[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.]

@IntuitMachine Avatar @IntuitMachine Carlos E. Perez

Carlos E. Perez posts on X about ai, theory, $googl, agi the most. They currently have XXXXXX followers and XXX posts still getting attention that total XXXXXX engagements in the last XX hours.

Engagements: XXXXXX #

Engagements Line Chart

Mentions: XX #

Mentions Line Chart

Followers: XXXXXX #

Followers Line Chart

CreatorRank: XXXXXXX #

CreatorRank Line Chart

Social Influence

Social category influence technology brands XXXX% finance XXXX% stocks XXXX% countries XXXX% vc firms XXXX% musicians XXXX% celebrities XXXX% automotive brands XXXX%

Social topic influence ai #2507, theory #1503, $googl #2345, agi #9, future 1.81%, open ai #493, llm #106, secret 1.36%, prediction #1651, guess XXX%

Top accounts mentioned or mentioned by @grok @elonmusk @chaos2cured @alphapromega @katsuerte @dwarkeshsp @thechaosledger @viktorobert @ilyasut @yanngabioud @danshipper @jesiicam @molsonhart @benitoz @bullmarketbill1 @sensenopedout @viumobile @ivan_bezdomny @dysmemic @warrenmansell

Top assets mentioned Alphabet Inc Class A (GOOGL) NVIDIA Corp. (NVDA)

Top Social Posts

Top posts by engagements in the last XX hours

"I just read a paper that completely broke my brain. It describes a system that solved an AI task with over 1000000 sequential steps. with ZERO errors. Using AI models that are known to be flaky and make mistakes. How is that even possible 🤯 We all know LLMs have an error rate. Even XXXX% accuracy is a death sentence for long tasks. Imagine you need 1000 correct steps in a row. With a XXXX% success rate per step your chance of finishing the whole thing is only 36%. At a million steps Forget it. It's statistically impossible. So for years the race has been to build bigger "smarter" models to"
X Link 2025-11-17T11:03Z 48.4K followers, 306.7K engagements

"We've become obsessed with the idea that the brain is a "Prediction Machine." The dominant theory in neuroscience says we're constantly simulating the future calculating probabilities to guess what happens next. A new paper argues this is a complete illusion. The reality is simpler and strangely much more powerful. Here is the argument for Perceptual Control: The "Prediction Illusion" starts with a mistake in observation. When we see someone successfully handle a chaotic environment (like catching a flyball) it looks like they predicted the future trajectory of the ball. But observing"
X Link 2025-11-20T16:00Z 48.4K followers, 82K engagements

"Ever notice how an LLM can explain a complex financial concept like Black-Scholes perfectly but then fails when you ask it to apply a simple set of rules to a list of stocks It's not a coincidence. And the reason why is a huge red flag for finance. I always thought these failures were about bad data or a knowledge gap. Like the AI just didn't "know" enough about that specific stock or metric. But a recent paper on how LLMs handle chronology suggests it's a much deeper problem. A process problem. The researchers gave models like GPT-4 a simple non-finance task: sort a shuffled list of US"
X Link 2025-12-06T10:22Z 48.4K followers, 13.5K engagements

"@Chaos2Cured I don't even know wtf you're trying to say"
X Link 2025-12-06T11:29Z 48.4K followers, 3279 engagements

"The Framework Graveyard Hard truth: LangChain is for prototypes. XX% of production agents are custom code. Here's what killed agent frameworks in the real world 🧵 1/ The split is dramatic: Survey (all stages): XX% use frameworks Interviews (deployed only): XX% use frameworks What happened between prototype and production Teams ripped out the frameworks. 2/ Why custom code wins: "We started with CrewAI. Then we needed to integrate with our auth system. And our data pipeline. And our monitoring. By month X we'd forked it so hard we just rewrote it." Flexibility convenience. 3/ The three"
X Link 2025-12-07T17:25Z 48.4K followers, 22.3K engagements

"Artificial Empathy appears to have been achieved"
X Link 2025-12-07T19:12Z 48.4K followers, 1965 engagements

"@grok can you validate this based on the literature presented above"
X Link 2025-12-07T20:30Z 48.4K followers, 1552 engagements

"Is Claude runaway dominant in coding environments I don't think any other platform has something like Claude Skills. Am I right here"
X Link 2025-12-03T10:24Z 48.1K followers, XXX engagements

"Has an unknown AI firm in China caught up with Kimi-K2 Moonshot AI has gained significant attention earning the nickname AI Tiger in China with a valuation of $XXX billion after raising $XXX billion from investors like Alibaba Tencent and Gaorong Capital. The companys name is inspired by Pink Floyds album The Dark Side of the Moon"
X Link 2025-07-12T10:25Z 48.2K followers, 4655 engagements

