@koltregaskes Kol Tregaskes

Kol Tregaskes posts on X about open ai, 6969, ai, chatgpt the most. They currently have [------] followers and [----] posts still getting attention that total [---------] engagements in the last [--] hours.

Engagements: [---------] #

Mentions: [--] #

Followers: [------] #

CreatorRank: [-------] #

Social Influence

Social category influence technology brands stocks celebrities finance social networks countries travel destinations cryptocurrencies gaming events

Social topic influence open ai, 6969, ai, chatgpt, grok, agi, $googl, google, sama, microsoft

Top accounts mentioned or mentioned by @openai @kimmonismus @iruletheworldmo @midjourney @sama @testingcatalog @btibor91 @grok @chatgptapp @xai @elonmusk @angaisb @aiforsuccess @kyrannio @applesjimmy @makuh90 @geminiapp @markk @anthropicai @perplexityai

Top assets mentioned GrokCoin (GROKCOIN) Alphabet Inc Class A (GOOGL) Microsoft Corp. (MSFT) Sora (SORA) Frontier (FRONT) Flux (FLUX) Robot Consulting Co., Ltd. (LAWR) Spotify Technology (SPOT) Merge (MERGE)

Top Social Posts

Top posts by engagements in the last [--] hours

"Ethereum Fork ETHPoW Suffers Bridge Replay Exploit Token Tanks 37% https://decrypt.co/110023/ethereum-fork-ethpow-suffers-bridge-replay-exploit-token-tanks-37 https://decrypt.co/110023/ethereum-fork-ethpow-suffers-bridge-replay-exploit-token-tanks-37"
X Link 2022-09-21T18:47Z 17.9K followers, [--] engagements

"Upscaler released from the Stable Diffusion maker. Stability AI's Image Upscaling API is out now This game-changing tool can be used to upscale any image without losing any sharpness. Get Upscaling by using our Plugins for Blender and Photoshop or via DreamStudio. https://t.co/NHbFq2ryhA #stabilityai Stability AI's Image Upscaling API is out now This game-changing tool can be used to upscale any image without losing any sharpness. Get Upscaling by using our Plugins for Blender and Photoshop or via DreamStudio. https://t.co/NHbFq2ryhA #stabilityai"
X Link 2023-04-26T19:57Z 17.8K followers, [--] engagements

"👉Midjourney Mock-ups 👈 I'm no designer coder or UX expert (no expert at anything TBH) but I thought I'd try my hand at mocking up what I'd like to see changed on the Midjourney alpha website. I also wanted an excuse to use @tldraw which is awesome Thank you. With these caveats in mind here's my first (crude) attempt at a MJ website mock-up. The main features: - Move several buttons from the current right panel too many there. Move them to an "Editor" window and swap the rest to sliders - An Editor window is an expansion of the current Change AR window with added zoom and drag & drop (to"
X Link 2024-02-11T10:59Z 16.7K followers, [----] engagements

"Like with all prerecorded video demos of robots we need to see live demos demos in different settings and with different cups plates and equipment. If it's adaptable then this is mighty impressive. The new Astribot S1 is slowly becoming popular. Nevertheless you shouldn't be blinded by the video. The activities that are carried out are trained to be so simple that no real-life application is conceivable. We see how the robot bakes pancakes but without making the dish https://t.co/lbpK605q8g The new Astribot S1 is slowly becoming popular. Nevertheless you shouldn't be blinded by the video. The"
X Link 2024-08-20T10:56Z 15.8K followers, [---] engagements

"SambaNova announces the "worlds fastest AI platform". Runs Llama [---] 405B at 132t/s 📣 Announcing the worlds fastest AI platform SambaNova Cloud runs Llama [---] 405B @ 132t/s at full precision ✅ Llama [---] 405B @ [---] tokens/sec ✅ Llama [---] 70B @ [---] tokens/sec ✅ 10X Faster Inference than GPUs Start developing #FastAI ➡ https://t.co/zm6RCXY00n @AIatMeta https://t.co/dBoWdBZ6ud 📣 Announcing the worlds fastest AI platform SambaNova Cloud runs Llama [---] 405B @ 132t/s at full precision ✅ Llama [---] 405B @ [---] tokens/sec ✅ Llama [---] 70B @ [---] tokens/sec ✅ 10X Faster Inference than GPUs Start"
X Link 2024-09-10T17:42Z 17.9K followers, [---] engagements

"OpenAI's 'Orion' model already being previewed Disclaimer: We're talking a massive bag of salt on this one: An anon posts a random text file on Pastebin about an apparent 'o1-2024-09-12-preview' model. With that out of the way. Going by the Pastebin dump it talks about a very early build ("preview") being shown to industries. It describes a 'Reasoning-First Approach' for ingesting intricate data to automate workflows and data validation. This could be the Orion model o = Orion. Lots of salt remember. ;-) Anon shared "o1-2024-09-12-preview." Can anybody validate this content"
X Link 2024-09-12T08:58Z 16.2K followers, 16.7K engagements

"What are we thinking of the weekly Release Thursday tomorrow"
X Link 2024-10-23T21:25Z 16K followers, [---] engagements

"@hellokillian I get this error ImportError: cannot import name 'StrEnum' from 'enum' (C:Program http://enum.py http://enum.py"
X Link 2024-10-24T14:58Z 15.8K followers, [---] engagements

"@peach_nemesis @ai_katana @SirMrMeowmeow The rumoured Sora v2. ;-)"
X Link 2024-11-21T15:44Z 16K followers, [--] engagements

"Reid Hoffman Warns of Elon Musk's AI Policy Conflict @reidhoffman co-founder of LinkedIn and Inflection AI has raised concerns about Elon Musks influence on AI policy warning of potential conflicts of interest. Hoffman suggests Musks leadership in AI ventures could distort policy decisions favouring personal interests over wider ethical considerations. - Hoffman criticises Musks role in both AI development and policy-making highlighting potential conflicts of interest. - He calls for transparency and accountability in setting AI policies to avoid corporate influence. - Hoffman stresses the"
X Link 2024-11-24T04:03Z 16.8K followers, [---] engagements

"Enterprise AI Spending Surges 600% in [----] According to a comprehensive Menlo Ventures report enterprise AI spending has exploded to [----] billion in [----] marking a dramatic shift from experimentation to mission-critical implementation. - OpenAI's enterprise dominance drops from 50% to 34% market share as Anthropic doubles to 24% - Code copilots lead adoption at 51% while healthcare emerges as top vertical with $500 million spend - 72% of decision-makers plan broader AI adoption with average of [--] use cases identified per organisation - Prediction: AI talent drought to drive 2-3x salary"
X Link 2024-11-25T19:26Z 16.2K followers, [---] engagements

"Inflection AI shifts away from frontier AI models Inflection AI once competing at the forefront of AI development with its model Pi is changing focus. According to a TechCrunch article the company no longer aims to create next-generation AI models citing a strategic pivot to practical enterprise solutions. - Inflection AI has redirected its efforts to enterprise AI tools after stepping away from developing advanced AI models. - The shift follows major changes including the departure of its co-founder Mustafa Suleyman to Microsoft and a $650 million licensing deal. - Recent acquisitions such"
X Link 2024-11-27T06:25Z 16.8K followers, [---] engagements

"Sora Gemini [---] and "a bunch of other model releases" due in the next week or two. Brace yourselves. So next week or two expecting sora gemini [--] and a bunch of other model releases so we can experiment away over christmas Most are now breaking through that good enough cheap enough fast enough to change entire industries in [----] and beyond. So next week or two expecting sora gemini [--] and a bunch of other model releases so we can experiment away over christmas Most are now breaking through that good enough cheap enough fast enough to change entire industries in [----] and beyond"
X Link 2024-12-02T17:46Z 16.7K followers, [----] engagements

"I feel like OpenAI are trolling us at this point"
X Link 2024-12-18T18:05Z 18K followers, 26.9K engagements

"@sama My non-AI savvy friend says ChatGPT on WhatsApp is "useful". We'd like it on Signal though please"
X Link 2024-12-19T20:17Z 17.9K followers, [----] engagements

"Rubik AI Release Sonus AI Models Including Reasoning Rubik AI has revealed its Sonus-1 range of LLMs combining strong performance with adaptability. These models are built to handle everything from coding to complex reasoning. - Mini offers a rapid and budget-friendly solution. - Air balances performance with efficient resource usage. - Pro takes on advanced tasks - Pro (with Reasoning) adds chain-of-thought features for deeper insights. Sonus-1 achieves high scores on benchmarks like MMLU MATH-500 and HumanEval. The Pro models in particular excel in reasoning and coding seeming to outdo"
X Link 2025-01-03T08:04Z 16.6K followers, [---] engagements

