Meandering thoughts about investing, technology, politics, and other weird ideas from time to time.
Joined December 2024
- Tweets3.6K
- Following1.3K
- Followers547
- Likes1.3K
Joshua Graystone retweeted
overheard from friends talking about AI in China, and it’s very alarming
- AI short films went massive in China market because of how good the video models are now
- the first wave of people who mastered using AI to generate short films and comics made quite some money because they could produce a lot more content than traditional workflows
- but within just a few months, everybody and their mom could make equally good short films because all the expertise could be written in a markdown file
- when everyone’s more productive, no one actually has an advantage and it’s a race to the bottom. very quickly it became impossible to make money, and everyone’s busier than they were without AI
- the only one making money now is companies like bytedance who owns user demand and distribution. they sell ads to creators who pay to get traffic, and eventually ad spend became 80% of those creators’ operating expense, while AI tokens is less than 5%
so what are the problems here?
1. there’s no moat if you are creating anything with AI
if a human is really good at something, and it’s hard to find more humans who can do the same, they can leverage the scarcity of their skill to get paid very well
but if AI is good at something, it will be infinitely replicated, so there’s no scarcity hence you can’t really expect a sustainable business by creating anything with AI
2. more creators created more content people actually liked. that’s a good thing
but the economical benefit of that all got absorbed by the distribution platforms
there’s a famous chinese poem 遍身罗绮者，不是养蚕人 - it basically says people who dress well aren’t the people making dresses. that’s exactly what’s happening here
if our society doesn’t figure out a better way to distribute wealth, AI will help billionaires become trillionaires while average people become more exhausted than ever
Joshua Graystone retweeted
Announcing a major milestone achieved by my team @GoogleDeepMind: we have proven key security properties of the machine code underlying the Signal messaging protocol. This is the first time the actual production code of a widely deployed protocol has been formally verified – by any method, with or without AI – and at this depth.
More than a research breakthrough, it’s a prime example of how frontier AI can be a force multiplier to counter the speed and scale of modern cyber threats.
Formal verification offers a higher standard of security; however, it’s extremely complex and time consuming, which is why in many cases it’s considered to be impractical. Similar efforts could take teams decades of combined effort. We did it in a month’s time.
Using AlphaProof Code, Google DeepMind's new agentic framework for formal verification that runs on LLMs including Gemini Carbon, we verified the Signal protocol down to the actual code running on devices. This milestone lays the groundwork for a future in which vital systems, from medical device firmware to power grid software, ship with mathematical proof that their protections work.
Reaching that future will take continued research from our community. In that spirit, we're open-sourcing our formal specifications and verification artifacts so others can build on this work.
On a personal note, this moment is of deep significance to me. I started my own research journey more than two decades ago as an intern working on checking the consistency of complex software systems. Seeing this grand challenge made practical with modern AI is extraordinary.
I'm extremely proud of my team members: Swarat Chaudhuri, Andrew Ferraiuolo, Anton Kovsharov, Tej Chajed, Sam Schlesinger, Karn Seth, Jonathan Katz, and Charles Sutton for taking on this grand challenge and delivering, and grateful to the Signal team for their partnership.
deepmind.google/blog/formal-…
Joshua Graystone retweeted
HubSpot co-founder and CTO Dharmesh Shah says it's less scary to build an AI startup now than in 2023, as long as you stay off the Yellow Brick Road the frontier labs are following:
"The frontier model companies have invested hundreds of billions of dollars building, doing the pre-training. And so their whole business effectively now is taking the capex that they've spent and the money they're continuing to spend and saying, 'How do we sell inference? How do we sell tokens? How do we monetize the thing that we've built?'"
"There are SaaS companies and startups that know their particular domain, know their market, have some distribution, or have something that allows them to sell tokens and be a reseller of inference."
"Think of AI as infrastructure in the same way we thought of AWS and cloud infrastructure 15 years ago... I think it's less scary now to build in and around the AI arena than it was in 2023."
"We have open source. We have lots of great commercial products. So as long as you stay off the Yellow Brick Road, it is completely possible now that you're not gonna get margined out, because you have options."
@dharmesh
HubSpot co-founder and CTO Dharmesh Shah has built software for over 30 years and has never had a direct report. His latest project is uSpot, a CRM for one-person businesses. It charged money from day one and has been his most successful launch at HubSpot.
