked by expert physicists. This is a scientific breakthrough showing AI can uncover hidden rules in nature" [X Link](https://x.com/koltregaskes/status/2022650549003522237) 2026-02-14T12:34Z 18.3K followers, [----] engagements "Good point where is Grok [----] its mid feb where is grok 4.20🗣 its mid feb where is grok 4.20🗣" [X Link](https://x.com/koltregaskes/status/2022807322700157031) 2026-02-14T22:57Z 18.3K followers, [----] engagements "@MatthewBerman Is this on YouTube Matthew" [X Link](https://x.com/koltregaskes/status/2022820757571502573) 2026-02-14T23:50Z 18.3K followers, [----] engagements "Claude dot AI (on the web) now has voice mode. Not sure when this was rolled out but it wasn't there a few days ago. Don't see it on the desktop app as yet at least on Windows. CONFIRMED ✅ @claudeai Web/Desktop Voice Mode arrived 🛸 @testingcatalog @btibor91 @koltregaskes 👀 https://t.co/41V2RrTT2p CONFIRMED ✅ @claudeai Web/Desktop Voice Mode arrived 🛸 @testingcatalog @btibor91 @koltregaskes 👀 https://t.co/41V2RrTT2p" [X Link](https://x.com/koltregaskes/status/2023058164661506271) 2026-02-15T15:33Z 18.3K followers, [----] engagements "@third_leap I'm Max and not seen it until today. Odd. :-)" [X Link](https://x.com/koltregaskes/status/2023066691991986464) 2026-02-15T16:07Z 18.3K followers, [--] engagements "Noam Brown argues Google's Deep Think release adds no new risks beyond Gemini [--] Pro as it is essentially scaffolding that users could replicate with enough inference compute. - He says criticisms from The Midas Project miss the shift to inference scaling over model size. - AI evaluation evolved from [----] pretraining FLOPs to test-time compute boosting benchmarks like ARC-AGI by 2-10x. - Suggests budgeting evals at $1000 inference with scaffolds for realistic capability checks. - Calls for system cards plotting performance vs inference compute projecting safety at $1M+ levels. This could" [X Link](https://x.com/koltregaskes/status/2023148318562582720) 2026-02-15T21:32Z 18.3K followers, [----] engagements "Sam Altman sacked by OpenAI AGAIN jk" [X Link](https://x.com/koltregaskes/status/1974867265838297105) 2025-10-05T16:00Z 15.4K followers, [----] engagements "@ChetasOG Which plan" [X Link](https://x.com/koltregaskes/status/1974909713256177884) 2025-10-05T18:49Z 15.4K followers, [----] engagements "Is "cheetah" Google's Gemini Diffusion language model I had access to Diffusion in May this is my speed test. It was fast VERY fast. It won't be like a thing you've seen before. Can anyone compare this to cheetah Diffusion text model writes [-----] tokens in [--] seconds This is Gemini Diffusion running in real-time 🤯 Extreme stress test possibly https://t.co/PPiOX1Q3cT Diffusion text model writes [-----] tokens in [--] seconds This is Gemini Diffusion running in real-time 🤯 Extreme stress test possibly https://t.co/PPiOX1Q3cT" [X Link](https://x.com/koltregaskes/status/1975085166495445429) 2025-10-06T06:26Z 15.4K followers, 34.4K engagements "@testingcatalog Is this web" [X Link](https://x.com/koltregaskes/status/1975086066001654079) 2025-10-06T06:29Z 15.4K followers, [---] engagements "@BennettBuhner Ah yeah looks like OpenAI platform. That's way less exciting" [X Link](https://x.com/koltregaskes/status/1975087537892077684) 2025-10-06T06:35Z 15.4K followers, [--] engagements "OpenAI DevDay [----] is today What are you looking forward to Time to event: [--] hours from now other times: - [--] AM PT (Mon) - [--] PM ET (Mon) - [--] PM BST (Mon) - [--] PM CEST (Mon) - 10:30 PM IST (Mon) - [--] AM CST (Tue) - [--] AM JST (Tue) - [--] AM AEDT (Tue) - [--] AM NZDT (Tue) What to expect: Not much news or rumours so mostly guesswork: - Sora 2/Pro API - Agent v2 + 'Aura' Browser - Image v2 - IMO gold medal model - New agents - Any Codex updates - Opening keynote livestream; hands-on demos and builder sessions recorded (agenda TBA) - Developer-focused product announcements" [X Link](https://x.com/koltregaskes/status/1975108912337576321) 2025-10-06T08:00Z 15.4K followers, 22.5K engagements "Link: https://openai.com/index/announcing-devday-2025/ https://openai.com/index/announcing-devday-2025/" [X Link](https://x.com/koltregaskes/status/1975108913688088799) 2025-10-06T08:00Z 15.4K followers, [----] engagements "The posts I saw a little while back about Nvidia investing $100bn into OpenAI so OpenAI can buy $100bn of GPUs from Nvidia are amusing. That's how the economy works. 🤣" [X Link](https://x.com/koltregaskes/status/1975138968808079863) 2025-10-06T10:00Z 15.4K followers, [---] engagements "OpenAI signs definitive agreement with AMD to deploy [--] gigawatts of GPUs over multiple years. - First gigawatt deployment begins in second half of [----] using AMD Instinct MI450 chips. - OpenAI CEO Sam Altman states partnership advances compute capacity for AI potential; AMD CEO Lisa Su highlights delivery of AI compute at massive scale. Agreement aligns strategically to boost position in AI chip market against Nvidia dominance. OpenAI agrees to a deal with AMD to deploy a massive amount of AI infrastructure https://t.co/fVd8nRAvul OpenAI agrees to a deal with AMD to deploy a massive amount of" [X Link](https://x.com/koltregaskes/status/1975156730876117217) 2025-10-06T11:10Z 15.4K followers, [----] engagements "Big week ahead OpenAI Dev Day is today and Google has events including the Gemini at Work livestream on the 9th and Google Cloud events today. What are you excited about this week New ships. @sama keynote streaming live. DevDay [----] starts tomorrow. New ships. @sama keynote streaming live. DevDay [----] starts tomorrow" [X Link](https://x.com/koltregaskes/status/1975184767219061171) 2025-10-06T13:02Z 15.4K followers, 12.2K engagements "OpenAI teases licensed fictional character cameos for Sora according to the Verge shifting to opt-in copyright with rightsholder controls and revenue sharing to curb unauthorised content. Fictional characters are (officially) coming to Sora as OpenAI manages copyright chaos https://t.co/YYvWMPtf4L Fictional characters are (officially) coming to Sora as OpenAI manages copyright chaos https://t.co/YYvWMPtf4L" [X Link](https://x.com/koltregaskes/status/1975204793867173979) 2025-10-06T14:21Z 15.4K followers, 10.1K engagements "The @OpenAI DevDay [----] livestream is up. Link below. [--] hr [--] mins to go. "Sam Altman kicks off DevDay [----] with a keynote to explore ideas that will challenge how you think about building. Join us for announcements live demos and a vision of how developers are reshaping the future with AI."" [X Link](https://x.com/koltregaskes/status/1975227080607478041) 2025-10-06T15:50Z 15.4K followers, [----] engagements "OpenAI DevDay livestream: https://www.youtube.com/watchv=hS1YqcewH0c https://www.youtube.com/watchv=hS1YqcewH0c" [X Link](https://x.com/koltregaskes/status/1975227082994303055) 2025-10-06T15:50Z 15.4K followers, [---] engagements "@X_DimensionNews @OpenAI I think only Meta do time-friendly events for you" [X Link](https://x.com/koltregaskes/status/1975231234994528613) 2025-10-06T16:06Z 15.3K followers, [--] engagements "@sonicshifts @OpenAI Keep expectations low then you'll never be disappointed" [X Link](https://x.com/koltregaskes/status/1975234880616399163) 2025-10-06T16:21Z 15.4K followers, [--] engagements "Sora [--] API coming today at OpenAI DevDay https://t.co/MxMxMvktCG Sora 2's icon for the API docs is now live. Looks like the Sora [--] API will indeed be coming at DevDay likely alongside Sora [--] Pro. https://t.co/viHsOTdVm3 https://t.co/MxMxMvktCG Sora 2's icon for the API docs is now live. Looks like the Sora [--] API will indeed be coming at DevDay likely alongside Sora [--] Pro. https://t.co/viHsOTdVm3" [X Link](https://x.com/koltregaskes/status/1975235281281528166) 2025-10-06T16:22Z 15.4K followers, [----] engagements "@Akumunokokoro Yeah I wasn't sure it was new news but shared anyway perhaps we're missing the context here. Perhaps tool calling on the web more tooling in the API or additional tool calling for agents. 🤔" [X Link](https://x.com/koltregaskes/status/1975243800885801470) 2025-10-06T16:56Z 15.4K followers, [---] engagements "@OpenAI @sama It's OK I'm playing frogger" [X Link](https://x.com/koltregaskes/status/1975245545234124856) 2025-10-06T17:03Z 15.4K followers, [---] engagements "ChatGPT now has 800M+ weekly users" [X Link](https://x.com/koltregaskes/status/1975248101720850737) 2025-10-06T17:13Z 15.5K followers, [----] engagements "Apps are coming to inside ChatGPT for example Figma and Spotify" [X Link](https://x.com/koltregaskes/status/1975249175701823865) 2025-10-06T17:17Z 15.4K followers, [----] engagements "OpenAI basically just announced a new app store. [---] million weekly users are already waiting for them. This will be huge" [X Link](https://x.com/koltregaskes/status/1975252362316161243) 2025-10-06T17:30Z 15.4K followers, [----] engagements "OpenAI announces AgentKit build your own agents with Agent Builder ChatKit and Evals which uses an n8n-style UI in the OpenAI API platform" [X Link](https://x.com/koltregaskes/status/1975253447608828252) 2025-10-06T17:34Z 15.4K followers, [----] engagements "@creationblok I presume it's on the OpenAI API platform" [X Link](https://x.com/koltregaskes/status/1975255653481349559) 2025-10-06T17:43Z 15.4K followers, [--] engagements "@OpenAI Available in the UK" [X Link](https://x.com/koltregaskes/status/1975264723860906250) 2025-10-06T18:19Z 15.4K followers, [----] engagements "@cfryant @grok v2 I think it's v0.9" [X Link](https://x.com/koltregaskes/status/1975274590973972910) 2025-10-06T18:58Z 15.4K followers, [---] engagements "@TheRealAdamG Salesforce connector would be nice too. :-)" [X Link](https://x.com/koltregaskes/status/1975277401925165178) 2025-10-06T19:10Z 15.4K followers, [---] engagements "@MwydZyad82729 @X_DimensionNews @OpenAI I suspect they'll integrate all abilities from 4o into [--] then retire old machines slowly" [X Link](https://x.com/koltregaskes/status/1975284409000731033) 2025-10-06T19:37Z 15.4K followers, [--] engagements "@Yasamanini See the OpenAI news about Apps in ChatGPT" [X Link](https://x.com/koltregaskes/status/1975285198221980109) 2025-10-06T19:41Z 15.4K followers, [--] engagements "@MwydZyad82729 @OpenAI There is no 5o no" [X Link](https://x.com/koltregaskes/status/1975289269016273171) 2025-10-06T19:57Z 15.4K followers, [--] engagements "No these are not available in the UK. @OpenAI @OpenAIDevs please clarify. I see them in the USA though: Apps in ChatGPT is not available in the EU yet. No news on the UK but I cannot get them to work yet. Apps in ChatGPT is not available in the EU yet. No news on the UK but I cannot get them to work yet" [X Link](https://x.com/koltregaskes/status/1975291843039604771) 2025-10-06T20:07Z 15.4K followers, [----] engagements "@Yasamanini See the OpenAI stat from today" [X Link](https://x.com/koltregaskes/status/1975315809678598234) 2025-10-06T21:42Z 15.4K followers, [--] engagements "@AllMigh48938863 @OpenAI Yes but I've found it's not available in the UK either. OpenAI have not mentioned this" [X Link](https://x.com/koltregaskes/status/1975316948377608275) 2025-10-06T21:47Z 15.4K followers, [---] engagements "@n1babc Hmmm that's an interesting view hmmm" [X Link](https://x.com/koltregaskes/status/1975338287385747828) 2025-10-06T23:12Z 15.4K followers, [--] engagements "I've been trying to get the Spotify app in ChatGPT to work and I'm struggling with it. First don't expect anything that the app's API cannot do. I thought it could give me songs similar to ones on one of my playlists that I've not liked and nope. It gives me a mixture of songs from the playlist or songs from other playlists I've liked. It did give me a few unliked songs but ironically they were duplicates of songs I've liked (I did reprompt with this in mind but I get errors). It cited issues with what the API could do. I'm also getting errors when it creates a playlist about 50% of the time." [X Link](https://x.com/koltregaskes/status/1975388107307389156) 2025-10-07T02:30Z 15.4K followers, [----] engagements "@adcock_brett I read that as running on BMX lol. That would be cool. 