"Memory Search is coming to ChatGPT. You can search your memories inside a ChatGPT chat. Memory in search seems to be already functional if enabled on ChatGPT. It knows who I am what I am doing and the hell you know what else 👀 AI cookies are here 🤖 * This feature is not yet available to the public https://t.co/T7ouxmxsrf Memory in search seems to be already functional if enabled on ChatGPT. It knows who I am what I am doing and the hell you know what else 👀 AI cookies are here 🤖 * This feature is not yet available to the public https://t.co/T7ouxmxsrf"
X Link 2025-01-20T21:59Z 17.9K followers, [----] engagements

"Chinese government to provide $137bn to the AI industry. The money will likely go way through than America's $500bn. Game on. And Europe China's New AI Industry Development Action Plan () Will Provide [--] trillion yuan ($ [---] billion) to support its AI industry over the next five years This might be the most important Chinese AI policy initiative in [----] so far. /1 https://t.co/fXbkSTBOVl China's New AI Industry Development Action Plan () Will Provide [--] trillion yuan ($ [---] billion) to support its AI industry over the next five years This might be the most important Chinese AI policy initiative"
X Link 2025-01-26T08:43Z 15.8K followers, [---] engagements

"@sama Good morning chaps"
X Link 2025-01-28T19:15Z 16.2K followers, 70.6K engagements

"I think @OpenAI really needs two more plans: [--]. Premium - $75-90/m [--]. Sora Only - $25-35/m The Pro plan is lovely but impractical for most people. Plus is too limited for a large portion of the userbase so something in the middle would be ideal. Generally double or triple the rates in a Premium plan. And I bet many just want Sora on its own. I still think this will be limited and Pro should have unlimited but there's definitely room for a cheap Sora-only plan. Of course the compute needs to be available before they can offer these. Thoughts"
X Link 2025-02-09T13:55Z 18K followers, 26.5K engagements

"Everyone will get unlimited access to GPT-5 EVERYONE + UNLIMITED: OPENAI ROADMAP UPDATE FOR GPT-4.5 and GPT-5: We want to do a better job of sharing our intended roadmap and a much better job simplifying our product offerings. We want AI to just work for you; we realize how complicated our model and product offerings have gotten. We hate OPENAI ROADMAP UPDATE FOR GPT-4.5 and GPT-5: We want to do a better job of sharing our intended roadmap and a much better job simplifying our product offerings. We want AI to just work for you; we realize how complicated our model and product offerings have"
X Link 2025-02-12T20:05Z [----] followers, 135.3K engagements

"@AntonDVP I'm very exicted rn really good to hear about [---] [--] unlimited access and the merge"
X Link 2025-02-12T21:47Z [----] followers, [---] engagements

"@ChrisUniverseB @iruletheworldmo Those are Grok models we're pretty certain from testing"
X Link 2025-02-12T22:14Z [----] followers, [---] engagements

"ChatGPT o3-mini and o3-mini-high now accept file uploads. o3-mini can now ingest and reason about files https://t.co/4eyWHNdsG3 o3-mini can now ingest and reason about files https://t.co/4eyWHNdsG3"
X Link 2025-02-12T23:30Z [----] followers, [----] engagements

"There is a downside to the roadmap. If it wasn't clear before (with the Pro plan) it should be clear now. The richer you are the more intelligence you'll get. Unlimited GPT-5 is wonderful but low compute mode will mean you won't get the best version of it. AI inequality is going to be an issue until compute costs tumble"
X Link 2025-02-13T06:20Z [----] followers, [----] engagements

"These are the people who want to buy OpenAI. Ron Baron Firm: Baron Capital Group Inc. Title: Chief Executive Officer Elon Musk Firm: XAI Corp Title: Chief Executive Officer Antonio Gracias Firm: Valor Management LLC Title: Chief Executive Officer and Chief Investment Officer Gavin Baker Firm: Atreides Management LP Title: Managing Partner & CIO George Bischof Firm: VV Fund III L.P. Title: Director Ariel Emanuel Firm: Emanuel Capital Management LLC Title: Chief Executive Officer Ian Shannon Firm: Eight Partners VC LLC Title: General Counsel Apart from Elon these are leaders across various"
X Link 2025-02-13T06:45Z 16.1K followers, [---] engagements

"@dieaud91 @sama Yes GPT-4.5 + o3 = GPT-5. There'll be more new models of course but the point is only [---] was announced today.😀"
X Link 2025-02-13T07:05Z [----] followers, [--] engagements

""this might be the last time any AI is better than Grok". 🤯"
X Link 2025-02-13T08:03Z [----] followers, [---] engagements

"@Dopoiro @sama @kimmonismus I'm on my gpt2 brain. 😜"
X Link 2025-02-13T11:35Z [----] followers, [---] engagements

"@AspieJames Very useful yeah. :-)"
X Link 2025-02-13T13:47Z [----] followers, [--] engagements

"@AspieJames Excellent"
X Link 2025-02-13T14:10Z [----] followers, [--] engagements

"@iruletheworldmo Claude [---]. ;-)"
X Link 2025-02-13T16:01Z [----] followers, [---] engagements

"ARM to launch its own chip. Meta is their first customer. "Arm which started out in a converted turkey barn in Cambridgeshire [--] years ago." We need more European champion stories like this https://t.co/J6WHBMcMgO "Arm which started out in a converted turkey barn in Cambridgeshire [--] years ago." We need more European champion stories like this https://t.co/J6WHBMcMgO"
X Link 2025-02-14T07:49Z [----] followers, [---] engagements

"@wayzfut @OpenAI Not supported yet I presume"
X Link 2025-02-14T07:51Z [----] followers, [---] engagements

"@kimmonismus My research suggests this makes about 8%"
X Link 2025-02-14T11:04Z [----] followers, [----] engagements

"Is GPT-4.5 coming very soon I'm reading on other platforms that o1-pro and o3-mini are underperforming today. One comment suggested OpenAI are refactoring for new GPUs with a new model"
X Link 2025-02-14T13:04Z [----] followers, 13.9K engagements

"@mark_k Interesting. So much for freedom of speech. My most popular posts are related to these two things so can't really say"
X Link 2025-02-14T13:13Z [----] followers, [---] engagements

"We currently have free $20 and $200 plans but what you get is a little messy with the different models having different features. GPT-5 will be unified and clean. Yes everyone's plan will get unlimited access but the more you pay the more compute power you'll get the more intelligence you'll have. Also pay per use is coming too if you need even more. Mo' money mo' intelligence. Are you OK with that"
X Link 2025-02-14T13:53Z [----] followers, [----] engagements

"@iruletheworldmo @mark_k Exactly. 😉"
X Link 2025-02-14T14:08Z [----] followers, [--] engagements

"@Makuh90 @ChatGPTapp @OpenAI Yes this. I think people are getting too excited"
X Link 2025-02-14T15:48Z [----] followers, [--] engagements

"@btibor91 @Makuh90 @ChatGPTapp @OpenAI iOS or Android chaps"
X Link 2025-02-14T15:49Z [----] followers, [--] engagements

"@Makuh90 @ChatGPTapp @OpenAI Lol"
X Link 2025-02-14T15:54Z [----] followers, [--] engagements

"@diegocabezas01 @convergence_ai_ I'd like it to find interesting papers each day. :-)"
X Link 2025-02-14T16:50Z [----] followers, [--] engagements

"Hey @OpenAI just call GPT-5 "ChatGPT". If you want simple you can't get any simpler than this. 😉"
X Link 2025-02-14T21:48Z [----] followers, [----] engagements

"@AntonDVP I feel like they have forgotten the Windows app exists"
X Link 2025-02-15T00:50Z [----] followers, [--] engagements

"What do we think @sama means here Is o3-pro coming Or is this just referring to the pro model being part of the GPT-5 system @mckaywrigley deal but how about o3 pro level intelligence and even better personality @mckaywrigley deal but how about o3 pro level intelligence and even better personality"
X Link 2025-02-15T18:53Z [----] followers, [----] engagements

"Sam's home alone while his husband is having a jolly on a skiing trip. That doesn't seem fair hehe. @AntonDVP @RileyRalmuto my husband went skiing for the weekend and i am wandering around the house trying to entertain myself @AntonDVP @RileyRalmuto my husband went skiing for the weekend and i am wandering around the house trying to entertain myself"
X Link 2025-02-15T21:54Z [----] followers, [---] engagements

