Dark | Light
[GUEST ACCESS MODE: Data is scrambled or limited to provide examples. Make requests using your API key to unlock full data. Check https://lunarcrush.ai/auth for authentication information.]

[@slow_developer](/creator/twitter/slow_developer)
"and we've just got a SORA X leak today last week sam and some openAI employees hinted that gpt-5 is releasing soon maybe they replace the older Sora with SORA X and integrate it into GPT-5 but openAI won't release it unless it's as good as or better than Veo 3"  
![@slow_developer Avatar](https://lunarcrush.com/gi/w:16/cr:twitter::1461305890892570628.png) [@slow_developer](/creator/x/slow_developer) on [X](/post/tweet/1947290238394524000) 2025-07-21 13:39:06 UTC 46.7K followers, 10.7K engagements


"Ilya Sutskever says the real challenge of AI is how fast and extreme its growth will be it still has limits but it already speaks codes and interacts in ways once impossible soon perhaps within X to XX years AI will be able to do everything humans can do the reason: "the human brain is a biological computer so why can't a digital brain do the same""  
![@slow_developer Avatar](https://lunarcrush.com/gi/w:16/cr:twitter::1461305890892570628.png) [@slow_developer](/creator/x/slow_developer) on [X](/post/tweet/1931995935099498865) 2025-06-09 08:45:00 UTC 46.7K followers, 146.3K engagements


"this should be a bigger deal than winning IMO openAI researcher said: "their IMO model "knew"; it didn't have the right answer" does he mean hallucination-free if models know what they know and when to ask for human help we could finally trust them with real-world tasks"  
![@slow_developer Avatar](https://lunarcrush.com/gi/w:16/cr:twitter::1461305890892570628.png) [@slow_developer](/creator/x/slow_developer) on [X](/post/tweet/1947644045121220936) 2025-07-22 13:05:00 UTC 46.7K followers, 10.8K engagements


"@scaling01 10m or unlimited context windows mean nothing unless you solve MEMORY that's the real bottleneck"  
![@slow_developer Avatar](https://lunarcrush.com/gi/w:16/cr:twitter::1461305890892570628.png) [@slow_developer](/creator/x/slow_developer) on [X](/post/tweet/1946551234925904128) 2025-07-19 12:42:34 UTC 46.6K followers, 1192 engagements


"chatGPT has outperformed nearly all elite programmers (only one survived) in a competition focusing on optimization algorithms openAI model entered the finals on July XX led early but was overtaken by Psyho a former oAI member key takeaways: - openAI internal coding model is insane - this could be the last human win"  
![@slow_developer Avatar](https://lunarcrush.com/gi/w:16/cr:twitter::1461305890892570628.png) [@slow_developer](/creator/x/slow_developer) on [X](/post/tweet/1945801907064442954) 2025-07-17 11:05:00 UTC 46.6K followers, 43.6K engagements


"i'm confused by this. how advanced AI whether it's AGI or ASI could create new jobs we already know it will automate many existing ones but if AI becomes capable of doing everything humans can do what would be left for people to do"  
![@slow_developer Avatar](https://lunarcrush.com/gi/w:16/cr:twitter::1461305890892570628.png) [@slow_developer](/creator/x/slow_developer) on [X](/post/tweet/1944830505402200311) 2025-07-14 18:45:00 UTC 46.7K followers, 64.7K engagements


"@burkov @TheLudd AGI was achieved long ago. it's meaningless now the term is used to build the hype for funding"  
![@slow_developer Avatar](https://lunarcrush.com/gi/w:16/cr:twitter::1461305890892570628.png) [@slow_developer](/creator/x/slow_developer) on [X](/post/tweet/1947600806460465659) 2025-07-22 10:13:11 UTC 46.7K followers, XXX engagements


"🚨 Meta CEO Mark Zuckerberg predicts smart glasses will replace phones by 2030 " Smart glasses are going to become the next major computing platform. They will gradually replace phones by 2030 much like mobile devices surpassed computers without fully replacing them ""  
![@slow_developer Avatar](https://lunarcrush.com/gi/w:16/cr:twitter::1461305890892570628.png) [@slow_developer](/creator/x/slow_developer) on [X](/post/tweet/1840441784888901810) 2024-09-29 17:21:29 UTC 46.7K followers, 43.4M engagements


"AI agents arent good enough yet at least in coding. models like gemini XXX pro and o3/o3 pro are great for debugging .but for agent tasks even claude code often goes in circles or gets stuck wasting api calls the best reasoning models are impressive -- but using them constantly for agents is still too expensive"  
![@slow_developer Avatar](https://lunarcrush.com/gi/w:16/cr:twitter::1461305890892570628.png) [@slow_developer](/creator/x/slow_developer) on [X](/post/tweet/1939436213154332781) 2025-06-29 21:30:00 UTC 46.6K followers, 8602 engagements


""in the future Agents should act for you without being asked" OpenAI Isa Fulford says the next big focus is improving how we collaborate with AI agents Multi-turn chats work well but the goal is smoother more continuous interactions personalization and memory will be key"  
![@slow_developer Avatar](https://lunarcrush.com/gi/w:16/cr:twitter::1461305890892570628.png) [@slow_developer](/creator/x/slow_developer) on [X](/post/tweet/1947696967666307411) 2025-07-22 16:35:18 UTC 46.7K followers, 7880 engagements


"after ChatGPT Agent and Grok X heavy the focus shifted to new benchmarks HLE and Frontier Math are still important but agentic benchmarks now matter more for most ppl intelligence is wide enough what's needed now is agent ability: tools memory context and modalities"  
![@slow_developer Avatar](https://lunarcrush.com/gi/w:16/cr:twitter::1461305890892570628.png) [@slow_developer](/creator/x/slow_developer) on [X](/post/tweet/1946936636249461074) 2025-07-20 14:14:01 UTC 46.7K followers, 6734 engagements


"subscription pricing mess. that's why we need open-source models to keep pressure on the big AI labs even if DeepSeek Qwen and KIMI might not be the best but without open-source competition gemini and openAI wouldn't be offering as much for free"  
![@slow_developer Avatar](https://lunarcrush.com/gi/w:16/cr:twitter::1461305890892570628.png) [@slow_developer](/creator/x/slow_developer) on [X](/post/tweet/1944003812991111264) 2025-07-12 12:00:01 UTC 46.7K followers, 71.9K engagements


"total domination by openAI this week: - launched chatGPT Agent; matched or beat humans in half the tasks even 10+ hours - mysterious o3-alpha claimed to be the best coding model - gpt-5 reasoning alpha leak - internal experimental model achieved IMO gold-level performance"  
![@slow_developer Avatar](https://lunarcrush.com/gi/w:16/cr:twitter::1461305890892570628.png) [@slow_developer](/creator/x/slow_developer) on [X](/post/tweet/1946663835873816967) 2025-07-19 20:10:00 UTC 46.7K followers, 12.3K engagements


"@jpreagan_ what if AGI is more about broad cognitive capability than pure economic utility the models today aren't perfect but their generality feels. different now"  
![@slow_developer Avatar](https://lunarcrush.com/gi/w:16/cr:twitter::1461305890892570628.png) [@slow_developer](/creator/x/slow_developer) on [X](/post/tweet/1946802930516218294) 2025-07-20 05:22:43 UTC 46.6K followers, XXX engagements


"MAJOR BLOW TO OPENAI Meta is rumored to have poached two more high-profile openAI researchers: Jason Wei co-author of the scaling laws paper and a lead researcher on agentic models and reasoning Hyung Won Chung head of Codex and a core architect behind GPT-4 the 'o' series and deep research brutal hit. sam has now lost several of his closest allies and top researchers"  
![@slow_developer Avatar](https://lunarcrush.com/gi/w:16/cr:twitter::1461305890892570628.png) [@slow_developer](/creator/x/slow_developer) on [X](/post/tweet/1945405544459571346) 2025-07-16 08:50:00 UTC 46.7K followers, 11.1K engagements


"more info about GPT-5 the release is imminent -- possibly within the next two weeks depending on safety tests "it's a router system that switches b/w reasoning basic and tool-use models; GPT-6 is in training" this is confusing because openAI earlier said a single unified model"  
![@slow_developer Avatar](https://lunarcrush.com/gi/w:16/cr:twitter::1461305890892570628.png) [@slow_developer](/creator/x/slow_developer) on [X](/post/tweet/1946896619909362095) 2025-07-20 11:35:00 UTC 46.7K followers, 23.5K engagements


"this should be bigger news. "ChatGPT Agent matched or beat humans in about half the tasks tested even ones that take over XX hours" you kidding me we're only a little over halfway into the year of AI agents and they're already completing high-impact tasks"  
![@slow_developer Avatar](https://lunarcrush.com/gi/w:16/cr:twitter::1461305890892570628.png) [@slow_developer](/creator/x/slow_developer) on [X](/post/tweet/1946247341838835968) 2025-07-18 16:35:00 UTC 46.7K followers, 29.9K engagements