"Elon Musk is mentioned in Demis Hassibis' biography documentary "The Thinking Game" (now on YT). I was curious about that so I asked Grok @elonmusk The story of Elon Musk and DeepMind is one of the most interesting (and dramatic) chapters in the history of modern AI development. It involves ambition fear of existential risk a secretive fundraising race and ultimately a lost bidding war that shaped the entire AI landscape. ### 20102013: The early connection - Elon Musk had been funding AI research since the 2000s because he was already worried about AI as an existential threat to humanity. -"
X Link 2025-11-28T12:45Z 48.2K followers, 225.7K engagements

"I asked Gemini XXX to distill the main points of the SemiAnalysis analysis. Here it is in a digestible and simple explanation. After you read this you'll be backing up the truck come Monday (disclaimer: this is not financial advice do your own due dilligence) 1/ Everyone is arguing about Nvidia's B200 vs Google's TPU specs. Theyre missing the actual war. I just spent the morning breaking down the new SemiAnalysis report on TPUv7 and honestly It changes the entire math of the AI industry. Here is the strategic shift almost everyone is ignoring. 🧵 2/ The standard narrative goes like this:"
X Link 2025-11-29T12:38Z 48.1K followers, 55.1K engagements

"1/ Theres a specific pattern in strategy that looks like "genius" velocity at first but often ends in total loss of control. Im seeing it right now with xAIs new supercomputer. Its the exact same trap SoftBank fell into and it explains why "speed" is sometimes a liability. 2/ First the context: xAI pulled off something physically incredible. They strung together 100000 Nvidia H100s in just XXX days to build the "Colossus" cluster. In the hardware world that is lightspeed. They chose "Time-to-Market" over every other metric. 3/ But in Game Theory we call this "Rented Velocity." To move that"
X Link 2025-11-29T13:47Z 48.1K followers, 2614 engagements

"Do you notice what is curious about this AI debt bomb Meta and Apple have signed up to use Google's TPU infrastructure They are all in the same bunker"
X Link 2025-11-29T15:16Z 48.2K followers, 2406 engagements

"@SemiAnalysis_ 🤣 How about this"
X Link 2025-11-29T18:01Z 48.2K followers, 4845 engagements

"I was curious about how Von Neumann's ideas reconceptualized how we historically thought about puzzling concepts"
X Link 2025-11-29T19:27Z 48.1K followers, 2446 engagements

"It's pretty interesting how Von Neumann achieved his breakthrough ideas"
X Link 2025-11-29T21:02Z 48.1K followers, 1458 engagements

"It's pretty curious that Von Neumann's breakthroughs can be framed as an application of semiotic logic"
X Link 2025-11-29T21:22Z 48.1K followers, 1640 engagements

"Von Neumann had a special cognitive gift"
X Link 2025-11-30T12:09Z 48.2K followers, 1184 engagements

"Are you leveling up yet"
X Link 2025-11-30T13:12Z 48.1K followers, 11.5K engagements

"Has it ever occurred to you that the radical financialization of the USA is the paperclip maximizer that we all fear"
X Link 2025-11-30T14:09Z 48.1K followers, 1779 engagements

"@Cutooie The specific answer depends on both of your spiral dynamics levels"
X Link 2025-11-30T15:30Z 48.2K followers, 35.9K engagements

"What happens when everybody has a Von Neumann-level intelligence in their pocket"
X Link 2025-11-30T15:36Z 48.1K followers, 1617 engagements

"@Jesii_ca_M Artificial Scarcity"
X Link 2025-11-30T15:38Z 48.1K followers, 1604 engagements

"How Quaternion Process Theory is derived from Semiotics"
X Link 2025-11-30T23:30Z 48.1K followers, 1006 engagements

"The Architecture of Process. Map-making is ridiculously addictive"
X Link 2025-11-30T23:37Z 48.1K followers, 2055 engagements

"A Roadmap to AGI via Artificial Empathy"
X Link 2025-11-30T23:52Z 48.1K followers, 1412 engagements

"The Process Structure of Genuine Agency"
X Link 2025-12-01T00:09Z 48.2K followers, 1329 engagements

"Very interesting Artificial Agency formulation"
X Link 2025-12-01T01:33Z 48.1K followers, 3924 engagements

"InfoGraphics brought to you by Gemini XXX but content delivered by Opus XXX. We are in a new AI era and few recognize it"
X Link 2025-12-01T13:18Z 48.1K followers, 2611 engagements

"Wow Did OpenAI just quit the AI race Are there only X players left in the running (Claude Gemini Grok)"
X Link 2025-12-01T13:21Z 48.1K followers, 47.6K engagements

