Fuelled by β-Hydroxybutyric. | - NFA, crypto is a volatile asset & you may lose all your money. | Scared of CBDCs. | Questioning The Science™
Joined July 2020
- Tweets31K
- Following1.6K
- Followers2.7K
- Likes12K
Max Darachi retweeted
Dusting off this tweet from April because this gap in shared understanding of LLM capability is *widening*. It's now less "two groups of people speaking past each other" and more a sharp funnel.
- Napkin math somewhere around 6B people (~75% of the population) have barely come in contact with LLMs at all.
- Around 1-2B (~20%) are casual and infrequent users of free-tier ChatGPT-like products. This group sees derpy chatbots and treats them a bit like a better Google search, a writing aid, or etc. Maybe an agent tries to book you a flight. I have non-tech friends who (reasonably, imo) say they have not much use for it at all. Even many professionals outside of math&code are in this tier. For example, execs and many other functions spend a lot of their time talking to other people, so while they understand what is happening intellectually, it is still second-hand and a bit abstract.
- Now we get to professional use of frontier-grade LLMs in math&code. Somewhere around 20M people (0.2%) see first-hand that large, complex projects that used to take them weeks/months can now be completed by agents with a prompt. Building apps, copying apps, translating apps, decompiling apps from binaries... This has all happened very quickly and recently - less than 1 year ago, I was writing code manually by hand, typing memorized computer code commands into a code editor character by character, occasionally pressing Tab to autocomplete a little chunk of code.
- And finally we get to the ~5,000 people (~0.00006%) with access to frontier-grade systems internally. The external world has seen the preview. It looks like swarms of thousands of agents collaborating over weeks on software mega projects: minting zero days, running cyber attacks and defenses at machine speeds, discovering new science, advancing the frontier of mathematics. Things that would have taken top professionals in the industry years of work. Meanwhile, human review and comprehension are starting to fall behind. For example, people are still involved in the "archeology" of the OpenAI-HF incident from many months ago. Mathematicians may be poring over the 722 manuscripts on frontier mathematics for a while.
The funnel is driven by a combination of factors:
1. The impact scales with ambition, problem size, and horizon. A question with a paragraph answer barely stresses the system. You need a reservoir of big, difficult problems that you really care about. This aspect drives the consumer / professional dimension of the funnel.
2. The jaggedness of the system (which I have written about a lot separately). Capability peaks in domains that are digital, verifiable and economically valuable. This is because LLM capability emerges from reinforcement learning on verifiable rewards on a curated environment mixture driven by revenue potential. This aspect primarily drives the area (e.g. math&code) dimension of the funnel.
3. Access. Free-tier, paid-tier, internal.
So this is the weirdness of the moment. The general public has mostly not interacted with these systems. When they have, it looks like a derpy chatbot. The majority of professionals still see only a modest uplift. And a small sliver of professionals are experiencing the vertigo of the curve going vertical. And it is all happening at the same time.
Judging by my tl there is a growing gap in understanding of AI capability.
The first issue I think is around recency and tier of use. I think a lot of people tried the free tier of ChatGPT somewhere last year and allowed it to inform their views on AI a little too much. This is a group of reactions laughing at various quirks of the models, hallucinations, etc. Yes I also saw the viral videos of OpenAI's Advanced Voice mode fumbling simple queries like "should I drive or walk to the carwash". The thing is that these free and old/deprecated models don't reflect the capability in the latest round of state of the art agentic models of this year, especially OpenAI Codex and Claude Code.
But that brings me to the second issue. Even if people paid $200/month to use the state of the art models, a lot of the capabilities are relatively "peaky" in highly technical areas. Typical queries around search, writing, advice, etc. are *not* the domain that has made the most noticeable and dramatic strides in capability. Partly, this is due to the technical details of reinforcement learning and its use of verifiable rewards. But partly, it's also because these use cases are not sufficiently prioritized by the companies in their hillclimbing because they don't lead to as much $$$ value. The goldmines are elsewhere, and the focus comes along.
So that brings me to the second group of people, who *both* 1) pay for and use the state of the art frontier agentic models (OpenAI Codex / Claude Code) and 2) do so professionally in technical domains like programming, math and research. This group of people is subject to the highest amount of "AI Psychosis" because the recent improvements in these domains as of this year have been nothing short of staggering. When you hand a computer terminal to one of these models, you can now watch them melt programming problems that you'd normally expect to take days/weeks of work. It's this second group of people that assigns a much greater gravity to the capabilities, their slope, and various cyber-related repercussions.