"Anthropic co-founder Ben Mann: There's about a XX% chance we'll reach superintelligence within a few years it's not just speculation; it's based on data scaling laws and clear scientific trends even if we get there soon the societal effects may take time to unfold"  
![@slow_developer Avatar](https://lunarcrush.com/gi/w:16/cr:twitter::1461305890892570628.png) [@slow_developer](/creator/x/slow_developer) on [X](/post/tweet/1946977150344593883) 2025-07-20 16:55:00 UTC 46.7K followers, 22.4K engagements


"the AI community i know is slowly turning into a cult around certain AI labs without truly understanding how things work. the race to AGI isn't about picking your favorite lab support labs doing valuable work and use whatever model helps you most"  
![@slow_developer Avatar](https://lunarcrush.com/gi/w:16/cr:twitter::1461305890892570628.png) [@slow_developer](/creator/x/slow_developer) on [X](/post/tweet/1946286348853223715) 2025-07-18 19:10:00 UTC 46.6K followers, 6379 engagements


"where' the damn "ChatGPT Agent" for plus users"  
![@slow_developer Avatar](https://lunarcrush.com/gi/w:16/cr:twitter::1461305890892570628.png) [@slow_developer](/creator/x/slow_developer) on [X](/post/tweet/1947629452931768584) 2025-07-22 12:07:01 UTC 46.7K followers, 21.3K engagements


"Ex-DeepMind researcher Misha Laskin says labs quietly changed the meaning of AGI to superintelligence The goalposts are shifting. what counted as AGI before is now called superintelligence but the definition is still vague unless we solve safety we're playing an Indiana Jones game that ends badly"  
![@slow_developer Avatar](https://lunarcrush.com/gi/w:16/cr:twitter::1461305890892570628.png) [@slow_developer](/creator/x/slow_developer) on [X](/post/tweet/1946364362773438602) 2025-07-19 00:20:00 UTC 46.7K followers, 24.3K engagements


"MORE DRAMA daniel gross wanted ilya to sell the company to Meta crazy since he was SSI ceo at the time. ilya is a GOAT researcher but something feels off about SSI if they're close to a breakthrough why did their CEO leave sooner or later ilya has to step out and show what's going on"  
![@slow_developer Avatar](https://lunarcrush.com/gi/w:16/cr:twitter::1461305890892570628.png) [@slow_developer](/creator/x/slow_developer) on [X](/post/tweet/1947368478861570220) 2025-07-21 18:50:00 UTC 46.7K followers, 7137 engagements


"Elon AGI bet: scale fast scale hard he announced Colossus X will host over 550k gb200s and gb300s in a few weeks that's ridiculously massive. that means xAI will soon be training on more compute than any other AI lab by a huge margin colossus X already runs 230k GPUs (including 30k gb200s) for grok training colossus X takes that to a whole new level"  
![@slow_developer Avatar](https://lunarcrush.com/gi/w:16/cr:twitter::1461305890892570628.png) [@slow_developer](/creator/x/slow_developer) on [X](/post/tweet/1947891929368383985) 2025-07-23 05:30:00 UTC 46.7K followers, 2129 engagements


"why keep moving the AGI bar 10-15 years ago the AI we have today would've been called AGI frontier models now outperform most people in many tasks and understand things that seemed impossible in 2000 it's not that we've redefined intelligence -- we've just kept raising the bar"  
![@slow_developer Avatar](https://lunarcrush.com/gi/w:16/cr:twitter::1461305890892570628.png) [@slow_developer](/creator/x/slow_developer) on [X](/post/tweet/1946046014990725415) 2025-07-18 03:15:00 UTC 46.6K followers, 7772 engagements


"have you noticed every major lab is now talking about superintelligence -- because it's clear we're already crossing the AGI threshold the term 'AGI' feels meaningless now if you'd shown o3 gemini XXX or grok X to someone in 2017 theyd think we were well past it"  
![@slow_developer Avatar](https://lunarcrush.com/gi/w:16/cr:twitter::1461305890892570628.png) [@slow_developer](/creator/x/slow_developer) on [X](/post/tweet/1944592702290432042) 2025-07-14 03:00:03 UTC 46.7K followers, 682.8K engagements


"bullish on those who think openAI is done just because a model was delayed chatGPT is the 3rd most visited website globally ahead of X no X on the Apple App Store in the U.S unless someone delivers something significantly better and cheaper oAI stays on top no need to hit the panic button"  
![@slow_developer Avatar](https://lunarcrush.com/gi/w:16/cr:twitter::1461305890892570628.png) [@slow_developer](/creator/x/slow_developer) on [X](/post/tweet/1944635502746882263) 2025-07-14 05:50:08 UTC 46.6K followers, 11.1K engagements


"ARC-AGI X is already here we haven't even completed half of ARC-AGI X and now there's ARC-3 and wasn't the test meant to tell us when we've reached AGI now the models are getting close they keep making new tests and shifting the goalposts Turing test passed ARC-AGI X passed still not AGI"  
![@slow_developer Avatar](https://lunarcrush.com/gi/w:16/cr:twitter::1461305890892570628.png) [@slow_developer](/creator/x/slow_developer) on [X](/post/tweet/1946325355452912082) 2025-07-18 21:45:00 UTC 46.7K followers, 8741 engagements


"Grok X ranked 2nd with a XXXX% score on the updated SimpleBench it's a great model and truly SOTA in areas like: math science reasoning also this isn't even grok X heavy that one should be even better xAI needs to launch a grok-4-heavy API and it must be affordable"  
![@slow_developer Avatar](https://lunarcrush.com/gi/w:16/cr:twitter::1461305890892570628.png) [@slow_developer](/creator/x/slow_developer) on [X](/post/tweet/1946819864070611410) 2025-07-20 06:30:00 UTC 46.7K followers, 4626 engagements


"imagine the scenario: this is the moment when "poached" openAI researchers were quietly automated by a GPT-5 agent not officially fired -- just quietly automated but what are the odds of that happening"  
![@slow_developer Avatar](https://lunarcrush.com/gi/w:16/cr:twitter::1461305890892570628.png) [@slow_developer](/creator/x/slow_developer) on [X](/post/tweet/1947527023347273852) 2025-07-22 05:20:00 UTC 46.7K followers, 5710 engagements


"Amjad Masad says true AGI means an AI that can enter new situations and quickly learn how to achieve goals Even the latest models still struggle with tasks outside their training data. The real breakthrough will come when AI can train itself -- like AlphaGo playing billions of games against itself"  
![@slow_developer Avatar](https://lunarcrush.com/gi/w:16/cr:twitter::1461305890892570628.png) [@slow_developer](/creator/x/slow_developer) on [X](/post/tweet/1940856822052581573) 2025-07-03 19:35:00 UTC 46.7K followers, 161.4K engagements


"UBE - universal basic employment makes more sense than UBI job shortages are mainly a private sector issue so as long as humans can still add value its better to have them work UBI won't happen until almost all jobs are gone which will take a very long time"  
![@slow_developer Avatar](https://lunarcrush.com/gi/w:16/cr:twitter::1461305890892570628.png) [@slow_developer](/creator/x/slow_developer) on [X](/post/tweet/1945284748320120919) 2025-07-16 00:50:00 UTC 46.7K followers, 13.2K engagements


"the best bet for GPT-5 release is august it was recently spotted in a biotech repo as "gpt-5 reasoning" and sam said it's releasing soon o3-alpha (maybe their open-weight) is claimed to be SOTA coding model openAI just launched "ChatGPT Agent" so no need to drop gpt-5 yet so august makes the most sense"  
![@slow_developer Avatar](https://lunarcrush.com/gi/w:16/cr:twitter::1461305890892570628.png) [@slow_developer](/creator/x/slow_developer) on [X](/post/tweet/1947734641957605401) 2025-07-22 19:05:00 UTC 46.7K followers, 3939 engagements


"Mysterious Model Alert a new model "o3 Alpha" by openAI claims to be better than o3-pro and google Kingfall but what exactly is it full version of Codex delayed open-weight model fine-tuned o3 version for chatGPT Agent probably not GPT-5 since they won't share it before launch"  
![@slow_developer Avatar](https://lunarcrush.com/gi/w:16/cr:twitter::1461305890892570628.png) [@slow_developer](/creator/x/slow_developer) on [X](/post/tweet/1946404639609442503) 2025-07-19 03:00:03 UTC 46.6K followers, 9921 engagements