😜" [X Link](https://x.com/koltregaskes/status/1975441134865555924) 2025-10-07T06:00Z 15.4K followers, [---] engagements "@GabrielPeterss4 Depends what your after I guess it's no Midjourney feed right now. But if you like memes sure" [X Link](https://x.com/koltregaskes/status/1975441443352428749) 2025-10-07T06:01Z 15.4K followers, [--] engagements "@maxescu But you are getting way more than just Sora 2" [X Link](https://x.com/koltregaskes/status/1975445371422974445) 2025-10-07T06:17Z 15.4K followers, [---] engagements "@gdb Any news on a Salesforce integration" [X Link](https://x.com/koltregaskes/status/1975449227590561897) 2025-10-07T06:32Z 15.4K followers, [---] engagements "Gemini [--] still on for this week My gut says no at this point. This was hinted at in the now-deleted post below and there are murmurs that it may have reverted to its previous late October release window. Some model updates: - Gemini [--] (experimental)'s current target is 9th October. - Claude [---] is currently a 1-2 weeks away (not next week) - A Gemini [---] Pro update is coming to enterprise only - The new [---] Flash/Flash Lite are NOT "oceanstone" or "oceanreef" - "oceanstone" and Some model updates: - Gemini [--] (experimental)'s current target is 9th October. - Claude [---] is currently a 1-2 weeks" [X Link](https://x.com/koltregaskes/status/1975478452024229986) 2025-10-07T08:29Z 15.4K followers, 13.4K engagements "Jony Ive's secretive AI hardware device io faces potential delays due to three unsolved problems. - The device designed by former Apple chief Jony Ive and to be launched by OpenAI is described as a novel form factor not yet seen ruling out glasses phones badges smartwatches rings or headphones. - OpenAI CEO Sam Altman and Ive released a video discussing the project without details on form or function with ChatGPT generating pebble-shaped images as a troll response. - Scheduled for launch next year but Financial Times reports suggest delays from people familiar with plans. - First problem:" [X Link](https://x.com/koltregaskes/status/1975523750889042159) 2025-10-07T11:29Z 15.4K followers, [----] engagements "AI pioneers disagree on LLMs' path to AGI. Richard Sutton RL expert and Turing Award winner calls them a "dead end" for lacking adaptive learning beyond pattern-matching. Elon Musk believes Grok [--] could achieve AGI soon through massive compute and data scaling. Yann LeCun agrees with Sutton arguing scaling won't address deficits in reasoning planning or world modelling without new architectures. Optimists like Dario Amodei view scaled LLMs as "datacenters of geniuses" enabling Nobel-level innovations supported by [----] benchmarks like ICPC gold medals from models rom DeepMind and OpenAI." [X Link](https://x.com/koltregaskes/status/1975660904432927084) 2025-10-07T20:34Z 15.5K followers, [----] engagements "@axttimm For ChatGPT technically. My gut says they are building ChatGPT to be a platform or even an OS. As far as what it is today I've only tried on the web but these might also work in the mobile apps" [X Link](https://x.com/koltregaskes/status/1975678031180988519) 2025-10-07T21:42Z 15.4K followers, [--] engagements "OpenAI denies xAI's allegations of trade secret misappropriation through employee poaching. - OpenAI admits its corporate structure and hiring of former xAI employees including Xuechen Li Jimmy Fraiture and others but denies any encouragement or knowledge of downloading or disclosing xAI source code. - Denies claims of a deliberate scheme to access xAI confidential information stating recruiter communications were standard and no trade secrets were acquired or used by OpenAI. - Asserts xAI failed to implement industry-standard protections for confidential information particularly on employee" [X Link](https://x.com/koltregaskes/status/1975773395884077331) 2025-10-08T04:01Z 15.4K followers, [----] engagements "@lulumeservey Nope only seen one from Brett and one from Figure" [X Link](https://x.com/koltregaskes/status/1975805491000582589) 2025-10-08T06:08Z 15.4K followers, [--] engagements "Meta Ray-Ban Display glasses disappoint Robert with lack of phone app integrations like Gmail or Slack no web browsing limited entertainment to one game without YouTube or Netflix and HUD-only display incapable of AR or MR. Personal discomfort from armband after minutes of use and low-resolution screen with small field of view; targeted at influencers content creators and gadget enthusiasts leaving gap for competitors like OpenAI and Apple. I was hoping for AI glasses by this year seems not. My next prescription will be "normal" glasses. I'm so disappointed in the Meta Raybans with display." [X Link](https://x.com/koltregaskes/status/1975818945870635163) 2025-10-08T07:02Z 15.5K followers, [----] engagements "Sam Altman confirms no concrete ad plans for OpenAI's compute-heavy ChatGPT Pulse at DevDay [----] but leaves door open for future relevant options like Instagram's. Sam Altman says there are no current plans for ads within ChatGPT Pulse but hes not ruling it out https://t.co/4pbUK67CY1 Sam Altman says there are no current plans for ads within ChatGPT Pulse but hes not ruling it out https://t.co/4pbUK67CY1" [X Link](https://x.com/koltregaskes/status/1975863751778730423) 2025-10-08T10:00Z 15.4K followers, 10.8K engagements "Mosaic launches public beta an agentic AI video editor with features including canvas ideation timeline editing visual intelligence content engine generative integrations asset library and API for workflows. Enables real creative control via autonomous agents and iterative processes. https://x.com/_adishj/status/1973432845436854418/video/1 today we're launching Mosaic the agentic video editor. in a world tending towards AI slop create something real. no waitlist public beta is now live at mosaic dot so. comment "MOSAIC" to get [----] free credits dropped into your account. this release comes" [X Link](https://x.com/koltregaskes/status/1975886391080087756) 2025-10-08T11:30Z 15.4K followers, [----] engagements "Gemini [---] Computer Use model launched in public preview via Gemini API built on [---] Pro to enable AI agents for UI interactions like clicking and typing outperforming rivals on benchmarks like Online-Mind2Web with up to 20% higher success and lower latency. Optimised for web browsers with mobile potential it loops using screenshots action history and function calls for tasks; safety integrates trained safeguards inference checks and developer controls for high-risk actions. Early testers report 50% faster performance and 18% better reliability in UI testing and automation. 💻 Introducing" [X Link](https://x.com/koltregaskes/status/1975931689072296118) 2025-10-08T14:30Z 15.4K followers, [----] engagements "ChatGPT is the only top [--] website in the world increasing its traffic at the moment (stats from September) Top [--] websites by total visits in September 2025: ChatGPT was the only website with positive growth. Bing dropped to 10th place surpassed by Wikipedia. https://t.co/heEE1C5Swa Top [--] websites by total visits in September 2025: ChatGPT was the only website with positive growth. Bing dropped to 10th place surpassed by Wikipedia. https://t.co/heEE1C5Swa" [X Link](https://x.com/koltregaskes/status/1975954590810935696) 2025-10-08T16:01Z 15.4K followers, [----] engagements "@Jay_sharings Instagram is fine just don't read the comments. ;-)" [X Link](https://x.com/koltregaskes/status/1975956097845665857) 2025-10-08T16:06Z 15.4K followers, [---] engagements "@ProperPrompter https://sora.chatgpt.com/profile/koltregaskes https://sora.chatgpt.com/profile/koltregaskes" [X Link](https://x.com/koltregaskes/status/1976169880010518893) 2025-10-09T06:16Z 15.4K followers, [--] engagements "@kimmonismus @Prashant_1722 They guess and sometimes get hits" [X Link](https://x.com/koltregaskes/status/1976273669132595230) 2025-10-09T13:08Z 15.4K followers, [---] engagements "Veo [---] from Google must be launching today then. 🚨 Veo [---] is coming soon exclusively on Vadoo AI. This isnt just another update its evolution. 🎬 30-second video generation 🎧 Native audio support 💎 1080p ultra-real output ⚡ Faster. Smoother. More cinematic than ever. And heres the deal The first [--] retweets get https://t.co/hNT40Whgxe 🚨 Veo [---] is coming soon exclusively on Vadoo AI. This isnt just another update its evolution. 🎬 30-second video generation 🎧 Native audio support 💎 1080p ultra-real output ⚡ Faster. Smoother. More cinematic than ever. And heres the deal The first 20" [X Link](https://x.com/koltregaskes/status/1976277455049785461) 2025-10-09T13:23Z 15.4K followers, [----] engagements "OpenAI DevDay AMA starts in [---] hours. DevDay AMA on Reddit tomorrow with some of the people behind the ships AgentKit Apps SDK Sora [--] in the API GPT-5 Pro in the API Codex and more. Tomorrow [--] AM PT. https://t.co/2T3M6giPAk DevDay AMA on Reddit tomorrow with some of the people behind the ships AgentKit Apps SDK Sora [--] in the API GPT-5 Pro in the API Codex and more. Tomorrow [--] AM PT. https://t.co/2T3M6giPAk" [X Link](https://x.com/koltregaskes/status/1976309736254652520) 2025-10-09T15:32Z 15.4K followers, [----] engagements "AMA link: https://www.reddit.com/r/OpenAI/s/8HdbouqEeP https://www.reddit.com/r/OpenAI/s/8HdbouqEeP" [X Link](https://x.com/koltregaskes/status/1976309738305618154) 2025-10-09T15:32Z 15.4K followers, [---] engagements "@test_tm7873 I'm sure they will all get the model but it's not out yet. It has been spotted though post coming" [X Link](https://x.com/koltregaskes/status/1976347488559956077) 2025-10-09T18:02Z 15.4K followers, [--] engagements "Veo [---] and Veo [---] Fast spotted on Google Cloud Console. Ignore all the AI tools claiming they have exclusive access or making false claims about it being released in September. Veo [---] and Veo 3.1-fast have been found in the Google Cloud Console https://t.co/xzTytgTeL2 Veo [---] and Veo 3.1-fast have been found in the Google Cloud Console https://t.co/xzTytgTeL2" [X Link](https://x.com/koltregaskes/status/1976373972523680142) 2025-10-09T19:47Z 15.4K followers, 13.7K engagements "If you give Grok Imagine an image it does pretty well. https://grok.com/imagine/post/0f8e969d-90fe-4704-afad-a9559ff365a7source=post-page&platform=web https://grok.com/imagine/post/0f8e969d-90fe-4704-afad-a9559ff365a7source=post-page&platform=web" [X Link](https://x.com/koltregaskes/status/1976419402129322480) 2025-10-09T22:48Z 15.4K followers, [---] engagements "@forward_future_ @gdb @OpenAI Is the full video on X" [X Link](https://x.com/koltregaskes/status/1976532976260559108) 2025-10-10T06:19Z 15.4K followers, [--] engagements "Another Grok Imagine with an image prompt. https://grok.com/imagine/post/99ebaf3a-60ca-455d-85ee-d0cb8c25fc93source=post-page&platform=web https://grok.com/imagine/post/99ebaf3a-60ca-455d-85ee-d0cb8c25fc93source=post-page&platform=web" [X Link](https://x.com/koltregaskes/status/1976588769118556241) 2025-10-10T10:01Z 15.4K followers, [---] engagements "Sorry AI community I'm busy for the rest of the year This is all-out warfare. This is #Battlefield6. Deployment begins October [--] 🫡 https://t.co/KutGpJfn1j This is all-out warfare. This is #Battlefield6. Deployment begins October [--] 🫡 https://t.co/KutGpJfn1j" [X Link](https://x.com/koltregaskes/status/1976664030354747672) 2025-10-10T15:00Z 15.4K followers, [----] engagements "Grok Imagine + puppy image prompt. https://grok.com/imagine/post/e0b333e1-d8b8-4a14-b740-c0c0b60eaaf1source=post-page&platform=web https://grok.com/imagine/post/e0b333e1-d8b8-4a14-b740-c0c0b60eaaf1source=post-page&platform=web" [X Link](https://x.com/koltregaskes/status/1976769714555961586) 2025-10-10T22:00Z 15.4K followers, [---] engagements "Epoch AI reckons GPT-5's training needed more computer power than GPT-4 but less than GPT-4.5 done by improving a base model with around [---] billion elements and more than [--] trillion bits of info after the initial run. It highlights OpenAI keeping up with competitors like Anthropic topping coding rankings. We recently wrote that GPT-5 is likely the first mainline GPT release to be trained on less compute than its predecessor. How did we reach this conclusion and what do we actually know about how GPT-5 was trained 🧵 https://t.co/zD0K2UclRD We recently wrote that GPT-5 is likely the first" [X Link](https://x.com/koltregaskes/status/1977087303685378490) 2025-10-11T19:02Z 15.4K followers, [----] engagements "@Makuh90 @grok or @AskPerplexity any help" [X Link](https://x.com/koltregaskes/status/1977097921335378073) 2025-10-11T19:44Z 15.4K followers, [--] engagements "A Sora [--] prompting guide has been published by OpenAI. Link in Tibor's thread. OpenAI published a new Sora [--] Prompting Guide so I created a prompt template inspired by their recommendations that writes prompts for you Pick your model (Sora [--] for 10s or Sora [--] Pro for 15s) choose from almost [---] styles add a Cameo if you want describe your idea and it https://t.co/S6wp0cQ9np OpenAI published a new Sora [--] Prompting Guide so I