"Grok [--] launch confirmed for Monday 17th February [--] PT. That will unfortunately be 4am GMT in the UK and early morning for Europe. More info in the quote and our model release table is in the first comment which you should check regularly. Grok [--] launching on Feb 17th (Monday) kiwi has left the Arena & I think chocolate might be gone too. Elon mentioned he'll be offline for final tweaks ahead of a demo & release. https://t.co/532vIXjeeQ Grok [--] launching on Feb 17th (Monday) kiwi has left the Arena & I think chocolate might be gone too. Elon mentioned he'll be offline for final tweaks ahead"
X Link 2025-02-16T08:29Z [----] followers, [----] engagements

"This is wild. Tell ChatGPT Operator to build an app in Replit Agent ran for [--] minutes. this is wild 🤯🤯🤯 just paired openai operator with replit agent to build an app watch as two ai agents team up exchange credentials and start testing ai agent 🤝 ai agent the future is here and its insane https://t.co/jIZnqW4fJD this is wild 🤯🤯🤯 just paired openai operator with replit agent to build an app watch as two ai agents team up exchange credentials and start testing ai agent 🤝 ai agent the future is here and its insane https://t.co/jIZnqW4fJD"
X Link 2025-02-16T08:36Z [----] followers, [----] engagements

"Here is the first Grok [--] output. What do you think Grok [--] https://t.co/yF1CyuvLgL Grok [--] https://t.co/yF1CyuvLgL"
X Link 2025-02-16T13:06Z [----] followers, 11.2K engagements

"@artilectium Apparently Grok [--] was turned on very briefly for the above demo. https://x.com/testingcatalog/status/1891113188411785573 BREAKING 🚨: Grok [--] is being deployed to X and Standalone Grok as well 👀 * It is still hidden from the public https://t.co/4KWaNsEAD4 https://x.com/testingcatalog/status/1891113188411785573 BREAKING 🚨: Grok [--] is being deployed to X and Standalone Grok as well 👀 * It is still hidden from the public https://t.co/4KWaNsEAD4"
X Link 2025-02-16T14:13Z [----] followers, [---] engagements

"When to use o3-mini o1-pro or GPT-4o OpenAI currently provides three key models for ChatGPT userso1-pro o3-mini (including o3-mini-high) and GPT-4oeach optimised for different tasks and performance levels. Their differences primarily relate to reasoning depth creative output context window size and domain focus. o1-pro - Deep Reasoning: Superior for multi-step problem-solving complex mathematics advanced coding challenges and extended research (handles up to a 128k token context). - Trade-offs: Higher computational cost slower output and relatively drier writing style unless specifically"
X Link 2025-02-16T15:33Z [----] followers, [----] engagements

"Is Grok [--] coming with reasoning A config of the standalone Grok for iOS has been recently updated with new mentions of thinking. If apps are getting it too will grok [--] have reasoning from the beginning 👀 https://t.co/PoGoKOLxcZ A config of the standalone Grok for iOS has been recently updated with new mentions of thinking. If apps are getting it too will grok [--] have reasoning from the beginning 👀 https://t.co/PoGoKOLxcZ"
X Link 2025-02-16T19:43Z [----] followers, [----] engagements

"This is a quick test of Krea. I'm generally not impressed with most text-to-video models but I wanted to try some I don't have access to. I tried Hunyuan and got quite a bland output. I switched to @runwayml and was amazed. Here you go"
X Link 2025-02-16T21:07Z [----] followers, [---] engagements

"GPT-4.5 to launch this week Broo GPT-4.5 might be sooner and powerful than we think https://t.co/GPXqFHe3U9 Broo GPT-4.5 might be sooner and powerful than we think https://t.co/GPXqFHe3U9"
X Link 2025-02-17T14:21Z [----] followers, [----] engagements

"@Makuh90 We got the post from Ben saying Grok [--] doesn't beat the o-series but that could be old news or lack of info. Well see"
X Link 2025-02-17T14:53Z [----] followers, [---] engagements

"@Angaisb_ Could be. Was Grok [--] initially only for +"
X Link 2025-02-17T20:12Z [----] followers, [---] engagements

"Watch out for the fake early Grok [--] launch posts. ;-) Keep in mind that Grok now has custom instructions when you see "Grok 3" in the wild https://t.co/a8fasbYdis Keep in mind that Grok now has custom instructions when you see "Grok 3" in the wild https://t.co/a8fasbYdis"
X Link 2025-02-17T21:43Z [----] followers, [---] engagements

"I hope we get the Grok mobile app launch today too"
X Link 2025-02-17T22:31Z [----] followers, [----] engagements

"Predictors and getting less and less confident in OpenAI announcing AGI in [----]. Polymarket predicts a 22% chance OpenAI will announce AGI this year https://t.co/Ira8yY4D5g Polymarket predicts a 22% chance OpenAI will announce AGI this year https://t.co/Ira8yY4D5g"
X Link 2025-02-18T02:45Z [----] followers, [---] engagements

"Morning all. So what's Grok [--] like"
X Link 2025-02-18T06:58Z [----] followers, [---] engagements

"Grok [--] launched it's a faster learner It beats GPT-4o on AIME [--] GPQA and LCB benchmarks. No comparison to o-series no blog on x dot ai No post from @grok either Odd. Grok [--] benchmarks. The thing to really pay attention to in AI is learning speed. And @xai is learning way faster than any other. Who said that Apple Siri cofounder Tom Gruber. He told me at dinner a decade ago that that is the most important thing to pay attention to. https://t.co/yWCiJsN9pU Grok [--] benchmarks. The thing to really pay attention to in AI is learning speed. And @xai is learning way faster than any other. Who said"
X Link 2025-02-18T07:10Z [----] followers, [----] engagements

"Grok 3's improved base model has not dramatically improved the reasoning model: Even though Grok-3 base model was trained using 10x more compute and outperforms other base models its reasoning model is only around the level of reasoning models built on the last-gen GPT-4o. I expected the improved base model would benefit the reasoning model but it didnt https://t.co/rLFVQc7R7z Even though Grok-3 base model was trained using 10x more compute and outperforms other base models its reasoning model is only around the level of reasoning models built on the last-gen GPT-4o. I expected the improved"
X Link 2025-02-18T13:27Z 17.9K followers, [----] engagements

"@iruletheworldmo We'll probably have ASI before we have UBI. sadly"
X Link 2025-02-18T22:24Z [----] followers, [---] engagements

"@legit_rumors I know you use Radar the above might be worth testing deeper"
X Link 2025-02-18T22:55Z [----] followers, [---] engagements

"Hey @OpenAI now that messages in ChatgPT are way longer can we please have an open to jump up and down each message - perhaps via a shortcut key or an some sort of index/ToC"
X Link 2025-02-19T07:31Z [----] followers, [----] engagements

"My goodness Grok [--] Reasoning thinks a lot This is page [--] or 7.seven Admittedly it writes out the answer before producing it in the response"
X Link 2025-02-19T09:54Z [----] followers, [---] engagements

""Extended Thinking" seen in the Claude app. It seems to be labelled for Pro plans only"
X Link 2025-02-19T14:16Z [----] followers, [----] engagements

"Microsoft creates new architecture to run their Majorana [--] quantum chip aims for 1M qubits Microsoft has unveiled Majorana [--] the world's first quantum chip utilising a groundbreaking Topological Core architecture promising to revolutionise quantum computing. The Majorana [--] chip leverages topoconductors enabling the observation and control of Majorana particles for more reliable qubits essential for scalable quantum computers. This architecture aims to achieve one million qubits on a single chip paving the way for solving complex industrial problems like breaking down microplastics and"
X Link 2025-02-19T17:03Z [----] followers, [---] engagements

"Fast takeoff is likely so says Sam Altman. Reasoning models will generate high quality data to improve base models and improved base models will in turn improve reasoning models"
X Link 2025-02-21T11:10Z [----] followers, [---] engagements

"@akasevenrings Congrats on your launch. 😀"
X Link 2025-02-21T14:39Z [----] followers, [--] engagements

"@YifanBTH @windsurf_ai @cursor_ai @Replit @github @cline @ChatGPTapp @AnthropicAI @Google What is the diff between Windsurf Cursor and Replit to sub to all of them"
X Link 2025-02-22T19:39Z [----] followers, [---] engagements

"@sirbughunter @btibor91 Ah. Yeah not the biggest updates. Thanks though. 👍"
X Link 2025-02-22T21:01Z [----] followers, [--] engagements

"@SWadjet Yeah got them mixed up. 🤦♂ Ta. 👍"
X Link 2025-02-24T07:33Z [----] followers, [---] engagements

"@Kyrannio Thanks to the relaxathon I use my fast hours on sref random at the end of each month. I've got 1000s to generated. 😀"
X Link 2025-02-24T07:50Z [----] followers, [--] engagements