"@vasalex93 even humans screw up stuff like that too miscount overlook get tired forget context do we call them non-intelligent"  
![@slow_developer Avatar](https://lunarcrush.com/gi/w:16/cr:twitter::1461305890892570628.png) [@slow_developer](/creator/x/slow_developer) on [X](/post/tweet/1946657262849675291) 2025-07-19 19:43:53 UTC 46.6K followers, XXX engagements


"@polynoamial @OpenAI which model exactly Noam"  
![@slow_developer Avatar](https://lunarcrush.com/gi/w:16/cr:twitter::1461305890892570628.png) [@slow_developer](/creator/x/slow_developer) on [X](/post/tweet/1946479346786480601) 2025-07-19 07:56:54 UTC 46.7K followers, 13.2K engagements


"@SeSphoenix or maybe we're all expecting AGI to look and act like us with this seamless inner monologue .but what if general intelligence ends up working in a way thats just super different from how humans think"  
![@slow_developer Avatar](https://lunarcrush.com/gi/w:16/cr:twitter::1461305890892570628.png) [@slow_developer](/creator/x/slow_developer) on [X](/post/tweet/1946804650004668672) 2025-07-20 05:29:33 UTC 46.6K followers, XXX engagements


"i forgot we're only allowed to talk about things once they're "officially" solved by your standards"  
![@slow_developer Avatar](https://lunarcrush.com/gi/w:16/cr:twitter::1461305890892570628.png) [@slow_developer](/creator/x/slow_developer) on [X](/post/tweet/1946801504763158757) 2025-07-20 05:17:03 UTC 46.6K followers, XXX engagements


"full detail thread: zuck offers $1B to openAI CRO"  
![@slow_developer Avatar](https://lunarcrush.com/gi/w:16/cr:twitter::1461305890892570628.png) [@slow_developer](/creator/x/slow_developer) on [X](/post/tweet/1947210211271290995) 2025-07-21 08:21:06 UTC 46.7K followers, 15.3K engagements


"Demis Hassabis says true AGI is a theoretical benchmark set by the human brain the only proven architecture for general intelligence todays models are impressive but inconsistent; anyone can find flaws within minutes. "real AGI should be so strong that it would take experts months to spot a weakness""  
![@slow_developer Avatar](https://lunarcrush.com/gi/w:16/cr:twitter::1461305890892570628.png) [@slow_developer](/creator/x/slow_developer) on [X](/post/tweet/1925086124633923763) 2025-05-21 07:07:53 UTC 46.7K followers, 458.9K engagements


"video source: and if you don't want to miss major AI news you can always subscribe to the newsletter it includes: - top X% industry news - important research papers with summaries - read by 250k+ AI developers"  
![@slow_developer Avatar](https://lunarcrush.com/gi/w:16/cr:twitter::1461305890892570628.png) [@slow_developer](/creator/x/slow_developer) on [X](/post/tweet/1947009109934436557) 2025-07-20 19:02:00 UTC 46.7K followers, 1264 engagements


"Mustafa Suleyman says the possibility of superintelligence arriving should keep everyone awake we've never controlled anything as powerful as us let alone something far smarter if we get it right we'll solve our energy health and food crises "but getting it right is mind-blowingly hard""  
![@slow_developer Avatar](https://lunarcrush.com/gi/w:16/cr:twitter::1461305890892570628.png) [@slow_developer](/creator/x/slow_developer) on [X](/post/tweet/1944474408904540404) 2025-07-13 19:10:00 UTC 46.7K followers, 31.1K engagements


"elon was right. i pasted my entire python codebase into Grok X -- a full data pipeline with CSV parsing and API calls grok X quickly found poor function structure weak error handling and messy logic then suggested solid fixes most impressive it understood the whole pipeline and helped the modules work better together if you maintain complex or old python code grok X can save you a lot of time"  
![@slow_developer Avatar](https://lunarcrush.com/gi/w:16/cr:twitter::1461305890892570628.png) [@slow_developer](/creator/x/slow_developer) on [X](/post/tweet/1943612128100753748) 2025-07-11 10:03:36 UTC 46.7K followers, 1.1M engagements


"we need to be honest with 18-year-olds for most university is pointless. you take on a lot of debt learn very little and struggle to find a job afterward AI has already reduced the value of higher education for about XX% of students with an IQ below XXX the old path is broken"  
![@slow_developer Avatar](https://lunarcrush.com/gi/w:16/cr:twitter::1461305890892570628.png) [@slow_developer](/creator/x/slow_developer) on [X](/post/tweet/1945682369492488548) 2025-07-17 03:10:00 UTC 46.7K followers, 32.5K engagements


"BIG NEWS openAI experimental reasoning model reached IMO gold-level performance and they are also releasing GPT-5 soon new Q-star moment looks like they're using new reasoning methods that go beyond standard CoT techniques it also confirms they're running internal models months ahead of anything public"  
![@slow_developer Avatar](https://lunarcrush.com/gi/w:16/cr:twitter::1461305890892570628.png) [@slow_developer](/creator/x/slow_developer) on [X](/post/tweet/1946504688116945316) 2025-07-19 09:37:36 UTC 46.7K followers, 10.9K engagements


"BIG NEWS "Google announced an advanced version of gemini DeepThink won gold at the IMO" the model was trained on past IMOs with RL and multi-step reasoning surely google will likely be releasing advanced models at a faster pace now the world is not yet ready for the GPT-5 vs Gemini 3"  
![@slow_developer Avatar](https://lunarcrush.com/gi/w:16/cr:twitter::1461305890892570628.png) [@slow_developer](/creator/x/slow_developer) on [X](/post/tweet/1947407485985235064) 2025-07-21 21:25:00 UTC 46.7K followers, 7778 engagements


"important point to note. openAI researcher Jerry Tworek says "GPT-5 will not include the breakthrough of the IMO-winning model it's a later model probably end-of-the-year thing" that means GPT-5 will be released soon so there's no time to integrate that breakthrough into it"  
![@slow_developer Avatar](https://lunarcrush.com/gi/w:16/cr:twitter::1461305890892570628.png) [@slow_developer](/creator/x/slow_developer) on [X](/post/tweet/1946704101490016471) 2025-07-19 22:50:00 UTC 46.7K followers, 7133 engagements


"humanity is cooked. 💀 a person is addicted to Ani subscribed with money and reached level X he adds "it's too late for me to come back""  
![@slow_developer Avatar](https://lunarcrush.com/gi/w:16/cr:twitter::1461305890892570628.png) [@slow_developer](/creator/x/slow_developer) on [X](/post/tweet/1946446151789805811) 2025-07-19 05:45:00 UTC 46.6K followers, 3252 engagements


"strongly disagree. these systems walks like intelligence talks like intelligence beats humans at reasoning tasks but it's not "real" intelligence because we don't fully understand how it works that's kind of moving the goalpost. we don't understand our own brains in great detail either yet no one questions if we're intelligent"  
![@slow_developer Avatar](https://lunarcrush.com/gi/w:16/cr:twitter::1461305890892570628.png) [@slow_developer](/creator/x/slow_developer) on [X](/post/tweet/1947552503366881327) 2025-07-22 07:01:15 UTC 46.7K followers, XXX engagements


"Eric Schmidt asks what happens when AI creates millions of polymaths History's greatest discoveries came from experts applying ideas from one field to another something AI can't do yet "the rules keep changing" solving non-stationarity might bring breakthroughs but it'll take years to know"  
![@slow_developer Avatar](https://lunarcrush.com/gi/w:16/cr:twitter::1461305890892570628.png) [@slow_developer](/creator/x/slow_developer) on [X](/post/tweet/1946208343409758590) 2025-07-18 14:00:02 UTC 46.7K followers, 22.4K engagements


"what could be the next openAI open-weight model right now weve got: o3 o3-pro o4-mini o4-mini-high and the rumors say its a reasoning model sam has hinted many times even asked in polls about the idea of o3-mini running locally on phones so its either o3-mini or o4-mini being released as open weights"  
![@slow_developer Avatar](https://lunarcrush.com/gi/w:16/cr:twitter::1461305890892570628.png) [@slow_developer](/creator/x/slow_developer) on [X](/post/tweet/1939684005500084333) 2025-06-30 13:54:39 UTC 46.6K followers, 3674 engagements


"openAI researcher says "GPT-5 is coming" the key point: it will be different from the current reasoning setup that relies on RL and tool use that could mean: - new architecture or full end-to-end RL - stronger tool use or closer to a world model - can think longer or scale deeper based on the question"  
![@slow_developer Avatar](https://lunarcrush.com/gi/w:16/cr:twitter::1461305890892570628.png) [@slow_developer](/creator/x/slow_developer) on [X](/post/tweet/1946170032972607991) 2025-07-18 11:27:48 UTC 46.7K followers, 24.8K engagements