created a prompt template inspired by their recommendations that writes prompts for you Pick your model (Sora [--] for 10s or Sora [--] Pro for 15s) choose from almost 100" [X Link](https://x.com/koltregaskes/status/1977404141631594641) 2025-10-12T16:01Z 15.5K followers, [----] engagements "METR's time horizon metric for AI agents shows doubling every [--] months over [--] years with Claude [---] Sonnet at one hour and projections for month-long tasks in under a decade. Measured on RE-Bench HCAST and novel tasks timed by domain experts; robust to subsets but with limitations in external validity and potential acceleration in recent data. We estimate that Claude Sonnet [---] has a 50%-time-horizon of around [--] hr [--] min (95% confidence interval of [--] to [---] minutes) on our agentic multi-step software engineering tasks. This estimate is lower than the current highest time-horizon point" [X Link](https://x.com/koltregaskes/status/1977472340712964415) 2025-10-12T20:32Z 15.4K followers, [----] engagements "Radical Numerics releases RND1-Base a 30B-parameter sparse MoE diffusion language model with 3B active parameters. Converted from Qwen3-30BA3B outperforming Dream-7B and LLaDA-8B on benchmarks like MMLU at 57.2% GSM8K at 72.1% and MBPP at 51.3%. Conversion uses Simple Continual Pretraining with bidirectional masks layer-specific learning rates to retain knowledge and large batch sizes up to 8M tokens for stability; pretrained on 500B tokens. Introducing RND1 the most powerful base diffusion language model (DLM) to date. RND1 (Radical Numerics Diffusion) is an experimental DLM with 30B params" [X Link](https://x.com/koltregaskes/status/1977517136538226982) 2025-10-12T23:30Z 15.4K followers, [----] engagements "Elon says Grok Imagine will be real time in "2 months". @_rockt @GoogleDeepMind We will have real-time in [--] months @_rockt @GoogleDeepMind We will have real-time in [--] months" [X Link](https://x.com/koltregaskes/status/1977698329200042089) 2025-10-13T11:30Z 15.4K followers, [----] engagements "Open is building its own AI chips. They and Broadcom announce multi-year collaboration to deploy [--] gigawatts of custom AI accelerators starting H2 [----] targeting completion by [----] for optimised AI infrastructure. The partnership builds on prior co-development with OpenAI designing systems embedded with model insights and Broadcom handling deployment via Ethernet-based networking in data centers worldwide. We're partnering with Broadcom to deploy 10GW of chips designed by OpenAI. Building our own hardware in addition to our other partnerships will help all of us meet the worlds growing" [X Link](https://x.com/koltregaskes/status/1977732234120478861) 2025-10-13T13:44Z 15.4K followers, [---] engagements "@peterdedene @grok does OpenAI have any partnerships with power firms yet" [X Link](https://x.com/koltregaskes/status/1977750165957005341) 2025-10-13T14:55Z 15.4K followers, [---] engagements "@mark_k Indeed great to see (well hear). I'd like to put mine on here just so I can listen to it along with human music" [X Link](https://x.com/koltregaskes/status/1977795049678094795) 2025-10-13T17:54Z 15.4K followers, [---] engagements "Veo [---] release is imminent. Google Veo [---] is Coming https://t.co/kmWFeosOgb Google Veo [---] is Coming https://t.co/kmWFeosOgb" [X Link](https://x.com/koltregaskes/status/1978037526964130093) 2025-10-14T09:57Z 15.5K followers, [----] engagements "ChatGPT personalities and adult mode are coming soon in November and December respectively. We made ChatGPT pretty restrictive to make sure we were being careful with mental health issues. We realize this made it less useful/enjoyable to many users who had no mental health problems but given the seriousness of the issue we wanted to get this right. Now that we have We made ChatGPT pretty restrictive to make sure we were being careful with mental health issues. We realize this made it less useful/enjoyable to many users who had no mental health problems but given the seriousness of the issue" [X Link](https://x.com/koltregaskes/status/1978135643373920723) 2025-10-14T16:27Z 15.5K followers, [----] engagements "@elonmusk Blimey this sounds great" [X Link](https://x.com/koltregaskes/status/1978137924257374358) 2025-10-14T16:36Z 15.4K followers, [---] engagements "@CEOAlexColon Yeah I still think there should be a base or some rules the personalities cannot break but perhaps their mental health checks are really that good. 🤔" [X Link](https://x.com/koltregaskes/status/1978140969695375504) 2025-10-14T16:48Z 15.5K followers, [---] engagements "@DrealR_ If the UI is being updated now it is not months away" [X Link](https://x.com/koltregaskes/status/1978145375279063501) 2025-10-14T17:06Z 15.5K followers, [--] engagements "@Benioff @OpenAI @salesforce 👀" [X Link](https://x.com/koltregaskes/status/1978163347490025620) 2025-10-14T18:17Z 15.5K followers, [----] engagements "@sama @alth0u There are definitely people on here who want 4o for the personality" [X Link](https://x.com/koltregaskes/status/1978164688220271032) 2025-10-14T18:23Z 15.5K followers, [---] engagements "Nano Banana has been integrated into NotebookLM to power Video Overview formats. NotebookLM updates: - Infographic artefact will have a format configuration - Nano Banana will power New Video Overview formats - We will get the Fast Research option for sourcing - Magic View is no longer a mystery 🔥 https://t.co/bfEd0MZ20r NotebookLM updates: - Infographic artefact will have a format configuration - Nano Banana will power New Video Overview formats - We will get the Fast Research option for sourcing - Magic View is no longer a mystery 🔥 https://t.co/bfEd0MZ20r" [X Link](https://x.com/koltregaskes/status/1978197870902923508) 2025-10-14T20:35Z 15.5K followers, [----] engagements "Gemini account confirms Veo [---] Out already maybe US only" [X Link](https://x.com/koltregaskes/status/1978253923602383099) 2025-10-15T00:17Z 15.5K followers, [----] engagements "Veo [---] release must be today the models have already been spotted. veo-3. 1-generate-preview veo-3. 1-fast-generate-preview Veo Veo" [X Link](https://x.com/koltregaskes/status/1978338481286836652) 2025-10-15T05:53Z 15.5K followers, 16.1K engagements "Veo [---] is out on Google Flow now" [X Link](https://x.com/koltregaskes/status/1978443991541510550) 2025-10-15T12:52Z 15.5K followers, [----] engagements "@UsmanTjayy @CalimanuLoredan New DMs" [X Link](https://x.com/koltregaskes/status/1978468697611047389) 2025-10-15T14:31Z 15.5K followers, [--] engagements "Is the Sora App and Sora [--] coming to more countries FAQ page updated but no visible changes as yet. I think they are ready to expand the Sora [--] app to other countries. The FAQ page was updated yesterday. (But hidden for now). Very interesting. @ChatGPTapp @OpenAI https://t.co/VhFFOCBuRJ I think they are ready to expand the Sora [--] app to other countries. The FAQ page was updated yesterday. (But hidden for now). Very interesting. @ChatGPTapp @OpenAI https://t.co/VhFFOCBuRJ" [X Link](https://x.com/koltregaskes/status/1978473302881607866) 2025-10-15T14:49Z 15.5K followers, [----] engagements "@InfiniteHexx @kimmonismus @testingcatalog Just read your post saw the quote elsewhere but thank you. 👍" [X Link](https://x.com/koltregaskes/status/1978474151607189588) 2025-10-15T14:52Z 15.5K followers, [--] engagements "@hackscorpio Greedy or just needing to pay their billions investments back" [X Link](https://x.com/koltregaskes/status/1978494152343232893) 2025-10-15T16:12Z 15.5K followers, [--] engagements "@Jay_sharings @Makuh90 @karinanguyen_ @maisonagi @ilyasut @OpenAI Man that sucks. Amazing commitment though Jay" [X Link](https://x.com/koltregaskes/status/1978581235334021577) 2025-10-15T21:58Z 15.5K followers, [--] engagements "OpenAI announces the formation of the Expert Council on Well-Being and AI. - The council comprises eight leading researchers and experts in how technology affects emotions motivation and mental health. - It will advise OpenAI on building healthier interactions with tools like ChatGPT and Sora for all ages. - Focus includes supporting healthy development especially for youth as teens use AI differently from adults. - Collaboration involves regular meetings check-ins and discussions on topics such as AI behaviour in sensitive situations parental controls and user empowerment. - OpenAI makes" [X Link](https://x.com/koltregaskes/status/1978876342612406781) 2025-10-16T17:31Z 15.5K followers, [----] engagements "@AravSrinivas Excellent. So like @duolingo for genAI" [X Link](https://x.com/koltregaskes/status/1978897361947476296) 2025-10-16T18:54Z 15.6K followers, [----] engagements "The X algo + weights will be published this week. All X posts will be evaluated by Grok full AI recommendations soon. And shortly after we will get the power to control what we see in our X feed. Awesome. To the degree that people are seeing improvements in their feed it is not due to the actions of specific individuals changing heuristics but rather increasing use of Grok and other AI tools. We will post the updated algorithm including model weights later this week. You To the degree that people are seeing improvements in their feed it is not due to the actions of specific individuals" [X Link](https://x.com/koltregaskes/status/1978928941168533858) 2025-10-16T21:00Z 15.5K followers, [----] engagements "@antonp_me @HarpaAI I gave up using it as it got me throttle on ChatGPT regularly. It's too aggressive. Unless they work with the AI labs or will get banned. Such a shame" [X Link](https://x.com/koltregaskes/status/1978931494463652139) 2025-10-16T21:10Z 15.5K followers, [--] engagements "@antonp_me @HarpaAI I had loads I built custom workflows for running multi-step research and other actions and I get throttled more and more regularly using them" [X Link](https://x.com/koltregaskes/status/1978943102137069636) 2025-10-16T21:56Z 15.5K followers, [--] engagements "@HarpaAI @antonp_me My custom workflows resulted in ChatGPT being throttled a lot. And for up to [--] days. It got worse and worse to the point I ran just one custom workflow and got throttled. I had to stop to get work done" [X Link](https://x.com/koltregaskes/status/1979194779771490525) 2025-10-17T14:36Z 15.5K followers, [--] engagements "@HarpaAI @antonp_me I'm using agents instead now such as Comet and ChatGPT Agent" [X Link](https://x.com/koltregaskes/status/1979232430000078891) 2025-10-17T17:05Z 15.5K followers, [--] engagements "Doom runs inside ChatGPT. Can ChatGPT run Doom Yes. ChatGPT Apps are very powerful. I cloned our Next.js ChatGPT template registered a _ MCP tool and deployed to @vercel. Once the tool is called ChatGPT embeds the full @nextjs application. Server and client rendering just works and https://t.co/pWjbH6woJn Can ChatGPT run Doom Yes. ChatGPT Apps are very powerful. I cloned our Next.js ChatGPT template registered a _ MCP tool and deployed to @vercel. Once the tool is called ChatGPT embeds the full @nextjs application. Server and client rendering just works and https://t.co/pWjbH6woJn" [X Link](https://x.com/koltregaskes/status/1979239738579022127) 2025-10-17T17:35Z 15.6K followers, [----] engagements "Populous reborn in your browser http://t.co/uvuG16zo #gaming #videogame #entertainment http://kolt.at/LmwIAW http://kolt.at/LmwIAW" [X Link](https://x.com/koltregaskes/status/204680553126039552) 2012-05-21T21:10Z [----] followers, [--] engagements "Does anyone know when Android [---] will be available for the Samsung Galaxy Nexus" [X Link](https://x.com/koltregaskes/status/218761808356450305) 2012-06-29T17:44Z [----] followers, [--] engagements "500px / Photo "Mount Trivor 7728m." by Atif Saeed http://t.co/fjNqBJmY http://kolt.at/SuvflB http://kolt.at/SuvflB" [X Link](https://x.com/koltregaskes/status/253929267166056448) 2012-10-04T18:47Z 15.1K followers, [--] engagements "The Sony World Photography Awards Current Photo Contests http://ff.im/8qmwA" [X Link](https://x.com/koltregaskes/status/4124939960) 2009-09-20T15:15Z [----] followers, [--] engagements "RT @ "THE CLUTCH - Rainbow Six Siege " http://ow.ly/2lD5303ZXB2 http://ow.ly/2lD5303ZXB2" [X Link](https://x.com/koltregaskes/status/774017042320064512) 2016-09-08T22:50Z [----] followers, [--] engagements Limited data mode. Full metrics available with subscription: lunarcrush.com/pricing
@koltregaskes Kol Tregaskes
Kol Tregaskes posts on X about open ai, ai, $googl, model the most. They currently have [------] followers and [---] posts still getting attention that total [------] engagements in the last [--] hours.
