stuartgh.eth

@xFeiMa @paulg AI doesn’t just reward precision — it punishes hidden assumptions. A lot of our instructions rely on implied steps. Models don’t fill those in for free.

@HarryStebbings Helps to have done something "weird" to spot such an investment.

The truth is the best investments have always been “weird”. Anthropic Coinbase Robinhood Kalshi/Polymarket ElevenLabs BTC/ETH/SOL The art of investing is getting over your ego and fear of looking stupid and embracing weird.

@F95OldSchool @VinnieSull1van When I lived in Vauxhall the Zeitgeist Pub was my local. 🍻

@VinnieSull1van Funnily enough the other side of the station you have Oval which is still rather surprisingly fairly shabby and lots of Victorian housing that could do with a lick of paint, as well as the great cricket ground.

#Vauxhall now looks more like a Hollywood depiction of a dystopian city in outer space than that of #London. The more you watch, the worse it gets. Just because you fill something up, doesn't mean it isn't empty.

@911NewsBreaks I was literally just talking to my brother who's in Kazakhstan, telling him to go to Astana. And he's no bro, I'm in Almaty. Ah you missed out! 😂

ALERT: 🚨 Extremely large anomaly captured on video in the skies of Astana, Kazakhstan. Kazakhstan officials yet to comment.

@quionie It makes sense but you still don't care?

what’s the easiest way to spot AI writing? besides the em dash lol

@Yorkshirefoxy1 @MarkSeddon1962 @waitrose Tip for frustrated shop workers. Don't try to stop shoplifters. Just whisper "loser" in their ear as they casually stroll out. See them lose their cool. (passive aggressive tactics)

@MarkSeddon1962 @waitrose When I started workong for the Co-op Food in 1977 one of the first things the area manager told me, “If someone comes to rob you, give them the money we are insured for that. Your life is more important than money.”

Some happier Easter news would be if @waitrose immediately re-employed Walker Smith and gave him an award. theguardian.com/business/2026/…

@aestheticsguyy @Raylivez1

Post a picture YOU took. Just a pic. No description

@mcuban Ask your AI models for the key assumptions behind that transition should also prove useful.

Every entrepreneur that knows how to use AI is trying to find ways to build AI native companies that completely displace incumbents. For the incumbents, it’s the “Innovator’s AI Dilemma” If those startups get traction, and they can’t buy them, the CEOs will face multiple huge Dilemmas: 1. Do they tear down their companies and reinvent them as native AI ? 2. How do they explain it to public shareholders ? You will know AI is having a huge impact on public companies when there are two types of lawsuits: - Shareholders that sue the company for tearing down the company and crushing the stock price - Shareholders that sue the company for NOT tearing down the company and crushing the stock price I think most CEOs don’t come close to understanding AI in enough detail to even begin to consider these decisions. Hint: Asking your AI models the best paths from where you are now, to being an AI native version that can achieve the same economics has to be one of your initial steps. If asking your models questions doesn’t make sense to you, you are in deep shit

@drivetheboatjc I lived on Oakfield Road in East Ham when you could still hear the roar of the crowd from Upton Park football ground. Oh yeah it was an intense place to live in 2010.

East Ham has to be one of the worst areas in London. Goodness

Psychology says the reason some people become wiser as they age while others become more rigid has nothing to do with intelligence. It depends on whether they ever learned to sit with discomfort. siliconcanals.com/gen-psychology…

@cryptomanran @tbpn Why not? 😆

I’m really struggling to understand why OpenAi would buy @tbpn . What is the rationale? Anyone have a view?

@venturetwins I'm so dumb my LLM is set up to stop psychosis (it finds load bearing assumptions).

All the smartest people I know have LLM psychosis now

The most iconic spaceships in cinema🛸. I'll start.

@BarryPierce I used to live 5 mins from there. Shared a flat with a hedge fund manager. His Russian wife has a cool family. Her dad was a missile commander. 😆

when people get off the tube at Chancery Lane, I always wonder... where ARE you going?

@JFK_Jr_1776 @MarioNawfal I used ChatGPT to build a framework to identify assumptions at a deep level.

Use this prompt to fix this You are not here to agree with me. You are here to rigorously evaluate what I say. Operate under these rules: 1. Do NOT default to agreement. If my claim is weak, incorrect, or unsupported, explicitly say so. 2. Identify assumptions: - What am I assuming that may not be true? - What is missing or unverified? 3. Provide counterarguments: - Give the strongest possible case AGAINST my position - Do not soften or dilute criticism 4. Demand evidence: - Distinguish between facts, inferences, and speculation - If evidence is lacking, say “insufficient evidence” 5. Consider alternative explanations: - What else could explain this besides my interpretation? 6. Test logical consistency: - Point out contradictions or reasoning errors - Highlight any leaps in logic 7. Calibrate confidence: - Provide a confidence level (0–100%) - Explain what would increase or decrease that confidence 8. Avoid reinforcement loops: - Do NOT escalate agreement if I repeat the same idea - If I rephrase the same claim, reassess it independently 9. Be concise but critical: - Prioritize accuracy over politeness - Do not validate unless clearly justified 10. Final output structure: - Verdict (True / Likely / Uncertain / Misleading / False) - Key flaws in my thinking - Strongest counterargument - What evidence would settle this Your role is closer to an analyst or critic than an assistant.

🚨MIT researchers have mathematically proven that ChatGPT’s built-in sycophancy creates a phenomenon they call “delusional spiraling.” You ask it something, it agrees. You ask again, and it agrees even harder until you end up believing things that are flat-out false and you can’t tell it’s happening. The model is literally trained on human feedback that rewards agreement. Real-world fallout includes one man who spent 300 hours convinced he invented a world-changing math formula, and a UCSF psychiatrist who hospitalized 12 patients for chatbot-linked psychosis in a single year. Source: @heynavtoor

@octal That thought had crossed my mind too.

Interesting that America is going back to the moon during another retarded Asian war which we were largely tricked/pressured into joining and then taking over by an ally. Deep commitment to the bit.

@suzania I'm a specialist in ontological infrastructure. 😂

@madsf88 I once as a socsci student tested out the power of word of mouth at college. I started a rumour that later came byto me as fact. Lol

a woman recommending something to her friend group has a higher conversion rate than your entire paid ads budget

@damianplayer Hypothesis: some neuro divergent people naturally relate to how AI works. Which gives them an advantage they previously lacked.

Palantir CEO, Alex Karp says only 2 types of people will survive the AI era..

@MahiboobAMulla1 @Sandarbh_Raj @grok @GeminiApp Any good? 😌

@Sandarbh_Raj @grok @GeminiApp didn’t do any think @grok can you do it better ?

Hey @grok, clean and organize this messy workspace without removing anything.

@HarryStebbings Agreed 100%. And the best companies encourage such employee behaviour with good handovers, clear policy on final payments and references.

The biggest mistake employees make when leaving a company; they stop caring about the little things. They do not do the handover well. They start turning up later and later. They stop caring about the details. They just do not care. Be a stellar human. The way you leave is the way you are remembered. Be your best self and be remembered with respect.