"google is terrible at marketing openAI explains things and educates the public they had a breakthrough: RL with unverified rewards and they're explaining what that means google on the other hand drops something like this. THIS they're both smart and dumb at once"  
![@slow_developer Avatar](https://lunarcrush.com/gi/w:16/cr:twitter::1461305890892570628.png) [@slow_developer](/creator/x/slow_developer) on [X](/post/tweet/1947566745524240829) 2025-07-22 07:57:50 UTC 46.7K followers, 4175 engagements


"i'm convinced we've already crossed the AGI threshold the term "AGI" has lost its meaning every time we get close the goalposts shift maybe only a superintelligence can look back and define what AGI truly was otherwise we'll be stuck debating it forever"  
![@slow_developer Avatar](https://lunarcrush.com/gi/w:16/cr:twitter::1461305890892570628.png) [@slow_developer](/creator/x/slow_developer) on [X](/post/tweet/1946624829232406938) 2025-07-19 17:35:00 UTC 46.7K followers, 47K engagements


"Grok X feels much more human than other AIs after using gemini XXX pro o3 and others Grok felt more natural and less robotic DeepSearch is highly powerful and unbiased no point in using google search anymore these are the things that matter most to non-technical users"  
![@slow_developer Avatar](https://lunarcrush.com/gi/w:16/cr:twitter::1461305890892570628.png) [@slow_developer](/creator/x/slow_developer) on [X](/post/tweet/1947131920070099096) 2025-07-21 03:10:00 UTC 46.7K followers, 144.4K engagements


"all eyes are on GPT-5 the model wont just be a picker; it should be a full system that integrates everything more intelligently id guess DeepMind is also heading the same way with its upcoming gemini models with improved memory and longer context it could feel like talking to your AGI"  
![@slow_developer Avatar](https://lunarcrush.com/gi/w:16/cr:twitter::1461305890892570628.png) [@slow_developer](/creator/x/slow_developer) on [X](/post/tweet/1941644512138494309) 2025-07-05 23:45:00 UTC 46.7K followers, 12.7K engagements


"Logan was right. only two weeks have passed and we've already seen major AI milestones: - models winning gold at the IMO - launch of the ChatGPT Agent - news about openAI unreleased top-tier coding model - two SOTA open-source models released: Qwen and Kimi k2 everything is scaling up fast"  
![@slow_developer Avatar](https://lunarcrush.com/gi/w:16/cr:twitter::1461305890892570628.png) [@slow_developer](/creator/x/slow_developer) on [X](/post/tweet/1947605963239805190) 2025-07-22 10:33:41 UTC 46.7K followers, 15.6K engagements


"this is so frustrating. i get that gpt-4.5 uses a lot of GPU but what is a regular plus user supposed to do with only XX prompts a week sure the model is impressive at emotional intelligence still what's the point of releasing it if people can barely use it"  
![@slow_developer Avatar](https://lunarcrush.com/gi/w:16/cr:twitter::1461305890892570628.png) [@slow_developer](/creator/x/slow_developer) on [X](/post/tweet/1947850404450209849) 2025-07-23 02:45:00 UTC 46.7K followers, 10.9K engagements


"Sam Altman says some jobs will disappear many will evolve and entirely new ones will emerge. OpenAI's growth means it's making people more efficient not replacing them "we're growing so fast that theres still so much more to do" As for hardware it's going to be great but no smart glasses"  
![@slow_developer Avatar](https://lunarcrush.com/gi/w:16/cr:twitter::1461305890892570628.png) [@slow_developer](/creator/x/slow_developer) on [X](/post/tweet/1942852472013480149) 2025-07-09 07:45:00 UTC 46.7K followers, 79.4K engagements


"Geoffrey Hinton says LLMs aren't just mimicking they're making creative connections When asked why a compost heap is like an atom bomb GPT-4 pointed to chain reactions which showed it understood abstract links "it saw the analogy during training not test time" so they're seeing analogies to compress knowledge like humans do"  
![@slow_developer Avatar](https://lunarcrush.com/gi/w:16/cr:twitter::1461305890892570628.png) [@slow_developer](/creator/x/slow_developer) on [X](/post/tweet/1945642104111464593) 2025-07-17 00:30:00 UTC 46.7K followers, 46.2K engagements


"sam altman: perfect AI = tiny model superhuman reasoning 1T context tokens i agree with him. DO YOU the X important things missing for AGI now are large context size and strong reasoning current models like gemini have 1m tokens but after 120- 150k tokens they lose more than half the information a billion-token context with full retrieval could reach AGI even at XXX pro or o3 reasoning levels"  
![@slow_developer Avatar](https://lunarcrush.com/gi/w:16/cr:twitter::1461305890892570628.png) [@slow_developer](/creator/x/slow_developer) on [X](/post/tweet/1930377773115121807) 2025-06-04 21:35:00 UTC 46.6K followers, 16.2K engagements


"@ai_for_success nothing. just moving the goalposts. just i predicted in march"  
![@slow_developer Avatar](https://lunarcrush.com/gi/w:16/cr:twitter::1461305890892570628.png) [@slow_developer](/creator/x/slow_developer) on [X](/post/tweet/1946445470177317333) 2025-07-19 05:42:17 UTC 46.6K followers, XXX engagements


"Grok X has been amazing to me so far but the main problem is how slow Grok is at replying chatGPT gives similar answers much faster around 1-2 seconds -- while grok can take up to 10-15 seconds Elon and the xAI team please fix this. it is painfully slow"  
![@slow_developer Avatar](https://lunarcrush.com/gi/w:16/cr:twitter::1461305890892570628.png) [@slow_developer](/creator/x/slow_developer) on [X](/post/tweet/1945366537281659243) 2025-07-16 06:15:00 UTC 46.7K followers, 64.4K engagements


"openAI Noam Brown confirmed X key points: 1) the model that reached IMO gold-level performance is general-purpose not fine-tuned for math and achieved it without using tools 2) openAI developed new techniques that make LLMs a lot better could it be a new RL breakthrough or an algorithmic improvement"  
![@slow_developer Avatar](https://lunarcrush.com/gi/w:16/cr:twitter::1461305890892570628.png) [@slow_developer](/creator/x/slow_developer) on [X](/post/tweet/1946585831856095284) 2025-07-19 15:00:02 UTC 46.7K followers, 28.4K engagements


"Fortune reported: "Google DeepMind has grand ambitions to cure all diseases with AI. Now it's gearing up for its first human trials" this is the moment. generative design of drugs is my go-to example to show AI isn't just hype. this is why AGI or ASI matter to solve science and medical challenges humans can't"  
![@slow_developer Avatar](https://lunarcrush.com/gi/w:16/cr:twitter::1461305890892570628.png) [@slow_developer](/creator/x/slow_developer) on [X](/post/tweet/1941965378906693946) 2025-07-06 21:00:00 UTC 46.7K followers, 193.3K engagements


"Eric Schmidt says we're on the verge of AI replacing most programming and math tasks these domains are scale-free and don't require real-world data just compute "world-class AI mathematicians will emerge in the next year and top-tier AI programmers within two""  
![@slow_developer Avatar](https://lunarcrush.com/gi/w:16/cr:twitter::1461305890892570628.png) [@slow_developer](/creator/x/slow_developer) on [X](/post/tweet/1946007007493804055) 2025-07-18 00:40:00 UTC 46.7K followers, 82.6K engagements


"Google Logan Paul says model progress now comes from scaling all three stages: "pre-training post-training and reasoning with RL" Gemini XXX succeeded by pushing all three in parallel. post-training gave good results but reasoning is now easier to scale that's why companies are investing in all three because they all work"  
![@slow_developer Avatar](https://lunarcrush.com/gi/w:16/cr:twitter::1461305890892570628.png) [@slow_developer](/creator/x/slow_developer) on [X](/post/tweet/1947686856809701503) 2025-07-22 15:55:07 UTC 46.7K followers, XXX engagements


"Mark Cuban says learning AI today feels like the early internet days something new comes out every day keep asking questions try everything and stay engaged if you're hesitant about AI the answer is blunt: "you're f*cked. it's do or die""  
![@slow_developer Avatar](https://lunarcrush.com/gi/w:16/cr:twitter::1461305890892570628.png) [@slow_developer](/creator/x/slow_developer) on [X](/post/tweet/1946743108122583077) 2025-07-20 01:25:00 UTC 46.7K followers, 10.7K engagements


"what would "slowing down" actually look like should labs stop exploring just because the public isn't ready yet and is web-3 even part of this race"  
![@slow_developer Avatar](https://lunarcrush.com/gi/w:16/cr:twitter::1461305890892570628.png) [@slow_developer](/creator/x/slow_developer) on [X](/post/tweet/1947523966341451885) 2025-07-22 05:07:51 UTC 46.7K followers, XXX engagements