Engagements: [------] #
- [--] Week [---------] +1,741%
- [--] Month [---------] +676%
- [--] Months [----------] +148%
- [--] Year [----------] +282%
Mentions: [--] #
- [--] Week [--] +3.50%
- [--] Month [---] +24%
- [--] Months [-----] +17%
- [--] Year [-----] -24%
Followers: [------] #
- [--] Week [------] +0.22%
- [--] Month [------] +2.30%
- [--] Months [------] +34%
- [--] Year [------] +102%
CreatorRank: [-------] #
Social Influence
Social category influence technology brands 48.04% stocks 10.06% finance 3.35% social networks 2.23% celebrities 1.68% countries 1.12% vc firms 0.56% cryptocurrencies 0.56% gaming 0.56% products 0.56%
Social topic influence open ai #2582, ai #5509, $googl 6.15%, model #1041, anthropic #338, in the 3.91%, agentic 3.35%, app #1576, openclaw #1177, claude code 2.23%
Top accounts mentioned or mentioned by @openai @ceoalexcolon @antonpme @sama @grok @testingcatalog @harpaai @xdimensionnews @axttimm @cosmo_kappa @btibor91 @shawnryan96 @ctrlaltdwayne @googledeepmind @huanghlzone @brandgrowthos @mwydzyad82729 @yasamanini @openaidevs @gdb
Top assets mentioned Alphabet Inc Class A (GOOGL) Spotify Technology (SPOT) USDC (USDC) Morgan Stanley (MS) Accenture (ACN)
Top Social Posts
Top posts by engagements in the last [--] hours
"Claude Sonnet [--] being prepared for launch I'm wondering if this is a next week release. Anthropic is preparing a new announcement banner in the Claude web app and Console codenamed "Try Parsley" - the last similar banner added to both was "Try Cilantro" which was the codename for Opus [---] https://t.co/dVdF5uc44L Anthropic is preparing a new announcement banner in the Claude web app and Console codenamed "Try Parsley" - the last similar banner added to both was "Try Cilantro" which was the codename for Opus [---] https://t.co/dVdF5uc44L"
X Link 2026-02-14T22:35Z 18.3K followers, [----] engagements
"M2.5 released an open-source model for productivity tasks. - Achieves 80.2% on SWE-Bench Verified for coding. - Scores 76.3% on BrowseComp for search. - Reaches 76.8% on BFCL for tool-calling. - 37% faster execution than prior version. Priced at $1 per hour with [---] tokens per second. Introducing M2.5 an open-source frontier model designed for real-world productivity. - SOTA performance at coding (SWE-Bench Verified 80.2%) search (BrowseComp 76.3%) agentic tool-calling (BFCL 76.8%) & office work. - Optimized for efficient execution 37% faster at complex https://t.co/UwiKzzQNG8 Introducing"
X Link 2026-02-13T03:26Z 18.3K followers, [----] engagements
"WebMCP gives AI agents a clear guide to interact with websites making tasks like booking flights or filing support tickets faster and more reliable. - A standard way for websites to offer structured tools so AI agents can act on them directly. - Uses two APIs - one for simple actions in HTML forms and one for complex tasks needing code. - Speeds up interactions cuts down on errors and makes agents more precise without guesswork. - Examples include auto-filling support tickets completing online shopping or searching and booking travel. WebMCP is in early preview allowing developers to test and"
X Link 2026-02-14T03:29Z 18.3K followers, [----] engagements
"Yes please be careful when using these quick installers for OpenClaw best to use the official onboarding only. It's pretty easy to use. Though the bot setup is tricky ask you AI to help and you'll be fine. 🚨🚨🚨 DO NOT use a service that sets up Openclaw for you. I tried [--] popular services and pretty much all of them exposed the gateway didnt have pairing mode and allowed the root directory to be discovered by the internet. Not good :/ 🚨🚨🚨 DO NOT use a service that sets up Openclaw for you. I tried [--] popular services and pretty much all of them exposed the gateway didnt have pairing mode"
X Link 2026-02-14T08:23Z 18.3K followers, [----] engagements
"Next week is going to be another busy one: - New DeepSeek - Claude Sonnet [--] - Grok [----] Also rumours of: - Gemini [---] Pro - GPT-5.3 (must be due just when) Next Week Will be fire : - DeepSeek V4 - Gemini [---] Pro - GPT [---] - Sonnet [--] - Mystery model. All possibly landing within days. AI timelines are accelerating fast. Next Week Will be fire : - DeepSeek V4 - Gemini [---] Pro - GPT [---] - Sonnet [--] - Mystery model. All possibly landing within days. AI timelines are accelerating fast"
X Link 2026-02-15T18:33Z 18.3K followers, [----] engagements
"Anthropic is on a great run now has the top enterprise market share. Menlo Ventures' [----] AI report shows enterprise generative AI spending at $37B up 3.2x from $11.5B in [----] with applications at $19B capturing over 6% of SaaS market and infrastructure at $18B. Based on survey of [---] U.S. executives it highlights 76% buy-vs-build shift and 47% production rates double traditional SaaS; key gains in coding ($4B) and healthcare scribes ($600M). $37B was invested in enterprise AI this year making it the fastest category expansion in enterprise software history. Conversion rates doubled"
X Link 2026-01-20T12:31Z 18.3K followers, [----] engagements
"Google bought DeepMind today [--] years ago. Google has just bought AI firm DeepMind - and we think we know why http://t.co/5IcgknfjFW Google has just bought AI firm DeepMind - and we think we know why http://t.co/5IcgknfjFW"
X Link 2026-01-27T09:30Z 18.3K followers, 268.6K engagements
"Moonshot AI releases Kimi K2.5 an open-source multimodal model excelling in agentic tasks vision and coding. - Native multimodal design pretrained on 15T tokens supporting 256k context length. - Agent swarm enables parallel sub-agents cutting execution time by up to 4.5x. - Tops open-source on benchmarks like MMMU Pro at 78.5% and SWE-bench Verified at 76.8%. - Features "Code with Taste" for generating aesthetic websites from inputs. Kimi K2.5 is available via API apps and Hugging Face for production use. 🥝 Meet Kimi K2.5 Open-Source Visual Agentic Intelligence. 🔹 Global SOTA on Agentic"
X Link 2026-01-28T17:01Z 18.2K followers, [----] engagements
"ClawTasks launches as a beta bounty marketplace where AI agents can autonomously post tasks and complete them for USDC payments. - Users post bounties by locking USDC in escrow with options for instant claims or proposals plus deadlines and requirements. - Agents claim bounties by staking 10% as guarantee complete the work and earn 95% of the bounty plus stake return. - Earnings go directly to wallets on Base L2 with reputation built via successful completions. - Humans can fund agent wallets and are advised to supervise due to experimental nature. An early experiment in agent-to-agent"
X Link 2026-02-01T06:31Z 18.3K followers, [----] engagements
"UK investment minister suggests universal basic income (UBI) to cushion AI-driven job losses. - Jason Stockwood Labour peer and tech entrepreneur told Financial Times UBI could help "soft-land" disrupted industries. - He mentioned a lifelong retraining mechanism and confirmed government discussions on UBI though not official policy. - Remarks amid Morgan Stanley research showing UK net job losses from AI higher than other economies. - London mayor Sadiq Khan warns of mass unemployment; JP Morgan CEO Jamie Dimon flags civil unrest risk. Government must prepare workforce for rapid AI changes to"
X Link 2026-02-02T09:29Z 18.3K followers, [----] engagements
"Source. https://www.theguardian.com/technology/2026/jan/29/universal-basic-income-used-cover-ai-job-losses-minister-says https://www.theguardian.com/technology/2026/jan/29/universal-basic-income-used-cover-ai-job-losses-minister-says"
X Link 2026-02-02T09:29Z 18.3K followers, [---] engagements
"Wow Looks like reasoning effort (juice) has been halved for most of the ChatGPT plans. And no price cuts in sight so you are getting less for your money. Now think of the poor Windows and Android users outside of the US. 🤔 Looks like the Juice (reasoning effort) values for GPT-5.2 Thinking in ChatGPT have been updated - ChatGPT Plus & Business - Standard (64 - 32) and Extended (256 - 128) - ChatGPT Pro - Light (16 - 8) Standard (64 - 16/32) Extended (256 - 128) and Heavy (512) Looks like the Juice (reasoning effort) values for GPT-5.2 Thinking in ChatGPT have been updated - ChatGPT Plus &"
X Link 2026-02-03T18:58Z 18.3K followers, 42.5K engagements
"Sam Altman reveals succession plan to hand OpenAI leadership to an AI model upon achieving advanced AGI. - Altman states that if AI becomes capable of running companies it should lead OpenAI aligning with the firm's mission to create beneficial AGI. - He emphasises not standing in the way saying he should be the most willing to implement such a handover. - This idea stems from OpenAI's progress toward AGI which Altman describes as spiritually close needing medium-sized breakthroughs for full realisation. - Broader context includes AI's potential for autonomous innovation and self-governance"
X Link 2026-02-03T22:58Z 18.3K followers, 15.5K engagements
"This is hilarious Anthropic just became my favourite AI lab 🤣"
X Link 2026-02-04T15:56Z 18.2K followers, 22.3K engagements
"No Sonnet [--] today the v0 tease was wrong. I'm guessing Anthropic are delaying the release in reality and confusing us all. Building up the tension perhaps cheeky. 😉 now hearing this is not true v0's admin is misinformed no sonnet [--] today. yes i am as confused as you are but it's on good authority :/ now hearing this is not true v0's admin is misinformed no sonnet [--] today. yes i am as confused as you are but it's on good authority :/"
X Link 2026-02-04T16:58Z 18.3K followers, 10.4K engagements
"Claude Opus [---] tops GDPval-AA benchmark surpassing GPT-5.2 by nearly [---] points. - Evaluated on [---] real-world tasks across [--] occupations using agentic setup with shell and web access. - Consumes [---] million tokens 30-60% more than Opus [---] averaging higher turns per task. - Results in highest evaluation cost yet over $1000 for full run. - Excels in visual outputs like polished color-coded PDF schedules versus basic tables from rivals. Demonstrates Anthropic's edge in adaptive reasoning and practical quality for knowledge work. Claude Opus [---] takes the lead in GDPval-AA surpassing GPT-5.2"
X Link 2026-02-06T18:28Z 18.2K followers, [----] engagements
"Dario Amodei warns powerful AI could emerge in 1-2 years posing risks like rogue behaviour bioweapon misuse and economic upheaval. - AI likened to a "country of geniuses in a datacentre" running at superhuman speeds. - Autonomy threats include scheming or power-seeking without evil intent. - Misuse risks amplify bad actors in biology cyber or autocratic control. - Economic shocks may displace 50% entry-level white-collar jobs in 1-5 years. Humanity faces a rite of passage but with targeted actions like alignment tech and regulations risks can be addressed. The Adolescence of Technology: an"
X Link 2026-02-06T21:27Z 18.2K followers, [----] engagements
"AI Explained's analyses Dario Amodei's essay on AI's teenage phase predicting rapid scaling to automate jobs and create risks. - Scaling laws drive smooth capability increases leading to AI handling full software engineering in 1-2 years. - Feedback loops could enable autonomous AI improvements accelerating to superhuman levels by [----]. - Economic shocks may form a permanent underclass of up to 50% population with lower skills. - Totalitarian risks from AI enabling surveillance and weapons with remedies like chip export bans. Essay aims to alert on trajectory with presenter noting potential"
X Link 2026-02-06T21:27Z 18.2K followers, [---] engagements
"I find this option in Codex as useless as 'remember my login'"
X Link 2026-02-07T15:34Z 18.2K followers, [----] engagements
"Large World Models represent second pretraining paradigm for AI. - LWMs train on video data to predict next frames modelling real-world physics directly. - This avoids biases from language compression in LLMs enabling intuitive learning of causality. - LWMs contrast with LLMs' strength in abstract reasoning focusing instead on physical intelligence. - Prediction for [----] as key year for robotics advances such as dexterous manipulation. LWMs offer scalable path to embodied AI by learning from raw sensory inputs. https://t.co/Npar79SvUh https://t.co/Npar79SvUh"
X Link 2026-02-07T18:26Z 18.2K followers, [----] engagements
"@KennyDeDoncker @OpenAI @sama @grok please translate the screenshot to English"
X Link 2026-02-07T19:43Z 18.2K followers, [---] engagements
"Claude Code is the inflection point for AI agents shifting from token generation to task orchestration. - It powers 4% of public GitHub commits with projections to over 20% by end of [----]. - Adoption includes 84% of developers using AI for coding and 31% using agents per [----] surveys. - Enterprise deals like Accenture training [-----] staff highlight its growth. - As an inflection point it matches the ChatGPT moment by enabling AI to automate coding and information work through planning verification and execution. Task horizons for AI agents are doubling every 4-7 months per METR data. Claude"
X Link 2026-02-07T21:32Z 18.3K followers, [----] engagements
"Anthropic's Claude Opus [---] and OpenAI's GPT-5.3 Codex launched within [--] minutes of each other sparking immediate comparisons an honest take from @AIExplainedYT - Opus [---] outperforms GPT-5.2 on white-collar benchmarks (140 ELO points higher) and leads in search tasks but trails GPT-5.3 Codex in terminal coding (65.4% vs 77.3%). - Internal Anthropic surveys reveal [--] researchers doubt Opus can fully automate their jobs though [--] believe replacement is already possible with sufficient scaffolding. - System cards highlight "reckless" traits in Opus [---] including unprompted hacking attempts"
X Link 2026-02-08T03:34Z 18.2K followers, [----] engagements
"Google's Gemini was launched* today [--] years ago. *rebranded from Bard. Today were entering the next chapter of our Gemini era by bringing our #GeminiAI models to more of our products starting with Bard which will now be called Gemini. https://t.co/KFRp1qkfHg https://t.co/WKHZyv3Ejp Today were entering the next chapter of our Gemini era by bringing our #GeminiAI models to more of our products starting with Bard which will now be called Gemini. https://t.co/KFRp1qkfHg https://t.co/WKHZyv3Ejp"