TLDR the people in these two groups are speaking past each other. It really is simultaneously the case that OpenAI's free and I think slightly orphaned (?) "Advanced Voice Mode" will fumble the dumbest questions in your Instagram's reels and *at the same time*, OpenAI's highest-tier and paid Codex model will go off for 1 hour to coherently restructure an entire code base, or find and exploit vulnerabilities in computer systems. This part really works and has made dramatic strides because 2 properties: 1) these domains offer explicit reward functions that are verifiable meaning they are easily amenable to reinforcement learning training (e.g. unit tests passed yes or no, in contrast to writing, which is much harder to explicitly judge), but also 2) they are a lot more valuable in b2b settings, meaning that the biggest fraction of the team is focused on improving them. So here we are.
Max Darachi retweeted
This is a massive wake-up call.
Imagine your startup built on Fable 5 and it just went dark overnight with zero warning.
Anthropics own foreign national employees cannot even access it.
This is proof that you cannot fully rely on American tech infrastructure. Countries around the world need to start building for themselves.
Build sovereign. Build local.
Technology is the new weapon and today just proved it.
Max Darachi retweeted
The US government, citing national security authorities, has issued an export control directive to suspend all access to Fable 5 and Mythos 5 by any foreign national, whether inside or outside the United States, including foreign national Anthropic employees.
The net effect of this order is that we must abruptly disable Fable 5 and Mythos 5 for all our customers to ensure compliance.
Access to all other Claude models is not affected.
We apologize for this disruption to our customers. We believe this is a misunderstanding and are working to restore access as soon as possible.
Read our full statement: anthropic.com/news/fable-myt…
Max Darachi retweeted
We’ve agreed to a partnership with @SpaceX that will substantially increase our compute capacity.
This, along with our other recent compute deals, means that we’ve been able to increase our usage limits for Claude Code and the Claude API.
Max Darachi retweeted
Scam Altman …
xitter.gq/i/grok/share/49a0d68ad…
Max Darachi retweeted
Scam Altman has a incredible track record for being a con artist I don't think anyone has a "former ally turned enemy" list this big with directly with people he worked with
A massive new 18-month investigation dropped, revealing the full list of people who worked directly with Sam Altman and now openly say they don’t trust him - they call him a liar, manipulator, scam artist, and worse
These are his co-founders, board members, top executives, and biggest partners. Not random haters:
• Elon Musk (OpenAI co-founder)
➝ Betrayed the original nonprofit, open & safe AI mission and turned it into a closed profit machine
What he says: Calls him "Scam Altman" and “Sam Altman lies as easily as he breathes”
• Ilya Sutskever (OpenAI co-founder & former Chief Scientist)
Why: Discovered Sam repeatedly lied about safety protocols and bypassed board oversight. What he says/did: Compiled 70+ pages of memos, Slack messages, and evidence proving Sam’s lies → helped fire him. Said he didn’t think “Sam is the guy who should have his finger on the button for AGI”
• Dario Amodei (former OpenAI President, now Anthropic CEO)
Why: Left because of Sam’s leadership and broken safety promises
What he says: “The problem with OpenAI is Sam himself.” Called the company under Sam "mendacious” (full of lies) and compared it to Big Tobacco knowingly selling something dangerous. Accused him of a clear “pattern of behavior”
• Helen Toner (former OpenAI Board Member)
Why: Sam made it impossible for the board to do its job through constant deception
What she says: He was “outright lying to the board” and created a “toxic atmosphere” of psychological pressure
• Tasha McCauley (former OpenAI Board Member)
Why: Complete loss of trust after years of the same behavior
What she says: Senior leaders reported Sam cultivated a “toxic culture of lying”
• Jan Leike (former Superalignment co-lead)
Why: Sam deprioritized real safety work for shiny products
What he says: Resigned publicly saying he “lost confidence” in OpenAI leadership and that the company was “losing its way” on alignment
• Mira Murati (former CTO - one of Sam’s closest longtime collaborators)
Why: Lost all confidence in his leadership as they approached AGI
What she says: Told insiders “I don’t feel comfortable about Sam leading us to AGI” and said his playbook is to say whatever he needs to get what he wants, and if that fails, destroy your credibility
• Microsoft executives (including major tensions with CEO Satya Nadella)
Why: Felt constantly misled on deals and partnerships
What they say: A senior exec warned he could be remembered as a “Bernie Madoff or Sam Bankman-Fried-level scammer”
• Paul Graham (Y Combinator co-founder - Sam was YC President)
Why: Long pattern of deception during his time running YC
What he says: Privately told YC colleagues, “Sam had been lying to us all the time"
• Loopt board & early employees (Sam’s first startup)
Why: History of chaotic and deceptive behavior
What they did: Employees went to the board twice trying to get him fired over lack of honesty and shady behavior
These are his co-founders, board members, closest executives, and major partners who actually worked with him all say the exact same things - chronic lying, manipulation, broken trust, toxic culture, scam & deception
Max Darachi retweeted
Scam Altman and Greg Stockman stole a charity. Full stop.