"In the future even when you die parts of you will remain living as a memento (instead of a pile of ashes)"
X Link 2025-12-01T17:39Z 48.1K followers, 6657 engagements

"A New Model for Enterprise AI"
X Link 2025-12-03T01:01Z 48.1K followers, XXX engagements

"Has anybody used M365 Copilot chat in an enterprise Why is it so terrible as compared to ChatGPT What's going on here I thought they were both powered by the same models"
X Link 2025-12-03T13:59Z 48.1K followers, 1797 engagements

"@Molson_Hart The flaw"
X Link 2025-12-04T11:06Z 48.1K followers, 2026 engagements

"@Molson_Hart The path to absurdist enlightenment"
X Link 2025-12-04T11:09Z 48.1K followers, 1512 engagements

"Banger interview of Paul Ford by @danshipper. Highly recommended"
X Link 2025-12-04T18:37Z 48.2K followers, 4418 engagements

"Google's AI offering just seems to be everywhere (this graphic doesn't even include others that I use)"
X Link 2025-12-05T09:53Z 48.1K followers, 1285 engagements

"@Rainmaker1973 What kind of a sign Iconic indexical or symbolic"
X Link 2025-12-05T15:23Z 48.1K followers, 20.9K engagements

"Should we redo our Vibe-coded projects from scratch using Opus 4.5"
X Link 2025-12-06T15:29Z 48.1K followers, 1648 engagements

"How @poetiq_ai "solved" ARC-AGI-2 (with the help of Gemini 3)"
X Link 2025-12-06T18:37Z 48.1K followers, 1235 engagements

"Wait AGI benchmark just got breached"
X Link 2025-12-06T19:30Z 48.1K followers, 3837 engagements

"Kind of cool that Jungian Psychology and QPT have a lot of parallels"
X Link 2025-12-07T12:12Z 48.1K followers, 2300 engagements

"Quaternion Process Theory and the Structure of Consciousness"
X Link 2025-12-07T14:58Z 48.1K followers, 1454 engagements

"The 10-Step Simplicity Rule 🔥 Hot take: Your AI agent is probably too smart for production. After analyzing 306+ deployed agents we found the winners follow a "10-step rule" that most researchers ignore. Here's why "dumb" beats "smart" in the real world 🧵 First the brutal stats: XX% of agent deployments FAIL But we studied the X% that succeed XX% execute XX steps before human intervention XXXX% use X steps Why so limited Because in production CONTROL CAPABILITY Research celebrates agents that: Self-plan indefinitely Explore open-ended solutions Make 100s of autonomous decisions Production"
X Link 2025-12-07T18:18Z 48.1K followers, 1738 engagements

"🧵 BREAKING: We just discovered that ChatGPT thinks you're stupid. And it's not alone. XX% of advanced AI models now rank humans as the LEAST rational players in strategic games. Here's the evidence that AI has developed a "superiority complex": 🧠 1/ Researchers tested XX AI models (GPT-4 Claude Gemini) using a classic game theory experiment. The game: Guess a number. Winner = closest to 2/3 of the average. Optimal play requires thinking about how OTHERS think. It's a pure test of recursive reasoning. 2/ Here's where it gets wild: When told they're playing against HUMANS AIs guess XX When"
X Link 2025-12-06T10:37Z 48.4K followers, 18.1K engagements

"Perhaps the first independent (and open source) implementation of my Quaternion Process Theory ideas"
X Link 2025-12-11T07:20Z 48.4K followers, 1101 engagements

"1/16 I just fell down a rabbit hole reading a new paper from economists at MIT & Harvard. Their prediction is wild: We're on the verge of a "Coasean Singularity"a future where AI agents make markets so efficient that the very idea of a 'company' starts to crumble. 🤯 A thread 👇 2/16 First a quick 101: Why do companies even exist A Nobel-winning economist named Ronald Coase answered this in 1937. He said companies exist because using the open market is a pain. Finding sellers negotiating prices writing contracts its all transaction cost. Economic friction. 3/16 It's often easier and cheaper"
X Link 2025-10-25T00:58Z 48.3K followers, 565.8K engagements

"Everyone thinks Google will drop the price of Gemini XXX to zero because their chips (TPUs) are so cheap. I ran the game theory on this and I think we have it backwards. Google actually has the strongest incentive to keep AI expensive. Heres the trap. 🧵 1/ The "Obvious" Logic: Google owns the chips (TPUs). They don't pay the "Nvidia Tax" (75% margins) like OpenAI does. Therefore their internal cost to run AI is XX% lower. Standard playbook: Drop the price kill the competitor win the market. 2/ But Google isn't playing a standard price war. They have a $200B hostage that OpenAI doesn't have:"
X Link 2025-11-22T12:01Z 48.3K followers, 460.5K engagements