"gpt-4.1 is underrated as a coding model sonnet X gemini XXX pro o3 and grok X work well even with vague prompts gpt-4.1 needs clear detailed instructions but performs great when used right it's nearly free in cursor and copilot with good prompts gpt-4.1 comes close to the top models"  
![@slow_developer Avatar](https://lunarcrush.com/gi/w:16/cr:twitter::1461305890892570628.png) [@slow_developer](/creator/x/slow_developer) on [X](/post/tweet/1946857626522046800) 2025-07-20 09:00:03 UTC 46.7K followers, 45.5K engagements


"OpenAI CEO Sam Altman: I'm very interested in what it would mean to give everyone on Earth a free copy of GPT-5 running continuously for them Some economies will transform very quickly and run everything on AI at 1/100th the cost"  
![@slow_developer Avatar](https://lunarcrush.com/gi/w:16/cr:twitter::1461305890892570628.png) [@slow_developer](/creator/x/slow_developer) on [X](/post/tweet/1947772390571688044) 2025-07-22 21:35:00 UTC 46.7K followers, 76.1K engagements


"Zuck seems desperate: - meta $300m/4yr offer still isn't enough to poach some openAI top researchers - zuck even pitched Mark Chen (oAI CRO) with a $1B offer woah. that shows panic mode and the level of resistance suggests openAI is working on something big"  
![@slow_developer Avatar](https://lunarcrush.com/gi/w:16/cr:twitter::1461305890892570628.png) [@slow_developer](/creator/x/slow_developer) on [X](/post/tweet/1947207411019677815) 2025-07-21 08:09:58 UTC 46.7K followers, 267.3K engagements


"Roman Yampolskiy says were building something thousands of times smarter than us -- and we wont predict its moves Superintelligence wont stop at version XXX. It will keep improving itself indefinitely. The only hope A safety system that scales forever and never slips"  
![@slow_developer Avatar](https://lunarcrush.com/gi/w:16/cr:twitter::1461305890892570628.png) [@slow_developer](/creator/x/slow_developer) on [X](/post/tweet/1941726300957442363) 2025-07-06 05:10:00 UTC 46.7K followers, 28.2K engagements


"we were promised AGI but ended up with a virtual girlfriend instead 💀"  
![@slow_developer Avatar](https://lunarcrush.com/gi/w:16/cr:twitter::1461305890892570628.png) [@slow_developer](/creator/x/slow_developer) on [X](/post/tweet/1944950043183861782) 2025-07-15 02:40:00 UTC 46.6K followers, 5180 engagements


"here we go. first leak of GPT5 "gpt5reasoningalpha" found in a biologybenchmarksinspect repository important point: it was commit XX hours ago replacing the o3 model with the new gpt5 reasoningalpha entry if this repo is legit gpt5 isn't far off now"  
![@slow_developer Avatar](https://lunarcrush.com/gi/w:16/cr:twitter::1461305890892570628.png) [@slow_developer](/creator/x/slow_developer) on [X](/post/tweet/1946545812332540130) 2025-07-19 12:21:01 UTC 46.7K followers, 76.3K engagements


"New to web development Check out this comprehensive roadmap of HTML and CSS for beginners with project ideas:"  
![@slow_developer Avatar](https://lunarcrush.com/gi/w:16/cr:twitter::1461305890892570628.png) [@slow_developer](/creator/x/slow_developer) on [X](/post/tweet/1650469708041187332) 2023-04-24 12:00:17 UTC 46.7K followers, 95.6K engagements


"AI progress is speeding up at all major labs: xAI openAI Google China no single lab will dominate because new methods will quickly be copied or reverse-engineered there is no wall. and it's only a matter of time until intelligence is fully solved"  
![@slow_developer Avatar](https://lunarcrush.com/gi/w:16/cr:twitter::1461305890892570628.png) [@slow_developer](/creator/x/slow_developer) on [X](/post/tweet/1947446493650055403) 2025-07-22 00:00:00 UTC 46.7K followers, 21.7K engagements


"Meta CEO Mark Zuckerberg: "within 12-18 months most of the code is written by AI" It wont just be autocomplete. AI agents will set goals run tests find problems and write better code than top engineers"  
![@slow_developer Avatar](https://lunarcrush.com/gi/w:16/cr:twitter::1461305890892570628.png) [@slow_developer](/creator/x/slow_developer) on [X](/post/tweet/1917264443916050828) 2025-04-29 17:07:19 UTC 46.7K followers, 3.7M engagements


"Jack Dorsey says a future ruled by a few companies would be a nightmare DeepSeek shows open competitive models can outperform corporate ones "we can't depend on one system to keep us safe" give people access let them innovate stay ahead and build safer systems"  
![@slow_developer Avatar](https://lunarcrush.com/gi/w:16/cr:twitter::1461305890892570628.png) [@slow_developer](/creator/x/slow_developer) on [X](/post/tweet/1947329471721390430) 2025-07-21 16:15:00 UTC 46.7K followers, 3111 engagements


"Mark Zuckerberg says Meta's Superintelligence Labs offer researchers unmatched leverage the best talent doesn't want big teams they want massive compute "fewest people most GPUs" having the most compute per researcher is a strategic advantage"  
![@slow_developer Avatar](https://lunarcrush.com/gi/w:16/cr:twitter::1461305890892570628.png) [@slow_developer](/creator/x/slow_developer) on [X](/post/tweet/1945009814808953144) 2025-07-15 06:37:31 UTC 46.6K followers, 10.9K engagements


"damn. Grok XXX or XXX coming xAI is already working on v7 of their model grok X is based on v6 elon teased that v7 will bring major improvements in how it understands images and videos and it will also include a video generation model grok will test its games fix bugs add features and keep improving them until they're great"  
![@slow_developer Avatar](https://lunarcrush.com/gi/w:16/cr:twitter::1461305890892570628.png) [@slow_developer](/creator/x/slow_developer) on [X](/post/tweet/1944713490607775803) 2025-07-14 11:00:01 UTC 46.7K followers, 85.1K engagements


"CFO Sarah Friar revealed that OpenAI is working on: "Agentic Software Engineer (A-SWE)" unlike current tools like Copilot which only boost developers. A-SWE can build apps handle pull requests conduct QA fix bugs and write documentation"  
![@slow_developer Avatar](https://lunarcrush.com/gi/w:16/cr:twitter::1461305890892570628.png) [@slow_developer](/creator/x/slow_developer) on [X](/post/tweet/1911055984249667641) 2025-04-12 13:57:07 UTC 46.7K followers, 801.6K engagements


"@elonmusk i think invention has a lot of dependencies: tools freedom context e.g; if we locked Einstein in a room with no paper he wouldn't invent much either so maybe it's more about what the model is allowed to do than what it can do"  
![@slow_developer Avatar](https://lunarcrush.com/gi/w:16/cr:twitter::1461305890892570628.png) [@slow_developer](/creator/x/slow_developer) on [X](/post/tweet/1946799777246843313) 2025-07-20 05:10:11 UTC 46.7K followers, 2180 engagements

[GUEST ACCESS MODE: Data is scrambled or limited to provide examples. Make requests using your API key to unlock full data. Check https://lunarcrush.ai/auth for authentication information.]

@slow_developer "and we've just got a SORA X leak today last week sam and some openAI employees hinted that gpt-5 is releasing soon maybe they replace the older Sora with SORA X and integrate it into GPT-5 but openAI won't release it unless it's as good as or better than Veo 3"
@slow_developer Avatar @slow_developer on X 2025-07-21 13:39:06 UTC 46.7K followers, 10.7K engagements

"Ilya Sutskever says the real challenge of AI is how fast and extreme its growth will be it still has limits but it already speaks codes and interacts in ways once impossible soon perhaps within X to XX years AI will be able to do everything humans can do the reason: "the human brain is a biological computer so why can't a digital brain do the same""
@slow_developer Avatar @slow_developer on X 2025-06-09 08:45:00 UTC 46.7K followers, 146.3K engagements

"this should be a bigger deal than winning IMO openAI researcher said: "their IMO model "knew"; it didn't have the right answer" does he mean hallucination-free if models know what they know and when to ask for human help we could finally trust them with real-world tasks"
@slow_developer Avatar @slow_developer on X 2025-07-22 13:05:00 UTC 46.7K followers, 10.8K engagements

"@scaling01 10m or unlimited context windows mean nothing unless you solve MEMORY that's the real bottleneck"
@slow_developer Avatar @slow_developer on X 2025-07-19 12:42:34 UTC 46.6K followers, 1192 engagements

"chatGPT has outperformed nearly all elite programmers (only one survived) in a competition focusing on optimization algorithms openAI model entered the finals on July XX led early but was overtaken by Psyho a former oAI member key takeaways: - openAI internal coding model is insane - this could be the last human win"
@slow_developer Avatar @slow_developer on X 2025-07-17 11:05:00 UTC 46.6K followers, 43.6K engagements