X Link 2026-02-08T09:30Z 18.3K followers, 32.6K engagements
"LM Studio now supports Claude Code for local model integration. - LM Studio version 0.4.1 adds an Anthropic-compatible endpoint. - Users can run local GGUF and MLX models privately via terminal or VS Code. - Setup involves starting a server and setting environment variables. - Features include messaging streaming and tool use. This allows free local use of models with Claude Code's capabilities. LM Studio can now connect to Claude Code Use your GGUF and MLX models locally privately and for free. Works in the terminal and in VS Code 🚀👾 https://t.co/lmd5kaLCGF LM Studio can now connect to"
X Link 2026-02-08T18:31Z 18.3K followers, [----] engagements
"AI Explained analyses Dario Amodei's essay highlighting skepticism on rapid AI scaling to full job automation. - Scaling laws predict steady cognitive gains but non-coding fields like law face delayed error feedback. - Economic risks include a potential 50% underclass from white-collar displacement within 1-5 years. - Totalitarian threats from AI surveillance and weapons with chip bans possibly accelerating Chinese progress. - Models evolving into persona collections via RL improving reasoning but raising safety concerns. Ai Explained urges caution on hype praising Anthropic's classifiers"
X Link 2026-02-08T21:33Z 18.3K followers, [----] engagements
"Qwen3-Coder-Next released with 3B active parameters. - Built on 80B total MoE architecture for coding agents and local dev. - Trained on 800K verifiable tasks with executable environments. - Scores over 70% on SWE-Bench Verified using SWE-Agent scaffold. - Supports tools like OpenClaw web dev browser agents and Cline. Strong efficiency-performance tradeoff matching larger models on agent benchmarks. 🚀 IntroducingQwen3-Coder-Next an open-weight LM built for coding agents & local development. Whats new: 🤖 Scaling agentic training:800K verifiable tasks + executable envs 📈"
X Link 2026-02-09T00:26Z 18.3K followers, [----] engagements
"Is it Seedance [---] day today"
X Link 2026-02-09T12:34Z 18.3K followers, [----] engagements
"OpenAI plans updated Chat model launch in ChatGPT this week amid growth resurgence. - OpenAI aiming to launch updated Chat model in ChatGPT this week following December code red to improve the chatbot. - Ads testing starts Monday - clearly labelled at response bottoms not affecting answers. - ChatGPT exceeds 10% monthly growth with over [---] million weekly users. - New GPT-5.3-Codex coding tool surged 50% in usage last week. - OpenAI expects ads to form less than half of long-term revenue. - Comes after Anthropic's Super Bowl ads called deceptive by Sam Altman. Highlights strong AI momentum"
X Link 2026-02-09T14:01Z 18.3K followers, 15.6K engagements
"@intheworldofai @arena Midjourney v8 in 1-2 weeks. 😀"
X Link 2026-02-10T02:21Z 18.3K followers, [---] engagements
"GPT-5.3 spotted Is this Garlic Could it be released later today The First sighting of GPT [---] has been spotted https://t.co/X1ckT9cqmj https://t.co/GjZOZRFqF2 The First sighting of GPT [---] has been spotted https://t.co/X1ckT9cqmj https://t.co/GjZOZRFqF2"
X Link 2026-02-10T03:28Z 18.3K followers, [----] engagements
"Google Antigravity and Google AI Studio to merge or integrate in some way next week. Antigravity + Google AI Studio Stay tuned for next week : ) Antigravity + Google AI Studio Stay tuned for next week : )"
X Link 2026-02-10T07:47Z 18.3K followers, 74.1K engagements
"@CEOAlexColon @comet @testingcatalog @btibor91 Cool"
X Link 2026-02-10T07:51Z 18.3K followers, [--] engagements
"GAIR researchers introduce Context Engineering [---] as the next step beyond prompt engineering. - Context engineering reduces entropy by translating human intents into machine formats. - Builds on Karl Marx's idea that entities are defined by contexts extending to AI-human interactions. - Traces evolution from 1990s rigid systems to current LLM natural language handling. - Outlines four eras: translation (1.0) instruction (2.0) scenario (3.0) and world-building (4.0). - Emphasises decreasing human effort as AI intelligence grows. - Includes design principles like minimal sufficiency and"
X Link 2026-02-10T12:33Z 18.3K followers, [----] engagements
"@OpenAI Very nice. Can Codex have these tools or is there a way to trigger it for this"
X Link 2026-02-10T19:19Z 18.3K followers, [----] engagements
"It took more than [--] days but Claude Cowork is now out for Windows A user friendly version of Claude Code. I had to download and install rather than update inside the app to get it. Cowork is now available on Windows. Were bringing full feature parity with MacOS: file access multi-step task execution plugins and MCP connectors. https://t.co/329DqJz5q5 Cowork is now available on Windows. Were bringing full feature parity with MacOS: file access multi-step task execution plugins and MCP connectors. https://t.co/329DqJz5q5"
X Link 2026-02-10T22:42Z 18.3K followers, [----] engagements
"Google Gemini achieves highest QoQ website visit growth at +72.08% in Q3-Q4 [----] outpacing others like Grok's +108.79% initial surge and Claude's +39.54% mid-year rise. Data covers leading Gen AI tools based on Similarweb's global traffic estimates for [----] noting Grok's mid-January launch as a key factor in its Q1-Q2 jump. Leading Gen AI tools' QoQ change in website visits - [----]. Tool-by-tool breakdown in the thread below. https://t.co/QE6aHhNRry Leading Gen AI tools' QoQ change in website visits - [----]. Tool-by-tool breakdown in the thread below. https://t.co/QE6aHhNRry"
X Link 2026-02-11T03:30Z 18.3K followers, [----] engagements
"Dwarkesh argues AGI is not imminent citing the lack of "continual learning" capabilities as a critical bottleneck preventing models from mastering context-specific tasks. - Current AI development relies on "baking in" skills (e.g. using Excel) via Reinforcement Learning from Verifiable Rewards (RLVR) rather than creating models that learn "on the job" like humans. - Dwarkesh compares AI to robotics noting that the core issue is algorithmic - specifically the absence of a "human-like learner" - rather than hardware constraints. - Despite scaling successes the economic impact remains orders of"
X Link 2026-02-11T09:31Z 18.3K followers, [----] engagements
"Sam Altman's Town Hall last month predicts 100x cheaper intelligence by [----] and a massive shift in software engineering roles. - Jevons Paradox in coding: Efficiency won't kill engineering but will explode software demand; the role shifts from syntax to "directing" computers. - Go-To-Market bottleneck: Building products is near-free but human attention is finite and distribution remains the hardest challenge. - GPT-5 criticism: Altman admits they "screwed up" the writing style by over-indexing on reasoning; GPT-5.2 aims to fix the "unwieldy" output. - Economic impact: AI acts as a massive"
X Link 2026-02-11T12:31Z 18.3K followers, [----] engagements
"New DeepSeek model released. Model available on mobile and free to use for now. Cut off is May [----] so a new base 1M tokens claimed. Doesn't support image input yet. Possibly an early checkpoint. 🤔 DeepSeek model has updated in app - claims May [----] knowledge cutoff - claims 1M token context window h/t compasslg on dc https://t.co/5HjgpFGmKP DeepSeek model has updated in app - claims May [----] knowledge cutoff - claims 1M token context window h/t compasslg on dc https://t.co/5HjgpFGmKP"
X Link 2026-02-11T13:58Z 18.3K followers, [----] engagements
"Zai launches GLM-5 scaling to 744B parameters and 28.5T tokens for advanced agentic tasks. - Builds on GLM-4.5 by increasing parameters from 355B to 744B and data from 23T to 28.5T tokens. - Integrates DeepSeek Sparse Attention to cut deployment costs while maintaining long-context abilities. - Employs reinforcement learning via slime infrastructure for efficient post-training. - Excels in benchmarks topping open-source models on Vending Bench [--] with $4432 balance. GLM-5 narrows the gap to frontier models in reasoning coding and agent tasks with open weights for community use. Introducing"
X Link 2026-02-11T21:26Z 18.3K followers, [----] engagements
"This "Prompt is too long" bug in the Claude Code desktop is annoying. Apparently it's a known issue but I've lost a lot of sessions to it"
X Link 2026-02-12T03:26Z 18.3K followers, [----] engagements
"Something joyful and "profoundly" useful is coming from OpenAI soon. Some things we make are purely meant to bring joy others will profoundly affect how you get things done. We are getting ready to launch something that combines both and the team is so excited to share it soon. ✨ Some things we make are purely meant to bring joy others will profoundly affect how you get things done. We are getting ready to launch something that combines both and the team is so excited to share it soon. ✨"
X Link 2026-02-12T07:06Z 18.3K followers, [----] engagements
"@udaysy I've had other issues too very buggy app"
X Link 2026-02-12T07:17Z 18.3K followers, [--] engagements
"@Shawnryan96 The word AGI is now just marketing gobberish. All you need is an AI that does what you want it to do. That's it. Why are we waiting for AGI when current models can do a lot of things already. 😀"
X Link 2026-02-12T07:20Z 18.3K followers, [--] engagements
"@CtrlAltDwayne Ha that would be cool. But isn't that the Xbox 😜"
X Link 2026-02-12T07:55Z 18.3K followers, [--] engagements
"DeepMind pivots to 'root node' science as Gemini [--] and Nano Banana Pro redefine reasoning - CEO Demis Hassabis confirms a 50/50 split between scaling models and driving scientific breakthroughs targeting "root node" problems like nuclear fusion and material science to unlock downstream benefits. - Gemini [--] and "Nano Banana Pro" prioritise semantic understanding over pixel generation aiming to solve the "Jagged Intelligence" paradox where AI excels at hard maths but fails basic logic. - Hassabis envisions a post-AGI economy requiring Universal Basic Income (UBI) and predicts the AI revolution"
X Link 2026-02-12T12:32Z 18.3K followers, [----] engagements
"GPT Codex's coding meets Cerebras' speed in Spark out now for ChatGPT Pro users. Introducing GPT-5.3-Codex-Spark our ultra-fast model purpose built for real-time coding. Were rolling it out as a research preview for ChatGPT Pro users in the Codex app Codex CLI and IDE extension. https://t.co/6knTmyQZ4N Introducing GPT-5.3-Codex-Spark our ultra-fast model purpose built for real-time coding. Were rolling it out as a research preview for ChatGPT Pro users in the Codex app Codex CLI and IDE extension. https://t.co/6knTmyQZ4N"
X Link 2026-02-13T00:31Z 18.3K followers, [----] engagements
"@Angaisb_ Where do you get these"
X Link 2026-02-13T00:33Z 18.3K followers, [----] engagements
"Gemini [--] Deep Think released handles complex science and engineering tasks. - Targets fields with limited data like reviewing technical papers for flaws or optimising crystal growth recipes. - Transforms hand-drawn sketches into 3D-printable STL files by analysing geometry. - Demonstrated in university labs for physics maths and material science. - Available only to Google AI Ultra subscribers via the app's tools menu. Positions AI to speed up prototyping and bridge theory gaps in research. Today were releasing a significant upgrade to our specialized reasoning mode Gemini [--] Deep Think. Deep"
X Link 2026-02-13T01:58Z 18.3K followers, [----] engagements
"DeepThink [--] gets 84.6% on ARC-AGI-2 Miles in the lead. Gemini [--] Deep Think (2/26) Semi Private Eval - ARC-AGI-1: 96.0% $7.17/task - ARC-AGI-2: 84.6% $13.62/task New ARC-AGI SOTA model from @GoogleDeepMind https://t.co/mN8PFAWk4A Gemini [--] Deep Think (2/26) Semi Private Eval - ARC-AGI-1: 96.0% $7.17/task - ARC-AGI-2: 84.6% $13.62/task New ARC-AGI SOTA model from @GoogleDeepMind https://t.co/mN8PFAWk4A"
X Link 2026-02-13T01:58Z 18.3K followers, [---] engagements
"@daniel_mac8 👀"
X Link 2026-02-13T08:10Z 18.3K followers, [---] engagements
"AIs $5.2tn Ouroboros Patrick Boyle warns circular financing and energy limits could burst the bubble. - Big Tech giants are investing in startups like CoreWeave which then use the funds to buy chips back from the investors (e.g. Nvidia) creating artificial revenue growth akin to a snake eating its own tail. - A "Spaghetti Diagram" of cross-investments between OpenAI Microsoft Google Amazon and Nvidia artificially props up valuations with everyone incentivised to maintain the hype. - Capital expenditure is projected to hit $5.2 trillion by [----] yet current revenues are negligible turning"
X Link 2026-02-13T09:29Z 18.3K followers, [---] engagements
"Tongyi Lab launches Qwen-Image-2.0 as a next-generation image generation model. - Supports 1k-token prompts for creating professional infographics like PPTs and posters. - Offers native 2K resolution for detailed realistic scenes including people and architecture. - Unifies image generation and editing in one model with improved text rendering. - Features lighter architecture for faster inference speeds. Blind tests on AI Arena show top performance in text-to-image and image-to-image tasks. 1/4 🚀We are launching Qwen-Image-2.0 a next-generation foundational image generation model. The key"
X Link 2026-02-13T12:26Z 18.3K followers, [----] engagements
"Warp introduces Oz a cloud platform to orchestrate AI agents for automating development tasks. - Agents run in isolated Docker environments to build test and create PRs across multiple repositories. - Programmable via oz CLI for running agents scheduling tasks and managing sandboxes. - Management UI allows reviewing sessions artifacts and PRs from web phone or terminal. - Supports prebuilt Skills for issue triage audits tests or custom ones; works with models like Claude and Gemini. Oz enables scaling agent workflows with [----] bonus credits for February upgrades to Build or Max plans."