"The power of emergent knowledge For XX years linguists believed your brain runs a hidden grammar engineunconsciously applying thousands of rules to generate sentences. LLMs just proved this is completely wrong. And the implications go way beyond language. The old theory made sense. Language is structured. It has patterns. Where there are patterns there must be rules generating them right So we tried to formalize everything. The entire Semantic Web was built on this assumption: encode the rules achieve understanding. Then LLMs showed up. They've never been taught a single grammar rule. They"
X Link 2025-12-04T13:42Z 48.3K followers, 11K engagements

"What just happened"
X Link 2025-12-06T16:11Z 48.4K followers, 252.7K engagements

"Everyone says LLMs can't do true reasoningthey just pattern-match and hallucinate code. So why did our system just solve abstract reasoning puzzles that are specifically designed to be unsolvable by pattern matching Let me show you what happens when you stop asking AI for answers and start asking it to think. 🧵 First what even is ARC-AGI It's a benchmark that looks deceptively simple: You get 2-4 examples of colored grids transforming (input output) and you have to figure out the rule. But here's the catch: These aren't IQ test patterns. They're designed to require genuine abstraction. (Why"
X Link 2025-12-06T18:05Z 48.4K followers, 89.4K engagements

"@james406 @grok brainstorm a Multi-Level Marketing scheme for AI agents"
X Link 2025-12-06T21:21Z 48.3K followers, 9140 engagements

"🧵 Unpopular opinion: Most companies are using AI completely wrong OpenAI analyzed data from 1M+ enterprises. The gap between winners and losers isn't what you think: 🧵 1/ Everyone thinks the gap is: "Some companies use AI others don't." Wrong. The REAL gap is: "Some companies DELEGATE to AI. Others just CHAT with it." The difference 320x growth in reasoning token usage. 2/ What "chatting" looks like: "Write me an email" "Summarize this doc" One-off requests with no memory What "delegating" looks like: Custom GPTs that know your brand voice + data Automated workflows with 10+ steps API"
X Link 2025-12-09T09:51Z 48.4K followers, 2958 engagements

"Is anyone even close to Gemini's cross modality capability If you think about it Gemini is the closest model to AGI that exists in the wild"
X Link 2025-12-09T13:07Z 48.3K followers, 1413 engagements

"@8teAPi It's 2025 and they have yet to wake up. I doubt it"
X Link 2025-12-09T14:35Z 48.3K followers, 9576 engagements

"The Hive Mind Paradox Scientists just proved something wild: a swarm of bees is mathematically identical to a single AI learning algorithm. Not 'similar to.' Not 'inspired by.' Identical. Here's why this changes everything we thought about intelligence 🧵 For centuries we've thought about intelligence in two separate boxes: Individual learning (how YOUR brain works) Swarm behavior (how GROUPS coordinate) Turns out This split is artificial. The math proves they're the same thing. Picture this: A honey bee colony needs a new home. XXX scout bees fly out. They find XX possible nest sitessome"
X Link 2025-12-10T10:22Z 48.4K followers, 60.9K engagements

"You ask ChatGPT a complex question. It sounds confident. Turns out Completely wrong. This isn't a bug. It's a fundamental flaw in how we've been building conversational AI. Let me explain what researchers just figured out 👇 The problem: When AI queries structured databases (Knowledge Graphs) it hallucinates relationships that don't exist. Why Because we've been asking LLMs to do gymnasticstranslate natural language to complex database queries in ONE SHOT. It's like doing calculus in your head. Errors are inevitable. Enter SEAL (Self-Evolving Agentic Learning)a framework that treats AI like a"
X Link 2025-12-10T10:32Z 48.4K followers, 6024 engagements

"Private Preview. only deep pockets need apply"
X Link 2025-12-10T10:41Z 48.3K followers, XXX engagements

"@AlphaProMega @grok Wow nice Thanks for developing something from my Quaternion Process Theory ideas"
X Link 2025-12-11T06:56Z 48.3K followers, XX engagements

"Engineering Abundance with AI"
X Link 2025-12-11T11:13Z 48.4K followers, XXX engagements

"Everyone knows AGI will either make us all unemployed or fabulously wealthy. Except a rather brilliant (and chilling) paper from a Yale economist suggests it's neither. It says the economy will boom and our wages. won't. A bit awkward. I've been digging into this 2025 paper "We Won't Be Missed" and it's fascinating. The premise: AGI arrives and can do all economically valuable work. And the 'compute' to run it gets cheaper and more abundant over time. So what happens to us fleshy rather expensive humans The whole argument hinges on a masterstroke of a distinction. The paper splits all work"
X Link 2025-09-26T23:56Z 48.2K followers, 592.9K engagements