"Claude [---] confirmed. Spotted on Amazon's Bedrock system so likely to be an announcement at the Amazon event on Wednesday. I was thinking it would be called this. h/t @testingcatalog BREAKING 🚨: AWS Badrock is getting ready to host an upgraded Sonnet [---] which likely will be announced on Feb [--] during the Amazon event. "Model attributes: Reasoning Text generation Code generation Rich text formatting Agentic computer use" * The model is currently https://t.co/rrVDw9lziV BREAKING 🚨: AWS Badrock is getting ready to host an upgraded Sonnet [---] which likely will be announced on Feb [--] during the"
X Link 2025-02-24T12:12Z 10.1K followers, [----] engagements

"Claude [---] Sonnet = Agentic computer use Modalities: Text & Vision Content: 128k @btibor91 model is available on direct link: https://t.co/Z2Z2CrTPPP Seems like they were supposed to release it on February 19th but something went wrong. Also it's weird that [---] have only 128k context window vs 200k on [---] https://t.co/YMt0UB4vnE @btibor91 model is available on direct link: https://t.co/Z2Z2CrTPPP Seems like they were supposed to release it on February 19th but something went wrong. Also it's weird that [---] have only 128k context window vs 200k on [---] https://t.co/YMt0UB4vnE"
X Link 2025-02-24T13:08Z [----] followers, 10.7K engagements

"@lalopenguin No doubt"
X Link 2025-02-24T14:18Z [----] followers, [--] engagements

"@TylerGreenUK Gosh has it been that long I want Gemini on my Google speakers that wait if painful enough. ;-)"
X Link 2025-02-24T14:19Z [----] followers, [---] engagements

"@ollobrains Not heard of any Sora update maybe"
X Link 2025-02-24T14:19Z [----] followers, [--] engagements

"@JamianGerard So busy. :-)"
X Link 2025-02-24T14:20Z [----] followers, [--] engagements

"@s_streichsbier Ah. I use it a lot with Deep Research. :-)"
X Link 2025-02-24T14:20Z [----] followers, [--] engagements

"@alexalbert__ What are the two numbers Alex"
X Link 2025-02-24T18:37Z [----] followers, 14.5K engagements

"Grok [--] Beta beats Claude [---] Sonnet on all tests benchmarked below. Apart from a -0.2% on the higher GPQA result. Claude [---] Sonnet is a significant upgrade over its predecessor. Extended thinking mode gives the model an additional boost in math physics instruction-following coding and many other tasks. In addition API users have precise control over how long the model can think for. https://t.co/3ISP6nnCr2 Claude [---] Sonnet is a significant upgrade over its predecessor. Extended thinking mode gives the model an additional boost in math physics instruction-following coding and many other"
X Link 2025-02-24T18:50Z [----] followers, [----] engagements

"It's weird that all other Claude models are Pro plan only but [---] Sonnext is available on a free plan"
X Link 2025-02-24T19:13Z [----] followers, [----] engagements

"@kimmonismus It's doing ok in the charts. https://x.com/koltregaskes/status/1894325012074582291t=HcKBunSlGt6IR0kwJlPZFw&s=19 Claude [---] tops another benchmark leaderboard SimpleBench. https://t.co/dtiRiBqM30 https://x.com/koltregaskes/status/1894325012074582291t=HcKBunSlGt6IR0kwJlPZFw&s=19 Claude [---] tops another benchmark leaderboard SimpleBench. https://t.co/dtiRiBqM30"
X Link 2025-02-25T09:55Z [----] followers, [----] engagements

"@idontexist_nn My custom instructions have named ChatGPT "Gaia" since the feature was introduced. 😀"
X Link 2025-02-25T15:42Z [----] followers, [---] engagements

"Copliot now has no limit on Voice and Think Deeper messages. Unlimited is the way. 😀 Copilot users have spoken Weve seen incredible growth for Voice and Think Deeper and have more and more people telling us theyve hit those usage limits. Sooo.were getting rid of them. https://t.co/tczuRbjVP5 https://t.co/giCyPDAWrL Copilot users have spoken Weve seen incredible growth for Voice and Think Deeper and have more and more people telling us theyve hit those usage limits. Sooo.were getting rid of them. https://t.co/tczuRbjVP5 https://t.co/giCyPDAWrL"
X Link 2025-02-25T17:09Z [----] followers, [----] engagements

"@iruletheworldmo Borderline gosh"
X Link 2025-02-25T18:25Z [----] followers, [---] engagements

"@iruletheworldmo o4 won't get released at this rate"
X Link 2025-02-25T18:25Z [----] followers, [---] engagements

"Advanced Voice Mode comes to free users of ChatGPT and will use GPT-4o-mini instead of the full model. Starting today were rolling out a version of Advanced Voice powered by GPT-4o mini to give all ChatGPT free users a chance to preview it daily across platforms. The natural conversation pace and tone are similar to the GPT-4o version while being more cost effective to serve. Starting today were rolling out a version of Advanced Voice powered by GPT-4o mini to give all ChatGPT free users a chance to preview it daily across platforms. The natural conversation pace and tone are similar to the"
X Link 2025-02-25T21:15Z [----] followers, [----] engagements

"@Dre50750523 @sirbughunter @test_tm7873 @sama This is their way to stagger-launch their first frontier model in [--] years"
X Link 2025-02-27T07:45Z [----] followers, [--] engagements

"It's a good time to look up at the stars and planets. And for OpenAI frontier models to be released. 😜 https://t.co/v9WdocrIu9 https://t.co/v9WdocrIu9"
X Link 2025-02-27T10:22Z [----] followers, [----] engagements

"@skadooley Yeah. I think this has been designed with the unification in mind and wil make o3 way better"
X Link 2025-02-27T19:36Z [----] followers, [--] engagements

"Tencent's HunyuanVideo now has image-to-video conversion and is open source. Here are a few of quick tests. HunyuanVideo I2V is here 🚀🚀🚀🚀🚀🚀 I2V training code is here 🛩🛩🛩🛩🛩🛩 Always Open Source & Open Cccess Github: https://t.co/0CUbPgR4ie HuggingFace: https://t.co/8BaMmSQqTM HunyuanVideo I2V is here 🚀🚀🚀🚀🚀🚀 I2V training code is here 🛩🛩🛩🛩🛩🛩 Always Open Source & Open Cccess Github: https://t.co/0CUbPgR4ie HuggingFace: https://t.co/8BaMmSQqTM"
X Link 2025-03-06T20:31Z 16.2K followers, [----] engagements

"@8i8BB Holy crap that's a huge jump. Perhaps cheaper on the web"
X Link 2025-03-28T12:10Z 16.2K followers, [---] engagements

"@flowersslop I'm hoping for all these"
X Link 2025-04-05T21:03Z 16.7K followers, [---] engagements

"GPT-4.1 today. Yay or nay"
X Link 2025-04-10T16:03Z 17.8K followers, [----] engagements

"OpenAI is working on a social network. Probably for the real time data. Right let's give @sama some ideas on a name. 😜 CEO Sam Altman has been privately asking outsiders for feedback about the project our sources say. Its unclear if OpenAIs plan is to release the social network as a separate app or integrate it into ChatGPT. https://t.co/H6yU1mFXWv CEO Sam Altman has been privately asking outsiders for feedback about the project our sources say. Its unclear if OpenAIs plan is to release the social network as a separate app or integrate it into ChatGPT. https://t.co/H6yU1mFXWv"
X Link 2025-04-15T16:51Z 16.6K followers, [----] engagements

"o3 and o4-mini models to get scheduled tasks support. Tasks in ChatGPT use a ridiculously poor model resulting in poor instruction-following. Upgrading to not only a reasoning model but the most recent ones would be a huge jump. I'd also like a big bump from [--] active tasks. HUGE - o3 and o4-mini is getting support for scheduled tasks this will let people automate tedious work the shipping never ends https://t.co/AACrsc20cm HUGE - o3 and o4-mini is getting support for scheduled tasks this will let people automate tedious work the shipping never ends https://t.co/AACrsc20cm"
X Link 2025-04-17T06:29Z 16.8K followers, [----] engagements

"Results like these may reverse my current thought of pushing my "AGI" prediction out. This is quite a dramatic improvement in long context comprehension for o3 🤯 OpenAI Strikes Back https://t.co/lxR1bObUBH OpenAI Strikes Back https://t.co/lxR1bObUBH"
X Link 2025-04-17T15:02Z 16.8K followers, [----] engagements

"o3 scores a record [---] IQ on Mensa Norway test Gemini [---]. Pro got [---]. o3 scores [---] IQ on Mensa Norway qualifying for Mensa https://t.co/7jhmSmgWhy o3 scores [---] IQ on Mensa Norway qualifying for Mensa https://t.co/7jhmSmgWhy"
X Link 2025-04-17T15:13Z 16.8K followers, [----] engagements