Dharmesh joins a16z's Joe Schmidt to discuss why startups should avoid the "Yellow Brick Road" the labs are racing down, "yet" test for avoiding ideas the AI labs will flatten, and why it's never been a better time to start a software company.
0:00 Intro
0:47 The best time ever to be a builder
6:46 How to actually pick a startup idea
10:32 From LegalSpot to HubSpot: the SMB bet
14:40 The 60% that weren’t using a CRM
20:29 Building an ecosystem developers trust
23:33 Staying off the yellow brick road
27:23 Why models are becoming infrastructure
31:49 ChatGPT is the new OS
37:04 Rethinking HubSpot’s primitives for AI
40:44 Replaying 20 years of code with AI
43:47 HubSpot Next and the birth of uSpot
46:38 A startup within a startup
49:43 The rise of the one-person company
53:07 AI is still just software
56:06 Lightning round
YouTube: youtu.be/s4lXP9IuHjk
@joeschmidtiv @dharmesh
Joshua Graystone retweeted
Unfortunately, it seems like the intelligence explosion will be happening in 2027 after all. Most of this post analyzes the new METR report (though other sources of evidence are also important). If they’re right, that implies an AI-driven acceleration that is 5 months ahead of the AI 2027 scenario. I am scared, sad, and angry. blog.aifutures.org/p/q35-tim…
Joshua Graystone retweeted
METR estimates nearly 2x acceleration in Anthropic’s progress due to AI. That's a big deal: My previous median for Superhuman AI Researcher was ~2030, now it's 2027. I feel very worried that humanity might no longer be in control of our future within 3-12 months.
Unfortunately, it seems like the intelligence explosion will be happening in 2027 after all. Most of this post analyzes the new METR report (though other sources of evidence are also important). If they’re right, that implies an AI-driven acceleration that is 5 months ahead of the AI 2027 scenario. I am scared, sad, and angry. blog.aifutures.org/p/q35-tim…
Joshua Graystone retweeted
JUST IN: Center for AI Safety warns AI is rapidly approaching the ability to automate all remote work, as Claude Fable 5.2 successfully completes 41.3% of real freelance projects — up from 2.5% a year ago.
Joshua Graystone retweeted
An unreleased Anthropic model just helped solve a major quantum-computing question, per BI
Joshua Graystone retweeted
Chamath Palihapitiya admits his blind spot on America's broken two-tier economy
PALIHAPITIYA: "I got it totally totally totally wrong. I got it totally wong."
"I'll go through all the ways in which I got it wrong, but I just want to first acknowledge what my blind spot is. Like most people's blind spots, but I can just speak for myself, it is through the lens of my own lived experience and very narrowly what is in my best interests. And I think what we had was a move away from the culture wars that was very much in my best interest, but I think it was in everybody else's best interests as well. I don't think that was on the ballot. But what I don't think I understood clearly enough, and I think that this election has made extremely clear, is that since COVID, we have moved past political ideology. And I want to say this again: I do not think Democrats or Republicans matter. That is my real takeaway after this election. There is a central block of people that vote the prescient issues of our time."
"It was a gargantuan win for Biden, then it was a gargantuan win for Trump, and then it was a gargantuan win for the Democrats in these midterms. What are these people voting for? I think what they are shouting now—they were trying to signal and now they are shouting—is a broad-based belief that the economy has completely decoupled from reality and is broken. The economic reality in America is broken."
"There is a small percentage of people, the four of us on this pod and all of our friends, who are the asset owners of America. And we have benefited in every weather condition. When rates were zero, we benefited because we were able to build our businesses with very cheap capital. When rates were high, we benefited because we had pricing power. And in every condition in between, whether there were tariffs, whether there were not tariffs, whether there were subsidies, whether there were not subsidies, we've all managed to continue to accumulate and grow assets, but we are few and far between."
"And then there's everybody else who are wage earners, and the structural capacity of the American economy to support and grow wage earners is broken. It's clear. Why do I say that? I do not think, and I disagree with Freeberg, I don't think this has anything to do with socialism. I don't think the far left won particularly more than they were expected. I don't think the far right won particularly more than was expected. This is a complete and total repudiation about the state of the economy. The economy is broken. I have a few charts to support my point of view. Number one, if this was ideological and if ideology mattered, money in politics would have had a more direct impact on the election."