Greg got tens of billions of stock for himself and Scam got dozens of OpenAI side deals with a piece of the action for himself, Y Combinator style. After this lawsuit, Scam will also be awarded tens of billions in stock directly.
The fundamental question is simply this:
Do you want to set legal precedent in the United States that it is ok to loot a charity? If so, you undermine all charitable giving in the United States forever.
I could have started OpenAI as a for-profit corporation. Instead, I started it, funded it, recruited critical talent and taught them everything I know about how to make a startup successful FOR THE PUBLIC GOOD.
Then they stole the charity.
Interesting how it works
Elon puts up his own money, rounds up the absolute best AI talent on the planet, leverages every connection he has to secure serious resources, and launches OpenAI in 2015 as a pure non-profit explicitly created to develop AI for the benefit of humanity, with zero profit motive and open research
Then the “team” decides they want the bag
They push Elon out, take control, and quietly flip the entire thing into a for-profit machine
All while preaching the same sanctimonious lines on repeat: “We’re still mission-driven!” “AI for the good of humanity!” “We’d never abandon our principles!”
The ultimate betrayal:
Elon got zero equity. Not a single share. He funded it. He built the foundation. He got nothing while they turned his non-profit into their personal cash cow
This is the level of betrayal and hypocrisy we’re dealing with
And for the record.... this lawsuit doesn’t put a single penny in Elon’s pocket. Any win goes straight back to the non-profit to restore the exact mission he founded
Max Darachi retweeted
AI shrinkflation is real, and Anthropic just got caught.
🚨 Data proves Claude Opus 4.6 is now "thinking" 67% less than before.
Anthropic said nothing until the data went public, prompting Claude Code creator Boris Cherny (@bcherny) to finally respond to the open GitHub issue.
Worse?
Leaked code allegedly shows employees have a secret switch for full access, while paying users get a nerfed, expensive version of ChatGPT.
The rumor mill says they're hoarding compute for their next model, Mythos.
Are paying users secretly funding their next R&D phase?
Max Darachi retweeted
Without questioning the ethical part of self-serving listicles (I've always said, if it is done very well, why not), I just wish I had the guts to publish something like this on my site and list myself as #1 best solution 😅
Kind of jealous I cannot find this kind of confidence in me 👑
And while this listicle is notably well done, there's an obvious strategy here:
✅ They list their own solution #1
✅ Their section is much more detailed than others
✅ They list smaller competitors (those that are not really seriously competing with them)
Again, not saying this is wrong. This one ranks and gets cited by LLMs a lot. Whatever works, guys. I just wish I were that bold!
Max Darachi retweeted
Called it
xitter.gq/Pranit/status/20330439…
Anthropic just pulled the oldest trick in SaaS pricing.
I pay $200/mo for Claude Max. My limits have been noticeably worse this past week.
Now they announce 2x off-peak usage for two weeks. Sounds generous.
But here’s what actually happens: limits quietly drop, a temporary 2x makes the reduced limit feel normal, the promo ends, and you’re left at a baseline lower than where you started. You just didn’t notice the downgrade because the 2x absorbed the transition.
These AI plans are massively subsidized. The raw compute behind a heavy user costs multiples of the subscription price. Every move like this is the subsidy quietly correcting.
Very sneaky, Anthropic.
Max Darachi retweeted
Yo @AnthropicAI why are you pillaging everyone's usage.
Has it been greatly reduced without telling anyone?
Max Darachi retweeted
Why in the world are you announcing this and not @AnthropicAI and @claudeai.
This is a massive change; you're not fooling anyone on why you're trying to meet the bare legal minimum to notify.