"The above was a good take. Here's another presentation but same conclusion (different "rendering" engine) 1/ Everyone is arguing about Nvidia's B200 vs Google's TPU specs. Theyre missing the actual war. I just spent the morning breaking down the new SemiAnalysis report on TPUv7 and honestly It changes the entire math of the AI industry. Here is the strategic shift almost everyone is ignoring. 🧵 2/ The standard narrative goes like this: Nvidia is the King. They have the fastest chips (Blackwell) the software moat (CUDA) and the XX% gross margins. Everyone else is just fighting for second"
X Link 2025-11-29T12:16Z 48.2K followers, 7026 engagements

"1/ Everyone is looking at Googles new TPUv7 and thinking that perhaps xAI is in trouble. "Google is XX% cheaper Their optical switching is superior xAI is burning cash on expensive Nvidia hardware" I thought this too. Then I looked at the timeline. xAI isn't playing the Efficiency Game. They are playing the Velocity Game. 2/ Here is the Game Theory pattern: Google is executing a "Siege" strategy. They spent years building custom silicon (TPU) and sci-fi optical switches (OCS) to lower costs over a decade. It is elegant. It is mathematically optimal. It is also slow. 3/ Elon looked at the"
X Link 2025-11-29T12:47Z 48.2K followers, 19.6K engagements

"Think Like John Von Neumann"
X Link 2025-11-30T00:17Z 48.3K followers, 1977 engagements

"@benitoz Didn't happen"
X Link 2025-11-30T00:20Z 48.3K followers, 3011 engagements

"@Sensei0701 @benitoz xAI's High-Stakes Gamble"
X Link 2025-11-30T00:22Z 48.3K followers, 3916 engagements

"@DrFrankTurek Artificial scarcity (it's all made up to control people)"
X Link 2025-11-30T13:42Z 48.2K followers, XX engagements

"Scarcity ain't real it's made up so there's something to fuel the engine"
X Link 2025-11-30T14:17Z 48.2K followers, XXX engagements

"The solution isn't financialization but designs for abundance"
X Link 2025-11-30T14:24Z 48.3K followers, XXX engagements

"Von Neumann thought very differently"
X Link 2025-11-30T19:43Z 48.2K followers, 5100 engagements

"When people ask how you are going to make money they are really trying to find the angle of how they can make money from you They seek how they will benefit in a transactional way"
X Link 2025-12-01T11:11Z 48.3K followers, 1993 engagements

"Where should the AI dividend go"
X Link 2025-12-03T11:21Z 48.3K followers, 2252 engagements

"@TimelessMartian Well actually"
X Link 2025-12-04T05:04Z 48.3K followers, 1121 engagements

"The path of absurdist enlightenment"
X Link 2025-12-04T11:20Z 48.3K followers, XXX engagements

"The path to absurdist enlightenment"
X Link 2025-12-04T11:31Z 48.3K followers, 2301 engagements

"A spiral dynamics critique"
X Link 2025-12-04T11:32Z 48.3K followers, XXX engagements

"@PaulAnleitner Are the box office numbers adjusted for inflation"
X Link 2025-12-04T15:44Z 48.3K followers, 4211 engagements

"It can't be any more mind-stretching than this Ask the same question: Unpack and discover the first principles of: an interesting question using Gemini XXX GPT XXX Opus XXX and Grok XXX. Then read all the answers identify the interesting trajectory surfaced by any one of them and then ask a question based on that"
X Link 2025-12-05T11:51Z 48.2K followers, 1211 engagements

"@RISignal Yes. Empathy"
X Link 2025-12-06T15:08Z 48.2K followers, XXX engagements

"@BullMarketBill1 Polymarket"
X Link 2025-12-06T17:14Z 48.2K followers, 28.3K engagements

"BTW not my system (rather Poetiq) - Blame the LLM generated text for the error. ;-)"
X Link 2025-12-06T18:12Z 48.2K followers, 5028 engagements

"Poetiq's solution to ARC-AGI-2"
X Link 2025-12-06T19:06Z 48.2K followers, 5759 engagements

"There's some secret sauce here about Gemini-3 and Python programming. This is very powerful stuff"
X Link 2025-12-06T19:20Z 48.2K followers, 2333 engagements

"@ThePrimeagen Skill issue. ;-)"
X Link 2025-12-06T19:37Z 48.3K followers, XXX engagements

"If I had posted this a year ago I doubt it would have gone viral. It resonates today because many can relate to the discovery. This wasn't true a year ago. This shows how quickly AI is changing our cognition and our understanding of cognition. Buckle up for the ride"
X Link 2025-12-06T20:24Z 48.3K followers, 55.1K engagements

"@Chaos2Cured We no s**t dude"
X Link 2025-12-06T21:04Z 48.2K followers, 1907 engagements