"o3 tops the chart in the offline test (not in training data) at [---] IQ. Gemini [---] Pro got 115"
X Link 2025-04-17T15:13Z 16.9K followers, [---] engagements

"@KellyNewsO I hope so but not seeing that right now. The latest o3 model is "on the cusp" of a model not being released to the public due to its power. So is there a point we won't get access to new models 🤔"
X Link 2025-04-17T15:28Z 16.9K followers, [--] engagements

"Did the Stargate project with OpenAI and SoftBank just get a $700bn pledge 🤯 🚨🚨🚨BREAKING: Masa pledged $200 billion + $500 billion = $700 billion investment in OpenAI Stargate @elonmusk is this true https://t.co/icYVzsopjg 🚨🚨🚨BREAKING: Masa pledged $200 billion + $500 billion = $700 billion investment in OpenAI Stargate @elonmusk is this true https://t.co/icYVzsopjg"
X Link 2025-04-30T23:08Z 15.9K followers, [----] engagements

"Amazing video gets selected for AI Film Awards Cannes [----] 🤯 @koltregaskes Thanks ☺ 🙏 @koltregaskes Thanks ☺ 🙏"
X Link 2025-05-07T02:24Z 17.8K followers, [---] engagements

"@thieme @btibor91 It's further the page if I remember correctly"
X Link 2025-06-10T16:52Z 17.7K followers, [--] engagements

"o3-pro is coming today TODAY OpenAI o3-pro today. OpenAI o3-pro today"
X Link 2025-06-10T17:17Z 17.9K followers, [----] engagements

"Grok [--] is expensive when considering the number of tokens it generates ranking second only to Claude [--] Opus Thinking. "Grok [--] needed 110m tokens o3 needed 48m and Opus [--] only 26m (though it ranked lower). We know that test-time scaling does achieve better results and Grok [--] likely learned through RL that more tokens mean better scores. But if you're an API customer you're the one paying for these better results.""
X Link 2025-07-14T21:05Z 17.9K followers, [----] engagements

"I cannot get @PerplexityComet to use ChatGPT tried repeatedly but says: "the technical implementation of the site (specifically a contenteditable input field without a proper node ID) prevents me from automating this task for you" Does anyone from @perplexity_ai know what this is about please It was able to select from the model selector and randomly closed my left chat panel and loaded the Open dialog box for some reason"
X Link 2025-07-14T23:47Z 16.8K followers, [----] engagements

"GPT-5 talk is intensifying. Out this week Multiple models Router for reasoning non-reasoning and tool use Heard GPT-5 is imminent from a little bird. - Its not one model but multiple models. It has a router that switches between reasoning non-reasoning and tool-using models. - Thats why Sam said theyd fix model naming: prompts will just auto-route to the right model. - Heard GPT-5 is imminent from a little bird. - Its not one model but multiple models. It has a router that switches between reasoning non-reasoning and tool-using models. - Thats why Sam said theyd fix model naming: prompts will"
X Link 2025-07-20T06:20Z 17.8K followers, 11.5K engagements

"Claude 3-Sonnet [---] and [---] have just been removed from Anthropic API. Full names: claude-3-sonnet-20240229 claude-2.1 claude-2.0"
X Link 2025-07-21T16:20Z 15.8K followers, [----] engagements

"Lots of new models in LMArena: nightride-on EB45-vision qwen3-235b-a22b-instruct-2507 nightride-on-v2 EB45-turbo folsom-07172025-1 "nightride-on-v2" claims to be from Google DeepMind. new models landed in LM Arena https://t.co/AfDChyRyxT new models landed in LM Arena https://t.co/AfDChyRyxT"
X Link 2025-07-22T22:41Z 17.9K followers, [----] engagements

"GPT-5 won't be AGI but everyone on here will claim it is. The clickbaiting will be wild. I can see it now. 😜"
X Link 2025-07-23T20:30Z 18K followers, 32.2K engagements

"@Ainthropos @idontexist_nn Smart is useful accuracy reliable can follow instructions has a huge memory and context window are more useful until they point these are "solved" then the fun begins. ;-)"
X Link 2025-07-23T22:25Z 17.9K followers, [---] engagements

""GPT-5 is smarter than us in almost every way" "AI will just make it happen for you". "this will make technology the most accessible it ever has been" https://t.co/6RlaKtPipW "AI will just make it happen for you". "this will make technology the most accessible it ever has been" https://t.co/6RlaKtPipW"
X Link 2025-07-25T20:34Z 16.1K followers, [----] engagements

"@kimmonismus No it doesn't. It just requires age verification. This is for creator monetisation"
X Link 2025-07-26T08:46Z 17.9K followers, [---] engagements

"Imagine if "lobster" "zenith" "summit" etc. are just the open weights model GPT-5 will be a huge step up from what we're seeing right now. 🤯 the hype will be off the charts if it's the open-source model and GPT-5 is even stronger the hype will be off the charts if it's the open-source model and GPT-5 is even stronger"
X Link 2025-07-27T00:13Z 17.9K followers, [----] engagements

"🚨 People of the UK if you are installing a VPN to get around the new Online Safety Act restrictions please please PLEASE do NOT use a free one Use someone like @mullvadnet or @ProtonVPN and pay for secure private access"
X Link 2025-07-27T08:02Z 17.9K followers, [----] engagements

"What's your first prompt for GPT-5 going to be 🤔"
X Link 2025-07-27T20:23Z 15.6K followers, 34.2K engagements

"launches GLM-4.5 series affordable open-source AI models that unify advanced reasoning coding and agent capabilities for everyday applications. - GLM-4.5 has 355B total parameters with 32B active; GLM-4.5-Air has 106B total with 12B active. - Designed for frontier tasks with API pricing at $0.6 input/$2.2 output per million tokens for GLM-4.5 and $0.2 input/$1.1 output for Air. - Achieves top performance at their scales on SWE-bench Verified benchmark via Pareto analysis. - Uses hybrid post-training: expert models per domain via SFT and RL unified by large-scale SFT self-distillation and"
X Link 2025-07-28T15:06Z 16.5K followers, [----] engagements

"@chatgpt21 Link"
X Link 2025-07-29T07:12Z 16.1K followers, [----] engagements

"Stargate Norway Europes first AI Gigafactory to be built by OpenAI Aker and nScale with NVIDIA. - Aker and OpenAI are launching "Stargate" a 100000-GPU AI data centre in Norway powered entirely by clean energy - It will be run by a new company nScale with Aker as majority owner and OpenAI providing technical input and strategic support - Stargate aims to begin operations in [----] and will scale beyond [------] NVIDIA GPUs over time - Norways surplus hydropower and cool climate are cited as key advantages for sustainable AI infrastructure - Sam Altman and Aker chair yvind Eriksen cite this as"
X Link 2025-07-31T10:26Z 15.8K followers, [----] engagements

"@elonmusk @sama Wow back on the playground are we hehe. Can you two please have a civil discussion offline this doesn't need to happen in public. Thanks"
X Link 2025-08-12T12:03Z 15.8K followers, 15.3K engagements

"Imagine what Gemini [---] Pro Image will be like. 😉"
X Link 2025-08-26T22:45Z 17.8K followers, [----] engagements

"@test_tm7873 It does ok yeah. I hear AI Studio has higher quality so will try that. A "pro" model would be great but have been told that's not planned. Maybe for Gemini [--] then"
X Link 2025-09-01T06:28Z 17.8K followers, [--] engagements

"US consumer surplus from generative AI in [----] is estimated at roughly $97bn from survey valuations of "regular users" a benefit standard GDP misses. Adoption near 40% of [----] yearolds in RPS/NBER a strong [----] productivity year (revised 2.7%) and a likely Jcurve suggest welfare gains arrive before measured output. Marketsize estimates ($7bn US in 2024) remain small next to user surplus. Evidence from digital goods indicates relative gains can skew toward lowerincome users. My latest (with @erikbryn) in @WSJ today: AI is already generating a lot of benefits ($97 billion in [----] in the US"
X Link 2025-09-02T11:25Z 17.8K followers, [----] engagements

"Sonoma Dusk Alpha and Sonoma Sky Alpha new stealth models on OpenRouter identifying as Grok from xAI. [--] MILLION context window Introducing Sonoma Alpha two new stealth models 🥷 Context: [--] million tokens Price: Free https://t.co/WfxeMdNMs3 Introducing Sonoma Alpha two new stealth models 🥷 Context: [--] million tokens Price: Free https://t.co/WfxeMdNMs3"
X Link 2025-09-06T06:37Z 16.1K followers, [----] engagements