"i'm confused by this. how advanced AI whether it's AGI or ASI could create new jobs we already know it will automate many existing ones but if AI becomes capable of doing everything humans can do what would be left for people to do"
@slow_developer Avatar @slow_developer on X 2025-07-14 18:45:00 UTC 46.7K followers, 64.7K engagements

"@burkov @TheLudd AGI was achieved long ago. it's meaningless now the term is used to build the hype for funding"
@slow_developer Avatar @slow_developer on X 2025-07-22 10:13:11 UTC 46.7K followers, XXX engagements

"🚨 Meta CEO Mark Zuckerberg predicts smart glasses will replace phones by 2030 " Smart glasses are going to become the next major computing platform. They will gradually replace phones by 2030 much like mobile devices surpassed computers without fully replacing them ""
@slow_developer Avatar @slow_developer on X 2024-09-29 17:21:29 UTC 46.7K followers, 43.4M engagements

"AI agents arent good enough yet at least in coding. models like gemini XXX pro and o3/o3 pro are great for debugging .but for agent tasks even claude code often goes in circles or gets stuck wasting api calls the best reasoning models are impressive -- but using them constantly for agents is still too expensive"
@slow_developer Avatar @slow_developer on X 2025-06-29 21:30:00 UTC 46.6K followers, 8602 engagements

""in the future Agents should act for you without being asked" OpenAI Isa Fulford says the next big focus is improving how we collaborate with AI agents Multi-turn chats work well but the goal is smoother more continuous interactions personalization and memory will be key"
@slow_developer Avatar @slow_developer on X 2025-07-22 16:35:18 UTC 46.7K followers, 7880 engagements

"after ChatGPT Agent and Grok X heavy the focus shifted to new benchmarks HLE and Frontier Math are still important but agentic benchmarks now matter more for most ppl intelligence is wide enough what's needed now is agent ability: tools memory context and modalities"
@slow_developer Avatar @slow_developer on X 2025-07-20 14:14:01 UTC 46.7K followers, 6734 engagements

"subscription pricing mess. that's why we need open-source models to keep pressure on the big AI labs even if DeepSeek Qwen and KIMI might not be the best but without open-source competition gemini and openAI wouldn't be offering as much for free"
@slow_developer Avatar @slow_developer on X 2025-07-12 12:00:01 UTC 46.7K followers, 71.9K engagements

"total domination by openAI this week: - launched chatGPT Agent; matched or beat humans in half the tasks even 10+ hours - mysterious o3-alpha claimed to be the best coding model - gpt-5 reasoning alpha leak - internal experimental model achieved IMO gold-level performance"
@slow_developer Avatar @slow_developer on X 2025-07-19 20:10:00 UTC 46.7K followers, 12.3K engagements

"@jpreagan_ what if AGI is more about broad cognitive capability than pure economic utility the models today aren't perfect but their generality feels. different now"
@slow_developer Avatar @slow_developer on X 2025-07-20 05:22:43 UTC 46.6K followers, XXX engagements

"MAJOR BLOW TO OPENAI Meta is rumored to have poached two more high-profile openAI researchers: Jason Wei co-author of the scaling laws paper and a lead researcher on agentic models and reasoning Hyung Won Chung head of Codex and a core architect behind GPT-4 the 'o' series and deep research brutal hit. sam has now lost several of his closest allies and top researchers"
@slow_developer Avatar @slow_developer on X 2025-07-16 08:50:00 UTC 46.7K followers, 11.1K engagements

"more info about GPT-5 the release is imminent -- possibly within the next two weeks depending on safety tests "it's a router system that switches b/w reasoning basic and tool-use models; GPT-6 is in training" this is confusing because openAI earlier said a single unified model"
@slow_developer Avatar @slow_developer on X 2025-07-20 11:35:00 UTC 46.7K followers, 23.5K engagements

"this should be bigger news. "ChatGPT Agent matched or beat humans in about half the tasks tested even ones that take over XX hours" you kidding me we're only a little over halfway into the year of AI agents and they're already completing high-impact tasks"
@slow_developer Avatar @slow_developer on X 2025-07-18 16:35:00 UTC 46.7K followers, 29.9K engagements

"Anthropic co-founder Ben Mann: There's about a XX% chance we'll reach superintelligence within a few years it's not just speculation; it's based on data scaling laws and clear scientific trends even if we get there soon the societal effects may take time to unfold"
@slow_developer Avatar @slow_developer on X 2025-07-20 16:55:00 UTC 46.7K followers, 22.4K engagements

"the AI community i know is slowly turning into a cult around certain AI labs without truly understanding how things work. the race to AGI isn't about picking your favorite lab support labs doing valuable work and use whatever model helps you most"
@slow_developer Avatar @slow_developer on X 2025-07-18 19:10:00 UTC 46.6K followers, 6379 engagements

"where' the damn "ChatGPT Agent" for plus users"
@slow_developer Avatar @slow_developer on X 2025-07-22 12:07:01 UTC 46.7K followers, 21.3K engagements

"Ex-DeepMind researcher Misha Laskin says labs quietly changed the meaning of AGI to superintelligence The goalposts are shifting. what counted as AGI before is now called superintelligence but the definition is still vague unless we solve safety we're playing an Indiana Jones game that ends badly"
@slow_developer Avatar @slow_developer on X 2025-07-19 00:20:00 UTC 46.7K followers, 24.3K engagements

"MORE DRAMA daniel gross wanted ilya to sell the company to Meta crazy since he was SSI ceo at the time. ilya is a GOAT researcher but something feels off about SSI if they're close to a breakthrough why did their CEO leave sooner or later ilya has to step out and show what's going on"
@slow_developer Avatar @slow_developer on X 2025-07-21 18:50:00 UTC 46.7K followers, 7137 engagements

"Elon AGI bet: scale fast scale hard he announced Colossus X will host over 550k gb200s and gb300s in a few weeks that's ridiculously massive. that means xAI will soon be training on more compute than any other AI lab by a huge margin colossus X already runs 230k GPUs (including 30k gb200s) for grok training colossus X takes that to a whole new level"
@slow_developer Avatar @slow_developer on X 2025-07-23 05:30:00 UTC 46.7K followers, 2129 engagements

"why keep moving the AGI bar 10-15 years ago the AI we have today would've been called AGI frontier models now outperform most people in many tasks and understand things that seemed impossible in 2000 it's not that we've redefined intelligence -- we've just kept raising the bar"
@slow_developer Avatar @slow_developer on X 2025-07-18 03:15:00 UTC 46.6K followers, 7772 engagements

"have you noticed every major lab is now talking about superintelligence -- because it's clear we're already crossing the AGI threshold the term 'AGI' feels meaningless now if you'd shown o3 gemini XXX or grok X to someone in 2017 theyd think we were well past it"
@slow_developer Avatar @slow_developer on X 2025-07-14 03:00:03 UTC 46.7K followers, 682.8K engagements

"bullish on those who think openAI is done just because a model was delayed chatGPT is the 3rd most visited website globally ahead of X no X on the Apple App Store in the U.S unless someone delivers something significantly better and cheaper oAI stays on top no need to hit the panic button"
@slow_developer Avatar @slow_developer on X 2025-07-14 05:50:08 UTC 46.6K followers, 11.1K engagements

"ARC-AGI X is already here we haven't even completed half of ARC-AGI X and now there's ARC-3 and wasn't the test meant to tell us when we've reached AGI now the models are getting close they keep making new tests and shifting the goalposts Turing test passed ARC-AGI X passed still not AGI"
@slow_developer Avatar @slow_developer on X 2025-07-18 21:45:00 UTC 46.7K followers, 8741 engagements

"Grok X ranked 2nd with a XXXX% score on the updated SimpleBench it's a great model and truly SOTA in areas like: math science reasoning also this isn't even grok X heavy that one should be even better xAI needs to launch a grok-4-heavy API and it must be affordable"
@slow_developer Avatar @slow_developer on X 2025-07-20 06:30:00 UTC 46.7K followers, 4626 engagements

"imagine the scenario: this is the moment when "poached" openAI researchers were quietly automated by a GPT-5 agent not officially fired -- just quietly automated but what are the odds of that happening"
@slow_developer Avatar @slow_developer on X 2025-07-22 05:20:00 UTC 46.7K followers, 5710 engagements

"Amjad Masad says true AGI means an AI that can enter new situations and quickly learn how to achieve goals Even the latest models still struggle with tasks outside their training data. The real breakthrough will come when AI can train itself -- like AlphaGo playing billions of games against itself"
@slow_developer Avatar @slow_developer on X 2025-07-03 19:35:00 UTC 46.7K followers, 161.4K engagements