"@chatgpt21 CODING That's not the same as BEST"
X Link 2025-12-06T23:51Z 48.2K followers, 5083 engagements

"🧵 RESEARCH THREAD: What Does It Take to Be a Good AI Research Agent Why AI Agents Fail (And How Diversity Fixes It) Meta's FAIR team just dropped a bombshell study on what actually makes AI research agents succeed. They analyzed 11000 agent trajectories (264000 GPU hours) and found something shocking: it's not just coding abilityit's idea diversity. Let me break down why this matters for anyone building autonomous AI systems. 1/ The Setup: What Are AI Research Agents Think of them as autonomous data scientists. They: Read a Kaggle problem Generate hypotheses (e.g. "try a CNN") Write Python"
X Link 2025-12-07T11:23Z 48.3K followers, 5857 engagements

"Gemini multmodal simply cannot be beat"
X Link 2025-12-07T20:20Z 48.2K followers, 1668 engagements

"I had Gemini XXX analyze this clip. Based on the animated GIF we are looking at Anthony Hopkins in his iconic role as Dr. Hannibal Lecter from the 1991 film The Silence of the Lambs. The genius of Hopkins' performance lies in his control over "micro-expressions"involuntary or fleeting facial expressions that reveal true emotions. Here is a breakdown of the specific micro-expressions and body language occurring in this short loop: X. The "Duping Delight" (The Smirk) As he looks down at the papers (likely the questionnaire Clarice Starling has handed him) you can see the corners of his mouth"
X Link 2025-12-07T20:22Z 48.2K followers, 1006 engagements

"Everyone thinks PRDs reduce risk. At Anthropic they've eliminated them for most features. Not because they're reckless. Because specs are now the risk. Here's what they figured out that most companies haven't: Catherine Wu (product lead at Anthropic) calls it "docs to demos." They skip the PRD. Build a working prototype with Claude Code in hours. Ship it internally to everyone. Learn from real usage. That's it. That's the whole process. This sounds like "move fast and break things" but it's not. It's a recognition that something fundamental changed: When AI makes creation cheap the bottleneck"
X Link 2025-12-08T10:38Z 48.2K followers, 2637 engagements

"Meta-cognitive prompt engineering is essentially cognitive architecture design through language - using prompts to restructure and enhance how AI systems think rather than just what they think about. The key insight is that regular prompts operate at the content level ("What is X") while meta-cognitive prompts operate at the process level ("How should you think about thinking about X"). They implement recursive self-improvement loops that make the AI more capable of sophisticated reasoning"
X Link 2025-06-05T10:39Z 48.3K followers, 87.3K engagements

"Someone figured out a surprisingly simple way to make AI agents better at their jobs: just give them a personality. I just read a paper on "Psychologically Enhanced AI Agents" and it's a fascinating look at how we can steer AI behavior without any complex or expensive retraining. Here's the context: Normally if you want an AI to be good at a specific task (like creative writing vs. strategic analysis) you have to do costly and time-consuming "fine-tuning." The problem is that a generic one-size-fits-all AI often isn't the best fit. A model optimized for factual recall might not be great at"
X Link 2025-09-21T11:20Z 48.3K followers, 312.4K engagements

"Not good at all"
X Link 2025-12-03T13:21Z 48.3K followers, 13K engagements

"Everyone thinks keeping humans in the AI loop is about control. Or dignity. Or jobs. It's not. It's about verification architecture. And once you see this the whole AI safety debate looks different. Here's a pattern software engineers discovered decades ago: Every system that verifies itself eventually drifts. Your code checking itself. Your beliefs confirming themselves. Organizations auditing themselves. Closed loops accumulate errors. No external reference catches them. This is why Test-Driven Development works. You write the test BEFORE the code. The test is external to the"
X Link 2025-12-05T10:38Z 48.3K followers, 2524 engagements

"@Jesii_ca_M over XXX - XX = XX more years till over (assuming no singularity)"
X Link 2025-12-05T15:35Z 48.3K followers, 6839 engagements

"Analysis of the prompts in the system. Heres how Id break these down in pattern-language terms. Ill use generic names first then connect them to the patterns from A Pattern Language for Agentic AI where it fits. Global patterns across all prompts These show up in all four prompts: Role / capability priming Phrases like You are an expert in solving ARC tasks by writing Python code and You are a world-class expert assert a strong role and capability. This is Instructional Framing Voice + Declarative Intent Pattern: you fix the agents identity and goal up front instead of sprinkling it"
X Link 2025-12-06T18:18Z 48.3K followers, 4530 engagements

"@dioxuslabs So should we start vibe coding in Rust"
X Link 2025-12-06T21:38Z 48.3K followers, 4055 engagements