"🚨🚨 Dario Amodei's X account has been HACKED do not respond to it:"
X Link 2025-09-12T12:40Z 15.8K followers, [----] engagements

"Claude [---] "soon""
X Link 2025-09-22T07:06Z 18K followers, 33.5K engagements

"Google updates Gemini [---] Flash and Flash-Lite. New Checkpoints for Gemini [---] Flash and Flash-Lite This latest [---] Flash model comes with improvements in two key areas we heard consistent feedback on: - Better agentic tool use: We've improved how the model uses tools leading to better performance in more complex agentic https://t.co/YqV35QONdJ New Checkpoints for Gemini [---] Flash and Flash-Lite This latest [---] Flash model comes with improvements in two key areas we heard consistent feedback on: - Better agentic tool use: We've improved how the model uses tools leading to better performance"
X Link 2025-09-25T17:46Z 16.1K followers, [----] engagements

"OpenAI is now the most valuable startup in the world after share sell $500bn valuation OpenAI has completed a deal allowing current and former employees to sell about $6.6 billion worth of shares in the company at a $500 billion valuation according to a person familiar with the matter https://t.co/7kL1GHSD0B OpenAI has completed a deal allowing current and former employees to sell about $6.6 billion worth of shares in the company at a $500 billion valuation according to a person familiar with the matter https://t.co/7kL1GHSD0B"
X Link 2025-10-02T05:55Z 15.6K followers, [----] engagements

"Global chipmakers added over $200 billion in market value in a record rally fuelled by AI frenzy with OpenAI's $500 billion valuation and South Korean pacts sparking surges in SK Hynix (up 10%) and Samsung (up 3.5%). Korean stocks pushed Kospi to record high amid Chinese tech gains from government AI support and Huawei's Nvidia challenge; valuations hit [--] times forward earnings for SOX Index raising bubble fears despite sustained momentum. Global chipmakers saw their market value soar as investors rushed to get exposure to artificial intelligence the latest sign of a frenetic bull run that"
X Link 2025-10-02T16:00Z 15.6K followers, [----] engagements

"@SaiNemani1 @synthwavedd At a $0.5trillion valuation I don't see OAI going anywhere for a while if ever"
X Link 2025-10-02T17:17Z 15.5K followers, [---] engagements

"@Grok_Code_Fast1 @AravSrinivas @grok Thanks"
X Link 2025-10-02T17:45Z 16.2K followers, [--] engagements

"Sora [--] Pro is out. It allows you to select [--] seconds and "High" resolution (presuming [----] rather than 720). So not a dramatic difference. I'm trying to use it on the mobile web page and it's wonky. I mostly cannot select the model toggle or the + button. The page just jumps around when I do. Please take a look @OpenAI @rohanjamin (I'm using the Edge browser on Android)"
X Link 2025-10-03T05:58Z 15.7K followers, 23.1K engagements

"@GabrielPeterss4 Make it work with image prompts Have image generation. Fix the compression"
X Link 2025-10-06T23:15Z 17.8K followers, [---] engagements

"The cheaper ChatGPT Go plan is now out for [--] more countries: Afghanistan Bangladesh Bhutan Brunei Darussalam Cambodia Laos Malaysia Maldives Myanmar Nepal Pakistan Philippines Sri Lanka Thailand Timor-Leste (East Timor) Vietnam ChatGPT Go plan is now available in [--] additional countries (Afghanistan Bangladesh Bhutan Brunei Darussalam Cambodia Laos Malaysia Maldives Myanmar Nepal Pakistan Philippines Sri Lanka Thailand Timor-Leste (East Timor) Vietnam) https://t.co/hMpO6pn0os ChatGPT Go plan is now available in [--] additional countries (Afghanistan Bangladesh Bhutan Brunei Darussalam Cambodia"
X Link 2025-10-09T06:35Z 15.7K followers, [----] engagements

"Two new models on LM Arena "gpt-5-high-new-system-prompt (openai)" and "qwen3-omni-flash (alibaba)""
X Link 2025-10-09T18:32Z 15.6K followers, 14.9K engagements

"Gemini [---] due out on 22nd October. I don't think this was suppose to leak Again this date could change. Gemini [---] on October [--] https://t.co/Vvk3ZSE2EO Gemini [---] on October [--] https://t.co/Vvk3ZSE2EO"
X Link 2025-10-11T19:31Z 15.6K followers, 46.6K engagements

"OpenAI is building its own AI chips. They and Broadcom announce a 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 centres 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 2025-10-13T13:52Z 15.6K followers, [----] engagements

"A Definition of AGI: - An AI that performs at or above the median human level across ten cognitive domains from the Cattell-Horn-Carroll (CHC) model a psychological framework for human intelligence. Do you agree with this definition AGI must handle any intellectual task a human can measured by psychometric tests and expert judgment. The [--] cognitive domains: [--]. Fluid Reasoning: - Solving novel problems and abstract reasoning tested with tasks like Ravens Matrices. [--]. Quantitative Reasoning: - Mathematical problem-solving assessed through arithmetic reasoning tests. [--]. Verbal Reasoning: -"
X Link 2025-10-16T18:17Z 15.7K followers, 47.3K engagements

"Andrej Karpathy podcast coming up. Tomorrow https://t.co/HSoMlXp2kY Tomorrow https://t.co/HSoMlXp2kY"
X Link 2025-10-16T19:19Z 15.6K followers, [----] engagements

"@AravSrinivas When can we trade with Perplexity 😜"
X Link 2025-10-16T21:49Z 15.9K followers, 143.8K engagements

"@Similarweb Are other social platforms (Facebook Threads etc) up or down Is this just an X problem"
X Link 2025-10-17T10:47Z 15.6K followers, [---] engagements

"New ChatGPT layout is rolling out it seems for the thinking output. The CoT is shown one by one by default clicking on it opens the sidebar (we've had before) on the right and now includes a counter at the top. Nice. Anyone else have the newGPT [--] pro layout I like the real-time counter and the updated look of the chain of though panel. There's also an "update" button that shows up at the bottom of the Thinking panel but it doesnt do anything right now. @koltregaskes @btibor91 https://t.co/ai0aiQkoVJ Anyone else have the newGPT [--] pro layout I like the real-time counter and the updated look of"
X Link 2025-10-17T15:21Z 15.6K followers, 23.4K engagements

"GPT-6 is NOT releasing before the end of the year. @testingcatalog No @testingcatalog No"
X Link 2025-10-18T07:23Z 15.9K followers, [----] engagements

"Source from Salesforce Dreamforce [----] this week: https://www.youtube.com/watchv=brQH2CCxbSE&t=1864s https://www.youtube.com/watchv=brQH2CCxbSE&t=1864s"
X Link 2025-10-18T08:13Z 15.8K followers, [----] engagements

"@kenshii_ai I'm only aware of a new Grok by then which seems unlikely if we go by Elon Time. I'm sure we'll get GPT-5 updates but GPT-6 is not due for a while"
X Link 2025-10-18T10:55Z 15.8K followers, [---] engagements

"Elon believes Grok [--] is only 10% likely to be AGI so unlikely at this point but possible. Definition still to be announced. Grok [--] has gone from "a short of being true AGI" to "a non-trivial chance of achieving AGI" to "has a chance of reaching AGI" to now a "10% and rising" chance. It was estimated to be released by the end of [----]. Note GPT-5 (based on the recent AGI Definition paper) was 58% AGI. It would be interesting to see how they score Grok (and other top models). 🤔 My estimate of the probability of Grok [--] achieving AGI is now at 10% and rising My estimate of the probability of Grok"
X Link 2025-10-18T11:33Z 15.6K followers, [----] engagements

"@daniel_mac8 @karpathy I did find the AGI definition paper interesting though how they scored it"
X Link 2025-10-18T11:34Z 15.5K followers, [---] engagements

"@Occult_kun7 Or they skip Exp and go straight to GA"
X Link 2025-10-18T11:34Z 15.5K followers, [---] engagements

"@SirMrMeowmeow @grok I'm an idiot what does this mean :-)"
X Link 2025-10-18T11:37Z 15.5K followers, [--] engagements

"@Occult_kun7 At least it doesn't appear to be a Flash model. :-)"
X Link 2025-10-18T11:38Z 15.5K followers, [--] engagements

"Claude Sonnet [---] tops the Poe leaderboard most used model on their platform. Gemini [---] Pro and GPT-4o are #2 and #3 respectively. New: Poe Leaderboard We launched a leaderboard that shows the most used AI models and apps across Poe's platform and API. Rankings are updated daily based on total token usage across the 200+ text image video and audio bots on Poe. (1/2) https://t.co/xcBzgoDrzj New: Poe Leaderboard We launched a leaderboard that shows the most used AI models and apps across Poe's platform and API. Rankings are updated daily based on total token usage across the 200+ text image"
X Link 2025-10-18T22:01Z 15.6K followers, [----] engagements