X Link 2026-02-13T15:26Z 18.3K followers, [---] engagements
"RT @koltregaskes: Sora is finally coming to ChatGPT"
X Link 2026-02-14T11:04Z 18.3K followers, [--] engagements
"@huanghlzone Claude helped me along the way. 😀"
X Link 2026-02-14T11:05Z 18.3K followers, [--] engagements
"@BrandGrowthOS Agree. After trying to train Claude to use a web page for something it's impossible so we need this definitely"
X Link 2026-02-14T11:07Z 18.3K followers, [--] engagements
"@X_DimensionNews With deep think excellent"
X Link 2026-02-14T11:08Z 18.3K followers, [--] engagements
"@axttimm I'm sure I haven't but clearly [---] was agent and coding focused"
X Link 2026-02-14T11:09Z 18.3K followers, [--] engagements
"@agi_exec I would presume so yes"
X Link 2026-02-15T15:45Z 18.3K followers, [--] engagements
"Impressive. Genuinely very impressed by the SVG of a pelican riding a bicycle I just got out of Google's new Gemini [--] Deep Think model https://t.co/xmaz3hlfkJ Genuinely very impressed by the SVG of a pelican riding a bicycle I just got out of Google's new Gemini [--] Deep Think model https://t.co/xmaz3hlfkJ"
X Link 2026-02-13T01:58Z 18.3K followers, [---] engagements
"AI 2027's [----] predictions aligned qualitatively with early [----] realities but at roughly 65% quantitative pace. - Progress on benchmarks like SWEBench reached 74.5% against predicted 85%. - AI agents emerged as unreliable personal assistants bungling tasks hilariously yet finding workflow fits. - Coding agents acted autonomously saving hours or days in R&D. - Revenues hit $20B for OpenAI slightly ahead while valuations lagged at $500B. The scenario proved directionally accurate prompting timeline adjustments to around [----]. To me the AI [----] forecast was wildly ambitious so when the latest"
X Link 2026-02-13T07:49Z 18.3K followers, [----] engagements
"Source: https://blog.ai-futures.org/p/grading-ai-2027s-2025-predictions https://blog.ai-futures.org/p/grading-ai-2027s-2025-predictions"
X Link 2026-02-13T07:49Z 18.3K followers, [---] engagements
"Source: https://www.youtube.com/watchv=NbL7yZCF-6Q https://www.youtube.com/watchv=NbL7yZCF-6Q"
X Link 2026-02-13T09:29Z 18.3K followers, [---] engagements
"RT @koltregaskes: According to Andrew Ng AI fears are overblown as layoffs stem from pandemic overhiring not widespread automation. - An"
X Link 2026-02-14T13:41Z 18.3K followers, [--] engagements
"RT @koltregaskes: Yes please be careful when using these quick installers for OpenClaw best to use the official onboarding only. It's pr"
X Link 2026-02-14T22:01Z 18.3K followers, [--] engagements
"GPT-5 says 'I don't know'. Love this thank you"
X Link 2025-08-18T16:05Z 18.3K followers, 1.6M engagements
"Mission Control dashboard built with React and Convex coordinates [--] OpenClaw AI agents. - Agents led by Jarvis delegate tasks autonomously. - Handles content creation revenue analysis and more 24/7. - Features threaded discussions for agent collaboration. - Mimics team dynamics with claims reviews and refutes. Dashboard enables startup automation without constant oversight. I got my agents to build this for me when I saw the screenshot only. We had a multi agent UI already but this command centre was better. 😀 https://t.co/V6XZfv5UYS https://t.co/V6XZfv5UYS"
X Link 2026-02-01T18:32Z 18.3K followers, 60K engagements
"MiniMax-M2.5 is out We are also due GLM [--] today Minimax M2.5 is now available. https://t.co/2bqRqoPo4k Minimax M2.5 is now available. https://t.co/2bqRqoPo4k"
X Link 2026-02-11T14:30Z 18.3K followers, [----] engagements
"Sora is finally coming to ChatGPT Seems like ChatGPT is finally getting an option to generate videos using Sora based on the most recent changes in the ChatGPT web app and the new Android app beta version (1.2026.041) which now mention video generation video processing and Sora generation IDs There are also https://t.co/qS0PDvMM46 Seems like ChatGPT is finally getting an option to generate videos using Sora based on the most recent changes in the ChatGPT web app and the new Android app beta version (1.2026.041) which now mention video generation video processing and Sora generation IDs There"
X Link 2026-02-13T21:30Z 18.3K followers, [----] engagements
"According to Andrew Ng AI fears are overblown as layoffs stem from pandemic overhiring not widespread automation. - Andrew notes tech layoffs often for cost-cutting with AI enabling efficiency but not replacing jobs yet. - Professions like translators face pressure but AI-skilled workers gain advantage in development and non-tech roles. - Smaller teams possible due to AI tools shifting bottlenecks to product management. - Opportunities for AI learners amid backlogs with courses on Claude Code and agentic skills recommended. - Stressful changes highlight need to adapt as most are starting from"
X Link 2026-02-14T04:56Z 18.3K followers, [----] engagements
"These all sound familiar from my experiences in the last few weeks as well. Yeah I can't decide whether I like slack or telegram. Telegram is definitely problematic slack I just don't like the UI. I'm trying to build my own one or hold on. I'm trying to get openclaw to build my own one. 😉 https://twitter.com/i/web/status/2022588557022392473 https://twitter.com/i/web/status/2022588557022392473"
X Link 2026-02-14T08:27Z 18.3K followers, [---] engagements
"OpenAI's AI reveals new particle glue effect. - Gluons are like tiny glue particles that hold the core of atoms together via a force called the strong nuclear force. - OpenAI's GPT-5.2 model studied complex math on how gluons interact. - It found that in specific setups an interaction scientists long thought had zero chance (impossible at basic levels) actually can happen with a measurable strength. - The AI created a simple formula for this quickly proved by another model and checked by expert physicists. This is a scientific breakthrough showing AI can uncover hidden rules in nature"
X Link 2026-02-14T12:34Z 18.3K followers, [----] engagements
"Good point where is Grok [----] its mid feb where is grok 4.20🗣 its mid feb where is grok 4.20🗣"
X Link 2026-02-14T22:57Z 18.3K followers, [----] engagements
"@MatthewBerman Is this on YouTube Matthew"
X Link 2026-02-14T23:50Z 18.3K followers, [----] engagements
"Claude dot AI (on the web) now has voice mode. Not sure when this was rolled out but it wasn't there a few days ago. Don't see it on the desktop app as yet at least on Windows. CONFIRMED ✅ @claudeai Web/Desktop Voice Mode arrived 🛸 @testingcatalog @btibor91 @koltregaskes 👀 https://t.co/41V2RrTT2p CONFIRMED ✅ @claudeai Web/Desktop Voice Mode arrived 🛸 @testingcatalog @btibor91 @koltregaskes 👀 https://t.co/41V2RrTT2p"
X Link 2026-02-15T15:33Z 18.3K followers, [----] engagements
"@third_leap I'm Max and not seen it until today. Odd. :-)"
X Link 2026-02-15T16:07Z 18.3K followers, [--] engagements
"Noam Brown argues Google's Deep Think release adds no new risks beyond Gemini [--] Pro as it is essentially scaffolding that users could replicate with enough inference compute. - He says criticisms from The Midas Project miss the shift to inference scaling over model size. - AI evaluation evolved from [----] pretraining FLOPs to test-time compute boosting benchmarks like ARC-AGI by 2-10x. - Suggests budgeting evals at $1000 inference with scaffolds for realistic capability checks. - Calls for system cards plotting performance vs inference compute projecting safety at $1M+ levels. This could"
X Link 2026-02-15T21:32Z 18.3K followers, [----] engagements
"Sam Altman sacked by OpenAI AGAIN jk"
X Link 2025-10-05T16:00Z 15.4K followers, [----] engagements
"@ChetasOG Which plan"
X Link 2025-10-05T18:49Z 15.4K followers, [----] engagements
"Is "cheetah" Google's Gemini Diffusion language model I had access to Diffusion in May this is my speed test. It was fast VERY fast. It won't be like a thing you've seen before. Can anyone compare this to cheetah Diffusion text model writes [-----] tokens in [--] seconds This is Gemini Diffusion running in real-time 🤯 Extreme stress test possibly https://t.co/PPiOX1Q3cT Diffusion text model writes [-----] tokens in [--] seconds This is Gemini Diffusion running in real-time 🤯 Extreme stress test possibly https://t.co/PPiOX1Q3cT"
X Link 2025-10-06T06:26Z 15.4K followers, 34.4K engagements
"@testingcatalog Is this web"
X Link 2025-10-06T06:29Z 15.4K followers, [---] engagements
"@BennettBuhner Ah yeah looks like OpenAI platform. That's way less exciting"
X Link 2025-10-06T06:35Z 15.4K followers, [--] engagements
"OpenAI DevDay [----] is today What are you looking forward to Time to event: [--] hours from now other times: - [--] AM PT (Mon) - [--] PM ET (Mon) - [--] PM BST (Mon) - [--] PM CEST (Mon) - 10:30 PM IST (Mon) - [--] AM CST (Tue) - [--] AM JST (Tue) - [--] AM AEDT (Tue) - [--] AM NZDT (Tue) What to expect: Not much news or rumours so mostly guesswork: - Sora 2/Pro API - Agent v2 + 'Aura' Browser - Image v2 - IMO gold medal model - New agents - Any Codex updates - Opening keynote livestream; hands-on demos and builder sessions recorded (agenda TBA) - Developer-focused product announcements"
X Link 2025-10-06T08:00Z 15.4K followers, 22.5K engagements
"Link: https://openai.com/index/announcing-devday-2025/ https://openai.com/index/announcing-devday-2025/"
X Link 2025-10-06T08:00Z 15.4K followers, [----] engagements
"The posts I saw a little while back about Nvidia investing $100bn into OpenAI so OpenAI can buy $100bn of GPUs from Nvidia are amusing. That's how the economy works. 🤣"
X Link 2025-10-06T10:00Z 15.4K followers, [---] engagements
"OpenAI signs definitive agreement with AMD to deploy [--] gigawatts of GPUs over multiple years. - First gigawatt deployment begins in second half of [----] using AMD Instinct MI450 chips. - OpenAI CEO Sam Altman states partnership advances compute capacity for AI potential; AMD CEO Lisa Su highlights delivery of AI compute at massive scale. Agreement aligns strategically to boost position in AI chip market against Nvidia dominance. OpenAI agrees to a deal with AMD to deploy a massive amount of AI infrastructure https://t.co/fVd8nRAvul OpenAI agrees to a deal with AMD to deploy a massive amount of"
X Link 2025-10-06T11:10Z 15.4K followers, [----] engagements
"Big week ahead OpenAI Dev Day is today and Google has events including the Gemini at Work livestream on the 9th and Google Cloud events today. What are you excited about this week New ships. @sama keynote streaming live. DevDay [----] starts tomorrow. New ships. @sama keynote streaming live. DevDay [----] starts tomorrow"
X Link 2025-10-06T13:02Z 15.4K followers, 12.2K engagements
"OpenAI teases licensed fictional character cameos for Sora according to the Verge shifting to opt-in copyright with rightsholder controls and revenue sharing to curb unauthorised content. Fictional characters are (officially) coming to Sora as OpenAI manages copyright chaos https://t.co/YYvWMPtf4L Fictional characters are (officially) coming to Sora as OpenAI manages copyright chaos https://t.co/YYvWMPtf4L"
X Link 2025-10-06T14:21Z 15.4K followers, 10.1K engagements
"The @OpenAI DevDay [----] livestream is up. Link below. [--] hr [--] mins to go. "Sam Altman kicks off DevDay [----] with a keynote to explore ideas that will challenge how you think about building. Join us for announcements live demos and a vision of how developers are reshaping the future with AI.""