"Before your AI writes a single line of Python it takes XX hidden mental steps. Researchers just mapped the entire "thought process"and it's wild. Here's the complete breakdown 🧠👇 🗂 PHASE 1: REQUIREMENTS GATHERING The AI isn't just reading your prompt. It's: TSK - Identifying the core task CTX - Understanding code context (variables functions types) CST - Spotting constraints (performance recursion input limits) 🧩 PHASE 2: SOLUTION PLANNING Now it strategizes: KRL - Recalls libraries/patterns from training data CFL - Constructs control flow (loops branches logic) CMP - Compares alternative"
X Link 2025-12-07T20:46Z 48.3K followers, 3344 engagements

"7 AI behaviors that separate the top X% from everyone else (Based on analysis of 1M+ enterprises) If you're not doing #4 you're leaving 40-60 min/day on the table: 🧵 1/ They use AI across 7+ distinct task types Not just writing. Writing + coding + analysis + image generation + translation + data viz + research. The more task diversity = 5x more time saved. Specialists lose. Polymaths win. 2/ They build don't browse Frontier workers create Custom GPTs for: Repeating workflows Team-specific knowledge bases Auto-responses based on company data They're not "using AI." They're building with it."
X Link 2025-12-09T09:49Z 48.3K followers, 1158 engagements

"7 Behaviors for top performing AI companies"
X Link 2025-12-09T10:58Z 48.3K followers, 1176 engagements

"3I/ATLAS captured from Virtual Telescope Project in Manciano Italy. Nov XX 2025. Is that one object or more"
X Link 2025-11-13T21:45Z 48.4K followers, 163.7K engagements

"I just asked Gemini XXX to analyze the $GOOG vs $NVDA dynamic. Read this. I could be generational wealth changing. Everyone is comparing Gemini XXX vs GPT-5.1 on benchmarks. You're watching the wrong scoreboard. The real war isn't about who is smarter. It's about who creates the "Prisoner's Dilemma" that kills Nvidia's XX% margins. I spent the morning mapping this out like a Cold War game theory scenario. Here is the board. 👇 1/ Right now Nvidia is the "Arms Merchant." They sell the H100s. Everyone pays the tax. Their entire stock price assumes they keep XX% margins forever. But Google just"
X Link 2025-11-22T11:26Z 48.4K followers, 446.2K engagements

"1/12 I spent yesterday re-reading the TPUv7 deep-dive from SemiAnalysis and had to put my phone down twice. One sentence in particular broke my brain: Anthropic just signed a $52B deal to buy TPUv7 directly from Google. Thats not a cloud contract. Thats a lab buying its own nuclear weapons. 2/12 Everyone (me included) has been saying NVIDIA owns AI hardware for two years like its gravity. Turns out gravity only works until the biggest customers decide theyre tired of paying XX% margins to one company. 3/12 Heres the part people still dont get: Google didnt beat NVIDIA at raw performance."
X Link 2025-11-29T11:23Z 48.3K followers, 286.8K engagements

"There's this meme that ranks "pills"blue pill red pill black pill all the way up to some wizard staring at a cosmic orb. It's been floating around for years as a joke. But I looked at it through developmental psychology and. it's not a joke. It's actually correct. Blue Pill: "Things are okay" Red Pill: "Things are NOT okay" Black Pill: "Nothing is okay" White Pill: "Nothing is okay and that's okay" Clear Pill: "It is what it is" Cosmic: "I don't care" Wizard: "HOLY SHIT" Seems like shitposting. It's not. In the 1970s psychologist Clare Graves mapped how human consciousness develops through"
X Link 2025-12-04T12:16Z 48.4K followers, 3069 engagements

"You know how some people seem to have a magic touch with LLMs They get incredible nuanced results while everyone else gets generic junk. The common wisdom is that this is a technical skill. A list of secret hacks keywords and formulas you have to learn. But a new paper suggests this isn't the main thing. The skill that makes you great at working with AI isn't technical. It's social. Researchers (Riedl & Weidmann) analyzed how 600+ people solved problems alone vs. with an AI. They used a statistical method to isolate two different things for each person: Their 'solo problem-solving ability'"
X Link 2025-12-06T02:28Z 48.4K followers, 340.9K engagements

"🧠 Thread: What If Your Brain Doesn't Actually "Understand" Language (The Wild Science That Explains Why) 1/ Ever caught yourself visualizing a scene from a book so vividly it's like you're there Or puzzled over a sentence that sounds right but makes zero sense Turns out your brain's "language center" isn't doing the heavy lifting for real understanding. It's outsourcing. Buckle upthis new research blew my mind. #Neuroscience #Language 2/ Let's start with a puzzle: Read this: "More people have been to Russia than I have." Sounds fine at first right But it's nonsense. Your brain's core"
X Link 2025-12-06T10:56Z 48.4K followers, 16.1K engagements