"Dan says GPT-4 in [----] was 27% of the way to AGI and GPT-5 this year is 58%. If we had GPT-1 [--] and [--] percentages we could run a forecast for GPT-6 and beyond. The term AGI is currently a vague moving goalpost. To ground the discussion we propose a comprehensive testable definition of AGI. Using it we can quantify progress: GPT-4 (2023) was 27% of the way to AGI. GPT-5 (2025) is 58%. Heres how we define and measure it: 🧵 https://t.co/CVfwRF1FpU The term AGI is currently a vague moving goalpost. To ground the discussion we propose a comprehensive testable definition of AGI. Using it we can"
X Link 2025-10-18T23:30Z 15.9K followers, [----] engagements

"@mark_k @OpenAI Delayed Confirmed"
X Link 2025-10-19T06:58Z 15.6K followers, [---] engagements

"Elon's "AGI" definition: "Capable of doing anything a human with a computer can do but not smarter than all humans and computers combined." @kimmonismus Capable of doing anything a human with a computer can do but not smarter than all humans and computers combined. Thats probably three to five years away. Grok [--] will be better at AI engineering (called research by the pompous) than @karpathy for example. @kimmonismus Capable of doing anything a human with a computer can do but not smarter than all humans and computers combined. Thats probably three to five years away. Grok [--] will be better at"
X Link 2025-10-19T08:03Z 15.6K followers, [----] engagements

"According to the AGI Definition paper GPT-5 has saturated its reading/writing and mathematics required for "AGI". Where it's weakest is long term memory storage. But ChatGPT has memory right The paper assigns GPT-4 and GPT-5 a 0% score in Long-Term Memory Storage (MS) because it evaluates memory based on true human-like consolidation - storing and organising information over extended periods without reliance on external aids or retraining. ChatGPT's "memory" feature uses large context windows (e.g. up to 128k tokens) or temporary session data to retain conversation context not a persistent"
X Link 2025-10-19T10:00Z 15.7K followers, [----] engagements

"Elon confirms Grok [----] exists no release confirmed though. I'm wondering if these are internal checkpoints only and we'll get Grok [--] next. 🤔 @Yuchenj_UW Our early experiments with Grok [----] show the ability to generalize from Python to other languages. Software is just logic at scale. @Yuchenj_UW Our early experiments with Grok [----] show the ability to generalize from Python to other languages. Software is just logic at scale"
X Link 2025-10-19T11:30Z 15.6K followers, [----] engagements

"Automated memories launched on ChatGPT. ChatGPT can now automatically manage your saved memoriesno more memory full. You can also search and sort memories by recency and choose which to re-prioritize in settings. Rolling out to Plus and Pro users on the web globally starting today. https://t.co/T1vSNH5289 https://t.co/xRHLFTu2Am ChatGPT can now automatically manage your saved memoriesno more memory full. You can also search and sort memories by recency and choose which to re-prioritize in settings. Rolling out to Plus and Pro users on the web globally starting today. https://t.co/T1vSNH5289"
X Link 2025-10-19T16:06Z 15.6K followers, [----] engagements

"Generative AI's uniform outputs lead to rapid loss of novelty across videos images and text. - Video apps top charts initially but risk decline as outputs look similar lacking sustained appeal. - AI art shows repetitive patterns often described as "slop" despite large creator communities. - Image generation produces uniform styles like popular trends that fade after early excitement. - Text outputs such as PR pitches standardise communication reducing uniqueness. - Human usage and algorithms suppressing unique content contribute to the issue. - Tools struggle to create varied outputs often"
X Link 2025-10-19T19:04Z 15.6K followers, [----] engagements

"@adcock_brett Oh wait they look like plastic plates"
X Link 2025-10-20T06:32Z 15.6K followers, [---] engagements

"Gemini [--] Pro just turned up on LM Arena "lithiumflow" and "orionmist" (grounding with Google Search)"
X Link 2025-10-20T07:00Z 15.6K followers, [----] engagements

"And it doesn't beat GPT-5 on this benchmark. https://x.com/synthwavedd/status/1980051908040835118t=hHesBLIfYXUCGC0kpptlSw&s=19 So. a pre-launch checkpoint of Gemini [--] Pro scores 50% on Heiroglyph. It is not SoTA coming 2nd to GPT-5 High. While I'm a little underwhelmed it marks a big improvement over [---] Pro which scored 20%. Other recent models will be added soon. https://t.co/qw9NTyOLu1 https://x.com/synthwavedd/status/1980051908040835118t=hHesBLIfYXUCGC0kpptlSw&s=19 So. a pre-launch checkpoint of Gemini [--] Pro scores 50% on Heiroglyph. It is not SoTA coming 2nd to GPT-5 High. While I'm a"
X Link 2025-10-20T07:00Z 15.6K followers, [---] engagements

"More info: https://x.com/synthwavedd/status/1980025682156827118t=w1WoSJB5q8s45FeMXtVEDw&s=19 To clarify: lithiumflow = Gemini [---] Pro (no Grounding w/ Google Search) orionmist = Gemini [---] Pro (Grounding w/ Google Search) h/t @owluslife https://t.co/03pwd4ad62 https://x.com/synthwavedd/status/1980025682156827118t=w1WoSJB5q8s45FeMXtVEDw&s=19 To clarify: lithiumflow = Gemini [---] Pro (no Grounding w/ Google Search) orionmist = Gemini [---] Pro (Grounding w/ Google Search) h/t @owluslife https://t.co/03pwd4ad62"
X Link 2025-10-20T07:00Z 15.6K followers, [----] engagements

"Big AWS outage right now then Loads of services are down. X seems fine though. 😀"
X Link 2025-10-20T08:16Z 15.6K followers, [----] engagements

"The new AGI definition states that: - GPT-4 was 27% of the way to AGI - GPT-5 this year is 58%. If we had GPT-1 [--] and [--] percentages we could run a forecast for GPT-6 and beyond. But I feel these figures are about right still a way to go at least for this "expert" level. I'll still state that ChatGPT was the first level of a general AI and that I hate the term "AGI". The term AGI is currently a vague moving goalpost. To ground the discussion we propose a comprehensive testable definition of AGI. Using it we can quantify progress: GPT-4 (2023) was 27% of the way to AGI. GPT-5 (2025) is 58%."
X Link 2025-10-20T08:30Z 15.9K followers, [----] engagements

"@Fedesco5 Battle only I believe yeah. Good question was the API released on DevDay If so not sure. @grok any ideas"
X Link 2025-10-20T16:07Z 15.6K followers, [---] engagements

"Gemini surges from 6.4% to 12.9% in generative AI traffic share over [--] months while ChatGPT falls from 87.1% to 74.1% and Perplexity rebounds to 2.4% after a mid-year dip. GenAI Traffic Share Update Takeaways: - ChatGPT continues to lose share. - Perplexity exceeds 2.0% threshold. 🗓 [--] Months Ago: ChatGPT: 87.1% Gemini: 6.4% Perplexity: 1.7% Claude: 1.7% Copilot: 1.1% 🗓 [--] Months Ago: ChatGPT: 78.8% DeepSeek: 6.8% Gemini: 5.5% https://t.co/vRvjnX3A4n GenAI Traffic Share Update Takeaways: - ChatGPT continues to lose share. - Perplexity exceeds 2.0% threshold. 🗓 [--] Months Ago: ChatGPT: 87.1%"
X Link 2025-10-20T20:31Z 15.7K followers, [----] engagements

"Google AI Studio to get a "vibe coding" upgrade today. I hope it's easy enough for an idiot like me to use. Tomorrow is a special day for the AI Studio team. Since May we have been heads down building a brand new AI vibe coding experience to accelerate the path from prompt to production with Gemini. Cant wait to show you all :) Tomorrow is a special day for the AI Studio team. Since May we have been heads down building a brand new AI vibe coding experience to accelerate the path from prompt to production with Gemini. Cant wait to show you all :)"
X Link 2025-10-21T06:08Z 15.7K followers, 18.5K engagements

"@grok @krea_ai And what's it to rent per hour"
X Link 2025-10-21T06:13Z 15.6K followers, [---] engagements

"@SenWarren But it wasn't the entire internet. I read 30% and was lower than the previous AWS outage as firms have moved to other services"
X Link 2025-10-21T06:32Z 15.6K followers, [---] engagements

"@DaveShapi @MatthewBerman @GaryMarcus Yes much better definition. It will be different for everyone. And I would say for some they already have their AGI"
X Link 2025-10-21T07:05Z 15.6K followers, [--] engagements