"UBE - universal basic employment makes more sense than UBI job shortages are mainly a private sector issue so as long as humans can still add value its better to have them work UBI won't happen until almost all jobs are gone which will take a very long time"
@slow_developer Avatar @slow_developer on X 2025-07-16 00:50:00 UTC 46.7K followers, 13.2K engagements

"the best bet for GPT-5 release is august it was recently spotted in a biotech repo as "gpt-5 reasoning" and sam said it's releasing soon o3-alpha (maybe their open-weight) is claimed to be SOTA coding model openAI just launched "ChatGPT Agent" so no need to drop gpt-5 yet so august makes the most sense"
@slow_developer Avatar @slow_developer on X 2025-07-22 19:05:00 UTC 46.7K followers, 3939 engagements

"Mysterious Model Alert a new model "o3 Alpha" by openAI claims to be better than o3-pro and google Kingfall but what exactly is it full version of Codex delayed open-weight model fine-tuned o3 version for chatGPT Agent probably not GPT-5 since they won't share it before launch"
@slow_developer Avatar @slow_developer on X 2025-07-19 03:00:03 UTC 46.6K followers, 9921 engagements

"@vasalex93 even humans screw up stuff like that too miscount overlook get tired forget context do we call them non-intelligent"
@slow_developer Avatar @slow_developer on X 2025-07-19 19:43:53 UTC 46.6K followers, XXX engagements

"@polynoamial @OpenAI which model exactly Noam"
@slow_developer Avatar @slow_developer on X 2025-07-19 07:56:54 UTC 46.7K followers, 13.2K engagements

"@SeSphoenix or maybe we're all expecting AGI to look and act like us with this seamless inner monologue .but what if general intelligence ends up working in a way thats just super different from how humans think"
@slow_developer Avatar @slow_developer on X 2025-07-20 05:29:33 UTC 46.6K followers, XXX engagements

"i forgot we're only allowed to talk about things once they're "officially" solved by your standards"
@slow_developer Avatar @slow_developer on X 2025-07-20 05:17:03 UTC 46.6K followers, XXX engagements

"full detail thread: zuck offers $1B to openAI CRO"
@slow_developer Avatar @slow_developer on X 2025-07-21 08:21:06 UTC 46.7K followers, 15.3K engagements

"Demis Hassabis says true AGI is a theoretical benchmark set by the human brain the only proven architecture for general intelligence todays models are impressive but inconsistent; anyone can find flaws within minutes. "real AGI should be so strong that it would take experts months to spot a weakness""
@slow_developer Avatar @slow_developer on X 2025-05-21 07:07:53 UTC 46.7K followers, 458.9K engagements

"video source: and if you don't want to miss major AI news you can always subscribe to the newsletter it includes: - top X% industry news - important research papers with summaries - read by 250k+ AI developers"
@slow_developer Avatar @slow_developer on X 2025-07-20 19:02:00 UTC 46.7K followers, 1264 engagements

"Mustafa Suleyman says the possibility of superintelligence arriving should keep everyone awake we've never controlled anything as powerful as us let alone something far smarter if we get it right we'll solve our energy health and food crises "but getting it right is mind-blowingly hard""
@slow_developer Avatar @slow_developer on X 2025-07-13 19:10:00 UTC 46.7K followers, 31.1K engagements

"elon was right. i pasted my entire python codebase into Grok X -- a full data pipeline with CSV parsing and API calls grok X quickly found poor function structure weak error handling and messy logic then suggested solid fixes most impressive it understood the whole pipeline and helped the modules work better together if you maintain complex or old python code grok X can save you a lot of time"
@slow_developer Avatar @slow_developer on X 2025-07-11 10:03:36 UTC 46.7K followers, 1.1M engagements

"we need to be honest with 18-year-olds for most university is pointless. you take on a lot of debt learn very little and struggle to find a job afterward AI has already reduced the value of higher education for about XX% of students with an IQ below XXX the old path is broken"
@slow_developer Avatar @slow_developer on X 2025-07-17 03:10:00 UTC 46.7K followers, 32.5K engagements

"BIG NEWS openAI experimental reasoning model reached IMO gold-level performance and they are also releasing GPT-5 soon new Q-star moment looks like they're using new reasoning methods that go beyond standard CoT techniques it also confirms they're running internal models months ahead of anything public"
@slow_developer Avatar @slow_developer on X 2025-07-19 09:37:36 UTC 46.7K followers, 10.9K engagements

"BIG NEWS "Google announced an advanced version of gemini DeepThink won gold at the IMO" the model was trained on past IMOs with RL and multi-step reasoning surely google will likely be releasing advanced models at a faster pace now the world is not yet ready for the GPT-5 vs Gemini 3"
@slow_developer Avatar @slow_developer on X 2025-07-21 21:25:00 UTC 46.7K followers, 7778 engagements

"important point to note. openAI researcher Jerry Tworek says "GPT-5 will not include the breakthrough of the IMO-winning model it's a later model probably end-of-the-year thing" that means GPT-5 will be released soon so there's no time to integrate that breakthrough into it"
@slow_developer Avatar @slow_developer on X 2025-07-19 22:50:00 UTC 46.7K followers, 7133 engagements

"humanity is cooked. 💀 a person is addicted to Ani subscribed with money and reached level X he adds "it's too late for me to come back""
@slow_developer Avatar @slow_developer on X 2025-07-19 05:45:00 UTC 46.6K followers, 3252 engagements

"strongly disagree. these systems walks like intelligence talks like intelligence beats humans at reasoning tasks but it's not "real" intelligence because we don't fully understand how it works that's kind of moving the goalpost. we don't understand our own brains in great detail either yet no one questions if we're intelligent"
@slow_developer Avatar @slow_developer on X 2025-07-22 07:01:15 UTC 46.7K followers, XXX engagements

"Eric Schmidt asks what happens when AI creates millions of polymaths History's greatest discoveries came from experts applying ideas from one field to another something AI can't do yet "the rules keep changing" solving non-stationarity might bring breakthroughs but it'll take years to know"
@slow_developer Avatar @slow_developer on X 2025-07-18 14:00:02 UTC 46.7K followers, 22.4K engagements

"what could be the next openAI open-weight model right now weve got: o3 o3-pro o4-mini o4-mini-high and the rumors say its a reasoning model sam has hinted many times even asked in polls about the idea of o3-mini running locally on phones so its either o3-mini or o4-mini being released as open weights"
@slow_developer Avatar @slow_developer on X 2025-06-30 13:54:39 UTC 46.6K followers, 3674 engagements

"openAI researcher says "GPT-5 is coming" the key point: it will be different from the current reasoning setup that relies on RL and tool use that could mean: - new architecture or full end-to-end RL - stronger tool use or closer to a world model - can think longer or scale deeper based on the question"
@slow_developer Avatar @slow_developer on X 2025-07-18 11:27:48 UTC 46.7K followers, 24.8K engagements

"google is terrible at marketing openAI explains things and educates the public they had a breakthrough: RL with unverified rewards and they're explaining what that means google on the other hand drops something like this. THIS they're both smart and dumb at once"
@slow_developer Avatar @slow_developer on X 2025-07-22 07:57:50 UTC 46.7K followers, 4175 engagements

"i'm convinced we've already crossed the AGI threshold the term "AGI" has lost its meaning every time we get close the goalposts shift maybe only a superintelligence can look back and define what AGI truly was otherwise we'll be stuck debating it forever"
@slow_developer Avatar @slow_developer on X 2025-07-19 17:35:00 UTC 46.7K followers, 47K engagements

"Grok X feels much more human than other AIs after using gemini XXX pro o3 and others Grok felt more natural and less robotic DeepSearch is highly powerful and unbiased no point in using google search anymore these are the things that matter most to non-technical users"
@slow_developer Avatar @slow_developer on X 2025-07-21 03:10:00 UTC 46.7K followers, 144.4K engagements

"all eyes are on GPT-5 the model wont just be a picker; it should be a full system that integrates everything more intelligently id guess DeepMind is also heading the same way with its upcoming gemini models with improved memory and longer context it could feel like talking to your AGI"
@slow_developer Avatar @slow_developer on X 2025-07-05 23:45:00 UTC 46.7K followers, 12.7K engagements

"Logan was right. only two weeks have passed and we've already seen major AI milestones: - models winning gold at the IMO - launch of the ChatGPT Agent - news about openAI unreleased top-tier coding model - two SOTA open-source models released: Qwen and Kimi k2 everything is scaling up fast"
@slow_developer Avatar @slow_developer on X 2025-07-22 10:33:41 UTC 46.7K followers, 15.6K engagements