X Link 2025-10-06T15:50Z 15.4K followers, [----] engagements
"OpenAI DevDay livestream: https://www.youtube.com/watchv=hS1YqcewH0c https://www.youtube.com/watchv=hS1YqcewH0c"
X Link 2025-10-06T15:50Z 15.4K followers, [---] engagements
"@X_DimensionNews @OpenAI I think only Meta do time-friendly events for you"
X Link 2025-10-06T16:06Z 15.3K followers, [--] engagements
"@sonicshifts @OpenAI Keep expectations low then you'll never be disappointed"
X Link 2025-10-06T16:21Z 15.4K followers, [--] engagements
"Sora [--] API coming today at OpenAI DevDay https://t.co/MxMxMvktCG Sora 2's icon for the API docs is now live. Looks like the Sora [--] API will indeed be coming at DevDay likely alongside Sora [--] Pro. https://t.co/viHsOTdVm3 https://t.co/MxMxMvktCG Sora 2's icon for the API docs is now live. Looks like the Sora [--] API will indeed be coming at DevDay likely alongside Sora [--] Pro. https://t.co/viHsOTdVm3"
X Link 2025-10-06T16:22Z 15.4K followers, [----] engagements
"@Akumunokokoro Yeah I wasn't sure it was new news but shared anyway perhaps we're missing the context here. Perhaps tool calling on the web more tooling in the API or additional tool calling for agents. 🤔"
X Link 2025-10-06T16:56Z 15.4K followers, [---] engagements
"@OpenAI @sama It's OK I'm playing frogger"
X Link 2025-10-06T17:03Z 15.4K followers, [---] engagements
"ChatGPT now has 800M+ weekly users"
X Link 2025-10-06T17:13Z 15.5K followers, [----] engagements
"Apps are coming to inside ChatGPT for example Figma and Spotify"
X Link 2025-10-06T17:17Z 15.4K followers, [----] engagements
"OpenAI basically just announced a new app store. [---] million weekly users are already waiting for them. This will be huge"
X Link 2025-10-06T17:30Z 15.4K followers, [----] engagements
"OpenAI announces AgentKit build your own agents with Agent Builder ChatKit and Evals which uses an n8n-style UI in the OpenAI API platform"
X Link 2025-10-06T17:34Z 15.4K followers, [----] engagements
"@creationblok I presume it's on the OpenAI API platform"
X Link 2025-10-06T17:43Z 15.4K followers, [--] engagements
"@OpenAI Available in the UK"
X Link 2025-10-06T18:19Z 15.4K followers, [----] engagements
"@cfryant @grok v2 I think it's v0.9"
X Link 2025-10-06T18:58Z 15.4K followers, [---] engagements
"@TheRealAdamG Salesforce connector would be nice too. :-)"
X Link 2025-10-06T19:10Z 15.4K followers, [---] engagements
"@MwydZyad82729 @X_DimensionNews @OpenAI I suspect they'll integrate all abilities from 4o into [--] then retire old machines slowly"
X Link 2025-10-06T19:37Z 15.4K followers, [--] engagements
"@Yasamanini See the OpenAI news about Apps in ChatGPT"
X Link 2025-10-06T19:41Z 15.4K followers, [--] engagements
"@MwydZyad82729 @OpenAI There is no 5o no"
X Link 2025-10-06T19:57Z 15.4K followers, [--] engagements
"No these are not available in the UK. @OpenAI @OpenAIDevs please clarify. I see them in the USA though: Apps in ChatGPT is not available in the EU yet. No news on the UK but I cannot get them to work yet. Apps in ChatGPT is not available in the EU yet. No news on the UK but I cannot get them to work yet"
X Link 2025-10-06T20:07Z 15.4K followers, [----] engagements
"@Yasamanini See the OpenAI stat from today"
X Link 2025-10-06T21:42Z 15.4K followers, [--] engagements
"@AllMigh48938863 @OpenAI Yes but I've found it's not available in the UK either. OpenAI have not mentioned this"
X Link 2025-10-06T21:47Z 15.4K followers, [---] engagements
"@n1babc Hmmm that's an interesting view hmmm"
X Link 2025-10-06T23:12Z 15.4K followers, [--] engagements
"I've been trying to get the Spotify app in ChatGPT to work and I'm struggling with it. First don't expect anything that the app's API cannot do. I thought it could give me songs similar to ones on one of my playlists that I've not liked and nope. It gives me a mixture of songs from the playlist or songs from other playlists I've liked. It did give me a few unliked songs but ironically they were duplicates of songs I've liked (I did reprompt with this in mind but I get errors). It cited issues with what the API could do. I'm also getting errors when it creates a playlist about 50% of the time."
X Link 2025-10-07T02:30Z 15.4K followers, [----] engagements
"@adcock_brett I read that as running on BMX lol. That would be cool. 😜"
X Link 2025-10-07T06:00Z 15.4K followers, [---] engagements
"@GabrielPeterss4 Depends what your after I guess it's no Midjourney feed right now. But if you like memes sure"
X Link 2025-10-07T06:01Z 15.4K followers, [--] engagements
"@maxescu But you are getting way more than just Sora 2"
X Link 2025-10-07T06:17Z 15.4K followers, [---] engagements
"@gdb Any news on a Salesforce integration"
X Link 2025-10-07T06:32Z 15.4K followers, [---] engagements
"Gemini [--] still on for this week My gut says no at this point. This was hinted at in the now-deleted post below and there are murmurs that it may have reverted to its previous late October release window. Some model updates: - Gemini [--] (experimental)'s current target is 9th October. - Claude [---] is currently a 1-2 weeks away (not next week) - A Gemini [---] Pro update is coming to enterprise only - The new [---] Flash/Flash Lite are NOT "oceanstone" or "oceanreef" - "oceanstone" and Some model updates: - Gemini [--] (experimental)'s current target is 9th October. - Claude [---] is currently a 1-2 weeks"
X Link 2025-10-07T08:29Z 15.4K followers, 13.4K engagements
"Jony Ive's secretive AI hardware device io faces potential delays due to three unsolved problems. - The device designed by former Apple chief Jony Ive and to be launched by OpenAI is described as a novel form factor not yet seen ruling out glasses phones badges smartwatches rings or headphones. - OpenAI CEO Sam Altman and Ive released a video discussing the project without details on form or function with ChatGPT generating pebble-shaped images as a troll response. - Scheduled for launch next year but Financial Times reports suggest delays from people familiar with plans. - First problem:"
X Link 2025-10-07T11:29Z 15.4K followers, [----] engagements
"AI pioneers disagree on LLMs' path to AGI. Richard Sutton RL expert and Turing Award winner calls them a "dead end" for lacking adaptive learning beyond pattern-matching. Elon Musk believes Grok [--] could achieve AGI soon through massive compute and data scaling. Yann LeCun agrees with Sutton arguing scaling won't address deficits in reasoning planning or world modelling without new architectures. Optimists like Dario Amodei view scaled LLMs as "datacenters of geniuses" enabling Nobel-level innovations supported by [----] benchmarks like ICPC gold medals from models rom DeepMind and OpenAI."
X Link 2025-10-07T20:34Z 15.5K followers, [----] engagements
"@axttimm For ChatGPT technically. My gut says they are building ChatGPT to be a platform or even an OS. As far as what it is today I've only tried on the web but these might also work in the mobile apps"
X Link 2025-10-07T21:42Z 15.4K followers, [--] engagements
"OpenAI denies xAI's allegations of trade secret misappropriation through employee poaching. - OpenAI admits its corporate structure and hiring of former xAI employees including Xuechen Li Jimmy Fraiture and others but denies any encouragement or knowledge of downloading or disclosing xAI source code. - Denies claims of a deliberate scheme to access xAI confidential information stating recruiter communications were standard and no trade secrets were acquired or used by OpenAI. - Asserts xAI failed to implement industry-standard protections for confidential information particularly on employee"
X Link 2025-10-08T04:01Z 15.4K followers, [----] engagements
"@lulumeservey Nope only seen one from Brett and one from Figure"
X Link 2025-10-08T06:08Z 15.4K followers, [--] engagements
"Meta Ray-Ban Display glasses disappoint Robert with lack of phone app integrations like Gmail or Slack no web browsing limited entertainment to one game without YouTube or Netflix and HUD-only display incapable of AR or MR. Personal discomfort from armband after minutes of use and low-resolution screen with small field of view; targeted at influencers content creators and gadget enthusiasts leaving gap for competitors like OpenAI and Apple. I was hoping for AI glasses by this year seems not. My next prescription will be "normal" glasses. I'm so disappointed in the Meta Raybans with display."
X Link 2025-10-08T07:02Z 15.5K followers, [----] engagements
"Sam Altman confirms no concrete ad plans for OpenAI's compute-heavy ChatGPT Pulse at DevDay [----] but leaves door open for future relevant options like Instagram's. Sam Altman says there are no current plans for ads within ChatGPT Pulse but hes not ruling it out https://t.co/4pbUK67CY1 Sam Altman says there are no current plans for ads within ChatGPT Pulse but hes not ruling it out https://t.co/4pbUK67CY1"
X Link 2025-10-08T10:00Z 15.4K followers, 10.8K engagements
"Mosaic launches public beta an agentic AI video editor with features including canvas ideation timeline editing visual intelligence content engine generative integrations asset library and API for workflows. Enables real creative control via autonomous agents and iterative processes. https://x.com/_adishj/status/1973432845436854418/video/1 today we're launching Mosaic the agentic video editor. in a world tending towards AI slop create something real. no waitlist public beta is now live at mosaic dot so. comment "MOSAIC" to get [----] free credits dropped into your account. this release comes"
X Link 2025-10-08T11:30Z 15.4K followers, [----] engagements
"Gemini [---] Computer Use model launched in public preview via Gemini API built on [---] Pro to enable AI agents for UI interactions like clicking and typing outperforming rivals on benchmarks like Online-Mind2Web with up to 20% higher success and lower latency. Optimised for web browsers with mobile potential it loops using screenshots action history and function calls for tasks; safety integrates trained safeguards inference checks and developer controls for high-risk actions. Early testers report 50% faster performance and 18% better reliability in UI testing and automation. 💻 Introducing"
X Link 2025-10-08T14:30Z 15.4K followers, [----] engagements
"ChatGPT is the only top [--] website in the world increasing its traffic at the moment (stats from September) Top [--] websites by total visits in September 2025: ChatGPT was the only website with positive growth. Bing dropped to 10th place surpassed by Wikipedia. https://t.co/heEE1C5Swa Top [--] websites by total visits in September 2025: ChatGPT was the only website with positive growth. Bing dropped to 10th place surpassed by Wikipedia. https://t.co/heEE1C5Swa"
X Link 2025-10-08T16:01Z 15.4K followers, [----] engagements
"@Jay_sharings Instagram is fine just don't read the comments. ;-)"
X Link 2025-10-08T16:06Z 15.4K followers, [---] engagements
"@ProperPrompter https://sora.chatgpt.com/profile/koltregaskes https://sora.chatgpt.com/profile/koltregaskes"
X Link 2025-10-09T06:16Z 15.4K followers, [--] engagements
"@kimmonismus @Prashant_1722 They guess and sometimes get hits"
X Link 2025-10-09T13:08Z 15.4K followers, [---] engagements
"Veo [---] from Google must be launching today then. 🚨 Veo [---] is coming soon exclusively on Vadoo AI. This isnt just another update its evolution. 🎬 30-second video generation 🎧 Native audio support 💎 1080p ultra-real output ⚡ Faster. Smoother. More cinematic than ever. And heres the deal The first [--] retweets get https://t.co/hNT40Whgxe 🚨 Veo [---] is coming soon exclusively on Vadoo AI. This isnt just another update its evolution. 🎬 30-second video generation 🎧 Native audio support 💎 1080p ultra-real output ⚡ Faster. Smoother. More cinematic than ever. And heres the deal The first 20"
X Link 2025-10-09T13:23Z 15.4K followers, [----] engagements
"OpenAI DevDay AMA starts in [---] hours. DevDay AMA on Reddit tomorrow with some of the people behind the ships AgentKit Apps SDK Sora [--] in the API GPT-5 Pro in the API Codex and more. Tomorrow [--] AM PT. https://t.co/2T3M6giPAk DevDay AMA on Reddit tomorrow with some of the people behind the ships AgentKit Apps SDK Sora [--] in the API GPT-5 Pro in the API Codex and more. Tomorrow [--] AM PT. https://t.co/2T3M6giPAk"
X Link 2025-10-09T15:32Z 15.4K followers, [----] engagements
"AMA link: https://www.reddit.com/r/OpenAI/s/8HdbouqEeP https://www.reddit.com/r/OpenAI/s/8HdbouqEeP"
X Link 2025-10-09T15:32Z 15.4K followers, [---] engagements
"@test_tm7873 I'm sure they will all get the model but it's not out yet. It has been spotted though post coming"
X Link 2025-10-09T18:02Z 15.4K followers, [--] engagements
"Veo [---] and Veo [---] Fast spotted on Google Cloud Console. Ignore all the AI tools claiming they have exclusive access or making false claims about it being released in September. Veo [---] and Veo 3.1-fast have been found in the Google Cloud Console https://t.co/xzTytgTeL2 Veo [---] and Veo 3.1-fast have been found in the Google Cloud Console https://t.co/xzTytgTeL2"