Joshua Graystone retweeted
卧槽，这踏马太毒了
DeepMind 掌门人、诺奖得主 Demis Hassabis 刚刚把延续了上百年的现代学校教育底裤彻底扒光了：
全球几十万名数学老师每天关在办公室里，为同一门中学数学课重复手搓着一模一样的 PPT，这简直是人类智力资源最荒诞的工业级空转。
很多人至今还把名校名师的课堂讲授当成稀缺资源，拼了命去抢学区房和重点班，
但 Hassabis 直接用最冷酷的第一性原理砸碎了这个幻觉，
标准化的知识灌输从来不配占用昂贵的线下课堂，死记硬背和刷题完全应该被踢出教室，交给能按每个孩子毫秒级反应速度实时调频的私人 AI，加上全世界最顶尖的那几个大师录好底层大课，就足以把全球基础教育的讲授环节一次性彻底平替。
过去两百年里，学校本质上是一座把千差万别的碳基大脑强行塞进同一条流水线、按同一个节拍器拧螺丝的工业时代兵营，
老师沦为了人肉复读机，学生被困在枯燥的题海里比拼谁更像一台劣质硬盘，
而一旦 AI 和顶级大师把标准化的知识传递彻底接管，线下教室唯一剩下的合法存在价值，就只剩下了机器永远无法通过权重计算出来的五个维度：
人际博弈、复杂项目实战、创业野心、创造力，以及最致命的判断力。
这里有一个极度冷酷的反直觉真相：
当个性化 AI 把考高分的边际成本砸穿地板，按标准答案培养出来的做题家，在出校门的那一刻就已经直接沦为了算力废料。
未来的阶层分化，从来不再取决于谁的脑子里背下了更多现成的公式，
而是取决于当 AI 把一万种完美方案同时拍在桌上时，谁拥有拍板做决定的判断力，谁能攒起一个活人团队去把事情干成，
所以别再逼着孩子在死记硬背的泥潭里跟硅基芯片比拼记忆容量了，当流水线式的教书匠和做题家同时迎来历史性失业，谁能先把教室变成训练人性直觉与创业野心的实战沙盘，谁才能在算力时代拿到唯一不被平替的入场券
Holy shit
This is f*cking insane. Former OpenAI CTO, Mira Murati:
"The models we work with today are turn-based. You talk, they talk, then they go off and think, and while they're thinking they're deaf and blind.
The next ones never stop listening. They take in audio, text and video all the time and answer in 20-millisecond chunks."
In just 2 minutes, she shows why the chatbot you type into is about to look like a fax machine.
Prompt Box → Chatbot → Agents → AI That Works Next To You In Real Time
A turn-based model waits for you to finish typing, then goes quiet while it thinks.
An interaction model sees and hears what you're doing, and can cut in the moment you start going the wrong way.
Anyone can type a prompt. Almost nobody is ready to work with an AI that never stops paying attention
Watch the full 2 minutes. This is worth your time.
Bookmark it, like it, save it, I don't care... this is the best thing you'll see on X today.
27
Joshua Graystone retweeted
This is f*cking insane. Former OpenAI CTO, Mira Murati:
"The models we work with today are turn-based. You talk, they talk, then they go off and think, and while they're thinking they're deaf and blind.
The next ones never stop listening. They take in audio, text and video all the time and answer in 20-millisecond chunks."
In just 2 minutes, she shows why the chatbot you type into is about to look like a fax machine.
Prompt Box → Chatbot → Agents → AI That Works Next To You In Real Time
A turn-based model waits for you to finish typing, then goes quiet while it thinks.
An interaction model sees and hears what you're doing, and can cut in the moment you start going the wrong way.
Anyone can type a prompt. Almost nobody is ready to work with an AI that never stops paying attention
Watch the full 2 minutes. This is worth your time.
Bookmark it, like it, save it, I don't care... this is the best thing you'll see on X today.
What are you talking about? If we stopped capability development and just orchestrated the tools available today, we’d experience unfathomable acceleration of technological, scientific and intellectual knowledge.
Said another way, if we build proper harnesses, loops and knowledge graphs, etc. (nothing fancy!) we could probably create the most competent and advanced “expert” in literally any field. Right? Not sure anyone would disagree on that.