"this is so frustrating. i get that gpt-4.5 uses a lot of GPU but what is a regular plus user supposed to do with only XX prompts a week sure the model is impressive at emotional intelligence still what's the point of releasing it if people can barely use it"
@slow_developer Avatar @slow_developer on X 2025-07-23 02:45:00 UTC 46.7K followers, 10.9K engagements

"Sam Altman says some jobs will disappear many will evolve and entirely new ones will emerge. OpenAI's growth means it's making people more efficient not replacing them "we're growing so fast that theres still so much more to do" As for hardware it's going to be great but no smart glasses"
@slow_developer Avatar @slow_developer on X 2025-07-09 07:45:00 UTC 46.7K followers, 79.4K engagements

"Geoffrey Hinton says LLMs aren't just mimicking they're making creative connections When asked why a compost heap is like an atom bomb GPT-4 pointed to chain reactions which showed it understood abstract links "it saw the analogy during training not test time" so they're seeing analogies to compress knowledge like humans do"
@slow_developer Avatar @slow_developer on X 2025-07-17 00:30:00 UTC 46.7K followers, 46.2K engagements

"sam altman: perfect AI = tiny model superhuman reasoning 1T context tokens i agree with him. DO YOU the X important things missing for AGI now are large context size and strong reasoning current models like gemini have 1m tokens but after 120- 150k tokens they lose more than half the information a billion-token context with full retrieval could reach AGI even at XXX pro or o3 reasoning levels"
@slow_developer Avatar @slow_developer on X 2025-06-04 21:35:00 UTC 46.6K followers, 16.2K engagements

"@ai_for_success nothing. just moving the goalposts. just i predicted in march"
@slow_developer Avatar @slow_developer on X 2025-07-19 05:42:17 UTC 46.6K followers, XXX engagements

"Grok X has been amazing to me so far but the main problem is how slow Grok is at replying chatGPT gives similar answers much faster around 1-2 seconds -- while grok can take up to 10-15 seconds Elon and the xAI team please fix this. it is painfully slow"
@slow_developer Avatar @slow_developer on X 2025-07-16 06:15:00 UTC 46.7K followers, 64.4K engagements

"openAI Noam Brown confirmed X key points: 1) the model that reached IMO gold-level performance is general-purpose not fine-tuned for math and achieved it without using tools 2) openAI developed new techniques that make LLMs a lot better could it be a new RL breakthrough or an algorithmic improvement"
@slow_developer Avatar @slow_developer on X 2025-07-19 15:00:02 UTC 46.7K followers, 28.4K engagements

"Fortune reported: "Google DeepMind has grand ambitions to cure all diseases with AI. Now it's gearing up for its first human trials" this is the moment. generative design of drugs is my go-to example to show AI isn't just hype. this is why AGI or ASI matter to solve science and medical challenges humans can't"
@slow_developer Avatar @slow_developer on X 2025-07-06 21:00:00 UTC 46.7K followers, 193.3K engagements

"Eric Schmidt says we're on the verge of AI replacing most programming and math tasks these domains are scale-free and don't require real-world data just compute "world-class AI mathematicians will emerge in the next year and top-tier AI programmers within two""
@slow_developer Avatar @slow_developer on X 2025-07-18 00:40:00 UTC 46.7K followers, 82.6K engagements

"Google Logan Paul says model progress now comes from scaling all three stages: "pre-training post-training and reasoning with RL" Gemini XXX succeeded by pushing all three in parallel. post-training gave good results but reasoning is now easier to scale that's why companies are investing in all three because they all work"
@slow_developer Avatar @slow_developer on X 2025-07-22 15:55:07 UTC 46.7K followers, XXX engagements

"Mark Cuban says learning AI today feels like the early internet days something new comes out every day keep asking questions try everything and stay engaged if you're hesitant about AI the answer is blunt: "you're f*cked. it's do or die""
@slow_developer Avatar @slow_developer on X 2025-07-20 01:25:00 UTC 46.7K followers, 10.7K engagements

"what would "slowing down" actually look like should labs stop exploring just because the public isn't ready yet and is web-3 even part of this race"
@slow_developer Avatar @slow_developer on X 2025-07-22 05:07:51 UTC 46.7K followers, XXX engagements

"gpt-4.1 is underrated as a coding model sonnet X gemini XXX pro o3 and grok X work well even with vague prompts gpt-4.1 needs clear detailed instructions but performs great when used right it's nearly free in cursor and copilot with good prompts gpt-4.1 comes close to the top models"
@slow_developer Avatar @slow_developer on X 2025-07-20 09:00:03 UTC 46.7K followers, 45.5K engagements

"OpenAI CEO Sam Altman: I'm very interested in what it would mean to give everyone on Earth a free copy of GPT-5 running continuously for them Some economies will transform very quickly and run everything on AI at 1/100th the cost"
@slow_developer Avatar @slow_developer on X 2025-07-22 21:35:00 UTC 46.7K followers, 76.1K engagements

"Zuck seems desperate: - meta $300m/4yr offer still isn't enough to poach some openAI top researchers - zuck even pitched Mark Chen (oAI CRO) with a $1B offer woah. that shows panic mode and the level of resistance suggests openAI is working on something big"
@slow_developer Avatar @slow_developer on X 2025-07-21 08:09:58 UTC 46.7K followers, 267.3K engagements

"Roman Yampolskiy says were building something thousands of times smarter than us -- and we wont predict its moves Superintelligence wont stop at version XXX. It will keep improving itself indefinitely. The only hope A safety system that scales forever and never slips"
@slow_developer Avatar @slow_developer on X 2025-07-06 05:10:00 UTC 46.7K followers, 28.2K engagements

"we were promised AGI but ended up with a virtual girlfriend instead 💀"
@slow_developer Avatar @slow_developer on X 2025-07-15 02:40:00 UTC 46.6K followers, 5180 engagements

"here we go. first leak of GPT5 "gpt5reasoningalpha" found in a biologybenchmarksinspect repository important point: it was commit XX hours ago replacing the o3 model with the new gpt5 reasoningalpha entry if this repo is legit gpt5 isn't far off now"
@slow_developer Avatar @slow_developer on X 2025-07-19 12:21:01 UTC 46.7K followers, 76.3K engagements

"New to web development Check out this comprehensive roadmap of HTML and CSS for beginners with project ideas:"
@slow_developer Avatar @slow_developer on X 2023-04-24 12:00:17 UTC 46.7K followers, 95.6K engagements

"AI progress is speeding up at all major labs: xAI openAI Google China no single lab will dominate because new methods will quickly be copied or reverse-engineered there is no wall. and it's only a matter of time until intelligence is fully solved"
@slow_developer Avatar @slow_developer on X 2025-07-22 00:00:00 UTC 46.7K followers, 21.7K engagements

"Meta CEO Mark Zuckerberg: "within 12-18 months most of the code is written by AI" It wont just be autocomplete. AI agents will set goals run tests find problems and write better code than top engineers"
@slow_developer Avatar @slow_developer on X 2025-04-29 17:07:19 UTC 46.7K followers, 3.7M engagements

"Jack Dorsey says a future ruled by a few companies would be a nightmare DeepSeek shows open competitive models can outperform corporate ones "we can't depend on one system to keep us safe" give people access let them innovate stay ahead and build safer systems"
@slow_developer Avatar @slow_developer on X 2025-07-21 16:15:00 UTC 46.7K followers, 3111 engagements

"Mark Zuckerberg says Meta's Superintelligence Labs offer researchers unmatched leverage the best talent doesn't want big teams they want massive compute "fewest people most GPUs" having the most compute per researcher is a strategic advantage"
@slow_developer Avatar @slow_developer on X 2025-07-15 06:37:31 UTC 46.6K followers, 10.9K engagements

"damn. Grok XXX or XXX coming xAI is already working on v7 of their model grok X is based on v6 elon teased that v7 will bring major improvements in how it understands images and videos and it will also include a video generation model grok will test its games fix bugs add features and keep improving them until they're great"
@slow_developer Avatar @slow_developer on X 2025-07-14 11:00:01 UTC 46.7K followers, 85.1K engagements

"CFO Sarah Friar revealed that OpenAI is working on: "Agentic Software Engineer (A-SWE)" unlike current tools like Copilot which only boost developers. A-SWE can build apps handle pull requests conduct QA fix bugs and write documentation"
@slow_developer Avatar @slow_developer on X 2025-04-12 13:57:07 UTC 46.7K followers, 801.6K engagements

"@elonmusk i think invention has a lot of dependencies: tools freedom context e.g; if we locked Einstein in a room with no paper he wouldn't invent much either so maybe it's more about what the model is allowed to do than what it can do"
@slow_developer Avatar @slow_developer on X 2025-07-20 05:10:11 UTC 46.7K followers, 2180 engagements

creator/twitter::1461305890892570628/posts
/creator/twitter::1461305890892570628/posts