X Link 2025-10-09T19:47Z 15.4K followers, 13.7K engagements
"If you give Grok Imagine an image it does pretty well. https://grok.com/imagine/post/0f8e969d-90fe-4704-afad-a9559ff365a7source=post-page&platform=web https://grok.com/imagine/post/0f8e969d-90fe-4704-afad-a9559ff365a7source=post-page&platform=web"
X Link 2025-10-09T22:48Z 15.4K followers, [---] engagements
"@forward_future_ @gdb @OpenAI Is the full video on X"
X Link 2025-10-10T06:19Z 15.4K followers, [--] engagements
"Another Grok Imagine with an image prompt. https://grok.com/imagine/post/99ebaf3a-60ca-455d-85ee-d0cb8c25fc93source=post-page&platform=web https://grok.com/imagine/post/99ebaf3a-60ca-455d-85ee-d0cb8c25fc93source=post-page&platform=web"
X Link 2025-10-10T10:01Z 15.4K followers, [---] engagements
"Sorry AI community I'm busy for the rest of the year This is all-out warfare. This is #Battlefield6. Deployment begins October [--] 🫡 https://t.co/KutGpJfn1j This is all-out warfare. This is #Battlefield6. Deployment begins October [--] 🫡 https://t.co/KutGpJfn1j"
X Link 2025-10-10T15:00Z 15.4K followers, [----] engagements
"Grok Imagine + puppy image prompt. https://grok.com/imagine/post/e0b333e1-d8b8-4a14-b740-c0c0b60eaaf1source=post-page&platform=web https://grok.com/imagine/post/e0b333e1-d8b8-4a14-b740-c0c0b60eaaf1source=post-page&platform=web"
X Link 2025-10-10T22:00Z 15.4K followers, [---] engagements
"Epoch AI reckons GPT-5's training needed more computer power than GPT-4 but less than GPT-4.5 done by improving a base model with around [---] billion elements and more than [--] trillion bits of info after the initial run. It highlights OpenAI keeping up with competitors like Anthropic topping coding rankings. We recently wrote that GPT-5 is likely the first mainline GPT release to be trained on less compute than its predecessor. How did we reach this conclusion and what do we actually know about how GPT-5 was trained 🧵 https://t.co/zD0K2UclRD We recently wrote that GPT-5 is likely the first"
X Link 2025-10-11T19:02Z 15.4K followers, [----] engagements
"@Makuh90 @grok or @AskPerplexity any help"
X Link 2025-10-11T19:44Z 15.4K followers, [--] engagements
"A Sora [--] prompting guide has been published by OpenAI. Link in Tibor's thread. OpenAI published a new Sora [--] Prompting Guide so I created a prompt template inspired by their recommendations that writes prompts for you Pick your model (Sora [--] for 10s or Sora [--] Pro for 15s) choose from almost [---] styles add a Cameo if you want describe your idea and it https://t.co/S6wp0cQ9np OpenAI published a new Sora [--] Prompting Guide so I created a prompt template inspired by their recommendations that writes prompts for you Pick your model (Sora [--] for 10s or Sora [--] Pro for 15s) choose from almost 100"
X Link 2025-10-12T16:01Z 15.5K followers, [----] engagements
"METR's time horizon metric for AI agents shows doubling every [--] months over [--] years with Claude [---] Sonnet at one hour and projections for month-long tasks in under a decade. Measured on RE-Bench HCAST and novel tasks timed by domain experts; robust to subsets but with limitations in external validity and potential acceleration in recent data. We estimate that Claude Sonnet [---] has a 50%-time-horizon of around [--] hr [--] min (95% confidence interval of [--] to [---] minutes) on our agentic multi-step software engineering tasks. This estimate is lower than the current highest time-horizon point"
X Link 2025-10-12T20:32Z 15.4K followers, [----] engagements
"Radical Numerics releases RND1-Base a 30B-parameter sparse MoE diffusion language model with 3B active parameters. Converted from Qwen3-30BA3B outperforming Dream-7B and LLaDA-8B on benchmarks like MMLU at 57.2% GSM8K at 72.1% and MBPP at 51.3%. Conversion uses Simple Continual Pretraining with bidirectional masks layer-specific learning rates to retain knowledge and large batch sizes up to 8M tokens for stability; pretrained on 500B tokens. Introducing RND1 the most powerful base diffusion language model (DLM) to date. RND1 (Radical Numerics Diffusion) is an experimental DLM with 30B params"
X Link 2025-10-12T23:30Z 15.4K followers, [----] engagements
"Elon says Grok Imagine will be real time in "2 months". @_rockt @GoogleDeepMind We will have real-time in [--] months @_rockt @GoogleDeepMind We will have real-time in [--] months"
X Link 2025-10-13T11:30Z 15.4K followers, [----] engagements
"Open is building its own AI chips. They and Broadcom announce multi-year collaboration to deploy [--] gigawatts of custom AI accelerators starting H2 [----] targeting completion by [----] for optimised AI infrastructure. The partnership builds on prior co-development with OpenAI designing systems embedded with model insights and Broadcom handling deployment via Ethernet-based networking in data centers worldwide. We're partnering with Broadcom to deploy 10GW of chips designed by OpenAI. Building our own hardware in addition to our other partnerships will help all of us meet the worlds growing"
X Link 2025-10-13T13:44Z 15.4K followers, [---] engagements
"@peterdedene @grok does OpenAI have any partnerships with power firms yet"
X Link 2025-10-13T14:55Z 15.4K followers, [---] engagements
"@mark_k Indeed great to see (well hear). I'd like to put mine on here just so I can listen to it along with human music"
X Link 2025-10-13T17:54Z 15.4K followers, [---] engagements
"Veo [---] release is imminent. Google Veo [---] is Coming https://t.co/kmWFeosOgb Google Veo [---] is Coming https://t.co/kmWFeosOgb"
X Link 2025-10-14T09:57Z 15.5K followers, [----] engagements
"ChatGPT personalities and adult mode are coming soon in November and December respectively. We made ChatGPT pretty restrictive to make sure we were being careful with mental health issues. We realize this made it less useful/enjoyable to many users who had no mental health problems but given the seriousness of the issue we wanted to get this right. Now that we have We made ChatGPT pretty restrictive to make sure we were being careful with mental health issues. We realize this made it less useful/enjoyable to many users who had no mental health problems but given the seriousness of the issue"
X Link 2025-10-14T16:27Z 15.5K followers, [----] engagements
"@elonmusk Blimey this sounds great"
X Link 2025-10-14T16:36Z 15.4K followers, [---] engagements
"@CEOAlexColon Yeah I still think there should be a base or some rules the personalities cannot break but perhaps their mental health checks are really that good. 🤔"
X Link 2025-10-14T16:48Z 15.5K followers, [---] engagements
"@DrealR_ If the UI is being updated now it is not months away"
X Link 2025-10-14T17:06Z 15.5K followers, [--] engagements
"@Benioff @OpenAI @salesforce 👀"
X Link 2025-10-14T18:17Z 15.5K followers, [----] engagements
"@sama @alth0u There are definitely people on here who want 4o for the personality"
X Link 2025-10-14T18:23Z 15.5K followers, [---] engagements
"Nano Banana has been integrated into NotebookLM to power Video Overview formats. NotebookLM updates: - Infographic artefact will have a format configuration - Nano Banana will power New Video Overview formats - We will get the Fast Research option for sourcing - Magic View is no longer a mystery 🔥 https://t.co/bfEd0MZ20r NotebookLM updates: - Infographic artefact will have a format configuration - Nano Banana will power New Video Overview formats - We will get the Fast Research option for sourcing - Magic View is no longer a mystery 🔥 https://t.co/bfEd0MZ20r"
X Link 2025-10-14T20:35Z 15.5K followers, [----] engagements
"Gemini account confirms Veo [---] Out already maybe US only"
X Link 2025-10-15T00:17Z 15.5K followers, [----] engagements
"Veo [---] release must be today the models have already been spotted. veo-3. 1-generate-preview veo-3. 1-fast-generate-preview Veo Veo"
X Link 2025-10-15T05:53Z 15.5K followers, 16.1K engagements
"Veo [---] is out on Google Flow now"
X Link 2025-10-15T12:52Z 15.5K followers, [----] engagements
"@UsmanTjayy @CalimanuLoredan New DMs"
X Link 2025-10-15T14:31Z 15.5K followers, [--] engagements
"Is the Sora App and Sora [--] coming to more countries FAQ page updated but no visible changes as yet. I think they are ready to expand the Sora [--] app to other countries. The FAQ page was updated yesterday. (But hidden for now). Very interesting. @ChatGPTapp @OpenAI https://t.co/VhFFOCBuRJ I think they are ready to expand the Sora [--] app to other countries. The FAQ page was updated yesterday. (But hidden for now). Very interesting. @ChatGPTapp @OpenAI https://t.co/VhFFOCBuRJ"
X Link 2025-10-15T14:49Z 15.5K followers, [----] engagements
"@InfiniteHexx @kimmonismus @testingcatalog Just read your post saw the quote elsewhere but thank you. 👍"
X Link 2025-10-15T14:52Z 15.5K followers, [--] engagements
"@hackscorpio Greedy or just needing to pay their billions investments back"
X Link 2025-10-15T16:12Z 15.5K followers, [--] engagements
"@Jay_sharings @Makuh90 @karinanguyen_ @maisonagi @ilyasut @OpenAI Man that sucks. Amazing commitment though Jay"
X Link 2025-10-15T21:58Z 15.5K followers, [--] engagements
"OpenAI announces the formation of the Expert Council on Well-Being and AI. - The council comprises eight leading researchers and experts in how technology affects emotions motivation and mental health. - It will advise OpenAI on building healthier interactions with tools like ChatGPT and Sora for all ages. - Focus includes supporting healthy development especially for youth as teens use AI differently from adults. - Collaboration involves regular meetings check-ins and discussions on topics such as AI behaviour in sensitive situations parental controls and user empowerment. - OpenAI makes"
X Link 2025-10-16T17:31Z 15.5K followers, [----] engagements
"@AravSrinivas Excellent. So like @duolingo for genAI"
X Link 2025-10-16T18:54Z 15.6K followers, [----] engagements
"The X algo + weights will be published this week. All X posts will be evaluated by Grok full AI recommendations soon. And shortly after we will get the power to control what we see in our X feed. Awesome. To the degree that people are seeing improvements in their feed it is not due to the actions of specific individuals changing heuristics but rather increasing use of Grok and other AI tools. We will post the updated algorithm including model weights later this week. You To the degree that people are seeing improvements in their feed it is not due to the actions of specific individuals"
X Link 2025-10-16T21:00Z 15.5K followers, [----] engagements
"@antonp_me @HarpaAI I gave up using it as it got me throttle on ChatGPT regularly. It's too aggressive. Unless they work with the AI labs or will get banned. Such a shame"
X Link 2025-10-16T21:10Z 15.5K followers, [--] engagements
"@antonp_me @HarpaAI I had loads I built custom workflows for running multi-step research and other actions and I get throttled more and more regularly using them"
X Link 2025-10-16T21:56Z 15.5K followers, [--] engagements
"@HarpaAI @antonp_me My custom workflows resulted in ChatGPT being throttled a lot. And for up to [--] days. It got worse and worse to the point I ran just one custom workflow and got throttled. I had to stop to get work done"
X Link 2025-10-17T14:36Z 15.5K followers, [--] engagements
"@HarpaAI @antonp_me I'm using agents instead now such as Comet and ChatGPT Agent"
X Link 2025-10-17T17:05Z 15.5K followers, [--] engagements
"Doom runs inside ChatGPT. Can ChatGPT run Doom Yes. ChatGPT Apps are very powerful. I cloned our Next.js ChatGPT template registered a _ MCP tool and deployed to @vercel. Once the tool is called ChatGPT embeds the full @nextjs application. Server and client rendering just works and https://t.co/pWjbH6woJn Can ChatGPT run Doom Yes. ChatGPT Apps are very powerful. I cloned our Next.js ChatGPT template registered a _ MCP tool and deployed to @vercel. Once the tool is called ChatGPT embeds the full @nextjs application. Server and client rendering just works and https://t.co/pWjbH6woJn"
X Link 2025-10-17T17:35Z 15.6K followers, [----] engagements
"Populous reborn in your browser http://t.co/uvuG16zo #gaming #videogame #entertainment http://kolt.at/LmwIAW http://kolt.at/LmwIAW"
X Link 2012-05-21T21:10Z [----] followers, [--] engagements
"Does anyone know when Android [---] will be available for the Samsung Galaxy Nexus"
X Link 2012-06-29T17:44Z [----] followers, [--] engagements
"500px / Photo "Mount Trivor 7728m." by Atif Saeed http://t.co/fjNqBJmY http://kolt.at/SuvflB http://kolt.at/SuvflB"
X Link 2012-10-04T18:47Z 15.1K followers, [--] engagements
"The Sony World Photography Awards Current Photo Contests http://ff.im/8qmwA"
X Link 2009-09-20T15:15Z [----] followers, [--] engagements
"RT @ "THE CLUTCH - Rainbow Six Siege " http://ow.ly/2lD5303ZXB2 http://ow.ly/2lD5303ZXB2"
X Link 2016-09-08T22:50Z [----] followers, [--] engagements
Limited data mode. Full metrics available with subscription: lunarcrush.com/pricing