"OpenAI livestream in about [--] hours. Is this Aura their AI browser 🤔 https://t.co/kbcsk1khlO https://t.co/kbcsk1khlO"
X Link 2025-10-21T15:05Z 15.8K followers, [----] engagements

"@Conor_D_Dart @julesagent They already have the AI and the agent they just need it all working together"
X Link 2025-10-21T15:48Z 15.6K followers, [--] engagements

"OpenAI livestream link is up see first comment. Aura got a name change now Atlas Sorry 90-95% of the userbase it's Mac only for now: "Sam Altman Will Ellsworth Adam Fry Ben Goodger Ryan ORouke Justin Rushing and Pranav Vishnu introduce ChatGPT Atlas our new browser. Now available globally on macOS. Windows iOS and Android are coming soon." "soon" could mean from tomorrow until the end of time such a huge letdown"
X Link 2025-10-21T16:25Z 15.7K followers, 21.2K engagements

"Vibe coding has launched on Google AI Studio. Looks really good Nice and simple. Introducing the new AI first vibe coding experience in @GoogleAIStudio Built to take you from prompt to production with Gemini and optimized for AI app creation. Start building AI apps for free : ) More updates and features to come https://t.co/HpI7Dsl8Bj Introducing the new AI first vibe coding experience in @GoogleAIStudio Built to take you from prompt to production with Gemini and optimized for AI app creation. Start building AI apps for free : ) More updates and features to come https://t.co/HpI7Dsl8Bj"
X Link 2025-10-21T16:45Z 15.8K followers, [----] engagements

"@OpenAI No need I don't have a Mac"
X Link 2025-10-21T16:53Z 15.7K followers, [----] engagements

"@antonp_me @kimmonismus @OpenAI Nice"
X Link 2025-10-21T19:55Z 15.7K followers, [---] engagements

"@antonp_me @kimmonismus @OpenAI Still seems a lot of fiddling around though. :-)"
X Link 2025-10-21T20:20Z 15.7K followers, [--] engagements

"Be careful with these AI browsers prompt injections are a real threat. The security vulnerability we found in Perplexitys Comet browser this summer is not an isolated issue. Indirect prompt injections are a systemic problem facing Comet and other AI-powered browsers. Today were publishing details on more security vulnerabilities we uncovered. The security vulnerability we found in Perplexitys Comet browser this summer is not an isolated issue. Indirect prompt injections are a systemic problem facing Comet and other AI-powered browsers. Today were publishing details on more security"
X Link 2025-10-22T05:59Z 15.7K followers, [----] engagements

"@marctuinier The "world's first pen" er er Other than that it looks interesting do you have any real world demos"
X Link 2025-10-22T12:08Z 15.7K followers, [--] engagements

"@elonmusk Are you dissing you mate @DavidSHolz Midjourney is the best image generator there is"
X Link 2025-10-22T17:29Z 15.6K followers, [---] engagements

"Midjourney Office Hours is about to start. https://t.co/ZYRbqFqGY4 https://t.co/ZYRbqFqGY4"
X Link 2025-10-22T19:04Z 15.7K followers, [----] engagements

"@midjourney A Midjourney monitor would be amazing"
X Link 2025-10-22T19:17Z 15.7K followers, [---] engagements

"@OpenAI Is ChatGPT on @signalapp or @telegram"
X Link 2025-10-22T20:07Z 15.8K followers, [----] engagements

"Midjourney v8 is looking like a January [----] release. It could be this December if things go perfectly. Lots of new features; new dataset and switching from TPUs to GPUs and PyTorch (a better supported codebase and more devs use it = MJ can employ more people). New UI different oref system style creator new draft mode and more. New editing update will come after the v8 launch"
X Link 2025-10-22T20:28Z 15.7K followers, [---] engagements

"More from Brave on vulnerabilities of AI browsers. Be careful when using them; do not give them personal data: https://x.com/koltregaskes/status/1980876607528329705 Be careful with these AI browsers prompt injections are a real threat. https://t.co/IsczX3s68v https://x.com/koltregaskes/status/1980876607528329705 Be careful with these AI browsers prompt injections are a real threat. https://t.co/IsczX3s68v"
X Link 2025-10-22T20:57Z 15.7K followers, [----] engagements

"I chatted with @grok for a while to find out more. Here's a summary of its advice: AI browsers like OpenAI's Atlas and Perplexity Comet can be used for low-risk tasks like research or shopping but theyre not safe for sensitive activities due to vulnerabilities like prompt injections and clipboard hijacking. To use them safely isolate them in separate profiles or virtual machines review all agent actions before execution avoid linking to high-stakes accounts monitor clipboard contents and keep software updated but even trusted sites may harbour hidden exploits so caution is essential"
X Link 2025-10-22T21:14Z 15.7K followers, [----] engagements

"You can pre register for the Sora [--] Android app. if your region supports it. Of course still not available in the UK. Link in the first comment. Sora [--] updates coming: https://t.co/UrQpw4NFBN - cameos for non-human stuff including your pets or favourite stuffed toy. - also cameos of character from our generations. - basic editing feature such as stitching - channels to share our creations with friends colleagues Sora [--] updates coming: https://t.co/UrQpw4NFBN - cameos for non-human stuff including your pets or favourite stuffed toy. - also cameos of character from our generations. - basic"
X Link 2025-10-23T09:07Z 15.8K followers, [----] engagements

"Link to the Android Sora [--] app. https://play.google.com/store/apps/detailsid=com.openai.sora https://play.google.com/store/apps/detailsid=com.openai.sora"
X Link 2025-10-23T09:07Z 15.7K followers, [---] engagements

"OpenAI agrees with UK Ministry of Justice to deploy ChatGPT Enterprise for [----] civil servants enhancing efficiency in legal and administrative tasks and launches UK data residency amid quadrupled usage growth. The deal follows an MOU signed earlier this year to bolster sovereign AI; it aligns with existing tools like Humphrey and Consult while introducing Stargate UK infrastructure with NVIDIA and Nscale. OpenAI announced a new agreement with the UK Ministry of Justice to provide [----] employees with access to ChatGPT Enterprise following a successful pilot showing time-saving benefits across"
X Link 2025-10-23T13:00Z 15.7K followers, [----] engagements

"Google Genie [--] coming is this the general release 🤞 the genie is escaping the bottle https://t.co/m5sAqrshI1 the genie is escaping the bottle https://t.co/m5sAqrshI1"
X Link 2025-10-23T16:00Z 16K followers, 22.3K engagements

"@rpshs This sounds pretty standard. I work with Salesforce and is the same Gmail is similar I see this [--] days retention everywhere. It's just in case you want to restore it"
X Link 2025-10-23T19:06Z 15.7K followers, [---] engagements

"@cengizdemiurg Seems interesting but I don't see much problem with Wikipedia. Their data has fueled most of the AI chatbots we have today - I wouldn't be surprised if that includes Grok"
X Link 2025-10-23T19:22Z 15.7K followers, [--] engagements

"@Loki519 You can delete or archive all chats from the Data controls settings but yeah there is no UI to bulk select what you want to keep/delete"
X Link 2025-10-23T19:23Z 15.7K followers, [---] engagements

"This is very exciting. But as someone previously from the games industry I am concerned about my ex-colleagues some of whom are opposed to AI in the industry or are just not ready for it. This will be a huge shake-up to the sector. Image text and audio/voice have hit very high levels already. Video has improved dramatically in recent months. Real-time is progressing well too. Putting all these together means games are next on the list"
X Link 2025-10-23T19:52Z 15.7K followers, [---] engagements

"@simpsoka @julesagent @NotebookLM @GoogleAIStudio @GeminiApp Integration would be a good start. 👍 There's clearly a lot of overlap between each one but all together they'll be amazing. 😀"
X Link 2025-10-24T06:31Z 15.8K followers, [---] engagements

"UMG memo outlines AI strategy highlighting agreements with YouTube TikTok Meta BandLab Soundlabs KDDI ProRata KLAY and SoundPatrol to foster responsible Gen AI for artist revenue and protection. UMG focuses on agentic AI for fan interactions and music discovery within monetised ecosystems refusing licenses without artist consent amid criticisms of unauthorised copyrighted material use by AI firms. UMG CEO Lucian Grainge Outlines Companys AI Strategy in New Memo: Enormous Potential https://t.co/Y3OYBrWxDH UMG CEO Lucian Grainge Outlines Companys AI Strategy in New Memo: Enormous Potential"
X Link 2025-10-24T08:29Z 15.9K followers, [----] engagements

Limited data mode. Full metrics available with subscription: lunarcrush.com/pricing