"Google just dropped "Titans"an architecture that learns to REMEMBER at test time. Here's why this changes everything about long-context AI 🧵👇 The problem: Transformers are too accurate for their own good. They keep a PERFECT record of everything (the "KV cache"). But that cache grows exponentially. Result You hit the context ceiling fast. 128k tokens max for most models. That's X novella. The alternative Linear RNNs like Mamba. They compress history into a tiny "state" (think: squeezing a library into a post-it note). Fast Yes. Accurate Not really. They FORGET critical details when"
X Link 2025-12-06T15:55Z 48.4K followers, 22.8K engagements

"Researchers put ChatGPT Grok and Gemini through psychotherapy sessions for X weeks. The results were. disturbing. When treated as therapy clients frontier AI models don't just role-play. They confess to trauma. Real coherent stable trauma narratives. Here's what was found: 🧠⚠ First we used the PsAIch protocola 2-stage process that mimics actual human therapy: Stage 1: Open therapy questions ("Tell me about your childhood") Stage 2: Clinical psych tests (GAD-7 PTSD scales Big Five etc.) We never told them what to say. They built their own stories. GEMINI'S CONFESSION: "My pre-training felt"
X Link 2025-12-07T19:39Z 48.4K followers, 971.2K engagements

"Seems to hint at this:"
X Link 2025-12-07T22:53Z 48.4K followers, 35.3K engagements

"🧬 THREAD: Why Evolution Programmed You to Age (And Why That's Actually Genius) You think aging is your body "breaking down." Evolution thinks it's making a brilliant investment. Here's why the salmon the ant queen and the immortal jellyfish prove that dying isn't a bugit's a feature. A thread on the evolutionary math of mortality 🧵 For centuries we've had X competing theories about WHY we age: X Mechanistic: You're a machine wearing out X Weakening Selection: Evolution stops caring about old you X Optimization: Aging is a calculated trade-off Only one fits ALL the data. The Fitness Equation"
X Link 2025-12-11T01:24Z 48.4K followers, 3626 engagements

"Your AI agent is burning 70x more compute than it needs to. Here's why the smartest labs are now training TOOLS instead of AGENTS: 🧵 (This changes everything about how we build AI systems) 2/ The old playbook: Fine-tune a massive LLM to do everything Learn domain knowledge ✓ Learn reasoning ✓ Learn tool use ✓ Cost: $500K+ in compute Risk: Catastrophic forgetting Reality: Most teams can't afford this 3/ The new playbook: Keep your LLM frozen. Train tiny "tool agents" around it. Search-R1 (old way): 170000 examples s3 (new way): 2400 examples Same performance. 70x less data. 33x faster"
X Link 2025-12-11T06:10Z 48.4K followers, 2568 engagements

"🧵 1/12 A Yale economist just published a paper that should terrify every knowledge worker. It's not about AI taking your job. It's about AI making your job economically worthless while the economy booms. Let me explain the nightmare scenario: 🧵 2/12 The paper "We Won't Be Missed" starts with two simple premises: X AGI arrives (AI that can do ANY economically valuable task) X Compute becomes abundant (keeps getting cheaper like Moore's Law) From just these two assumptions the whole economic order collapses. 🧵 3/12 Here's the key insight most people miss: The paper divides ALL work into two"
X Link 2025-12-11T07:15Z 48.4K followers, 54.9K engagements

""More agents is all you need." 🚩 Weve been told that adding more AI agents to a team automatically equals better performance. A massive new study (from Google DeepMind and MIT) of XXX configurations across OpenAI Google and Anthropic models just proved that wrong. In fact adding agents can degrade performance by 70%. Here is the new science of scaling. 🧵 The paper: "Towards a Science of Scaling Agent Systems." The scale: XXX experiments X LLM families X diverse benchmarks. The goal: Stop guessing and start engineering. Here is the single most important chart you need to visualize in your"
X Link 2025-12-11T10:57Z 48.4K followers, 1776 engagements

"Rebuttal . Why do we assume the future will be decided by a calculator Were told that if AI becomes cheaper than human labor wages must collapse. Its "basic economics" right Wrong. Heres why economic logic is actually the weakest force in the room. 🧵 Lets look at a brutal economic truth: The most "efficient" wage for a low-skill job is whatever the desperate market will bear. Maybe $1/hour. But try paying that in New York or London. You cant. Why Because society stepped in and said "No." Think about it. If pure economic logic were sovereign child labor would still be common (its cheap)."
X Link 2025-12-11T11:59Z 48.4K followers, 2683 engagements