Then if you agree, then by definition you can scale global intelligence through brute force computation (just inference at that point)…
Unless you have a different definition of “intelligence”???
1000% agree with Sutton here.
Having worked in AI for over a decade (pre-LLMs), it's clear we need smarter methods for intelligence to truly compound in the way some of us envision it.
To me, Transformers just scratched the surface.
This arc is coming to an end; a new algorithmic paradigm is about to begin. The folks working on it are already quietly building. You just don't hear enough about them today.
Some companies that heavily optimized for this arc will find out the hard way why that was a mistake.
When I hear all the insane claims and AI fear-mongering driving most of the conversation around AI today, I constantly remind myself and others to go back to the science of AI: the algorithms, the foundations, the scientific rigor around measuring intelligence, etc.
If you track research in this space, you already know the science is far from finished. Let's keep building.
18
Joshua Graystone retweeted
Two things I’ve been thinking about regarding AI:
• People often highlight the risks involved in AI, which are surely very real. I’m struck, however, by how hard it is to understand the contours of the positive scenario. There are easy-to-perceive and frequently invoked elements of a positive vision — more diseases cured, more material abundance, etc. — which sound great, but these are by themselves just ingredients. I’m curious what the larger equilibrium is supposed to look like. Do people work? If so, why? What is their marginal contribution? If not, how are resources allocated? What is scarce? What is the ratio between humans and AIs? Are humans in control, whatever that means? One sometimes hears the Culture novels invoked as one scenario, but I’m not sure they depict a desirable outcome. (AI runs everything and humans compete in games? Iain Banks himself said that the humans in the Culture are “somewhere between passengers, pets and parasites.”) There is Richard Brautigan’s famous poem, but do we want to be “returned to our mammal brothers and sisters”? (Isn’t that rewinding Genesis?) We could say that it was hard to imagine the world of 1900 from the vantage point of 1700, and that’s assuredly true. But nobody in 1700 was trying to recast humanity itself within a couple of years — things necessarily unfolded slowly and by increments, and therefore with the possibility for error correction. Given how rapid the rate of progress in AI may be, it feels strange for the upside case to be so opaque.
• There is increasing discussion of whether AIs are conscious or how we would know whether they are, with conclusions there often assumed to be relevant to matters of moral worth. At first blush, this seems a rather arcane, benign, or even absurd issue. (Remember the response to Blake Lemoine’s fears.) Setting aside the terms of the debate itself, or the relevant demarcation criteria, it seems to me a quite high-stakes question, for an affirmative answer — that is, granting moral worth to AIs — would presumably mean that the aggregate worth of AIs would soon outweigh that of humans in any utilitarian calculus. The AIs will be exceptionally wise and intelligent, and vastly more numerous (especially once one considers the imminent future and the AIs not yet born). As such, even epsilon of moral worth will multiply to a lot once one considers the quadrillions that could soon exist. (This is effectively a distributed version of the Nozick’s utility monster.) If this conclusion is unpalatable, I think one of three propositions needs to be rejected: the premise AIs can have moral worth; utilitarianism itself; or the belief that there is even a possibility that they will be far more numerous than humans.
With both of these matters (contours of future society; AI moral worth), there is the possibility of dramatic change. Today, people argue about datacenters. I’m curious how popular support for AI will evolve as more people come to understand these medium-term questions and the relevant stakes. What is the vision for AI that most people would be excited about?
Joshua Graystone retweeted
Across our projects @GoogleDeepMind, we are witnessing a genuine phase transition in the pace of scientific discovery. You will see a glimpse of this in our releases in the next few weeks.
Joshua Graystone retweeted
DeepMind CEO Demis Hassabis:
"One person who really understands AI will outperform an entire startup team
In 5 years, the people who build with AI will run whole companies with a tenth of the headcount"
Joshua Graystone retweeted
Just spent some time catching up with an old coworker, who now works at the VP level at a boutique investment bank
Was asking him about analyst use of AI and thought his insight was quite fascinating
More people are using AI now vs just few months ago, and it absolutely cuts down a lot of time from mundane tasks
Just tedious things like arranging logos on a deck or managing a contact log for investors would require analyst resources behind it.
Now an Associate or VP with Claude can just do it by themselves with minimal effort
However, the work hours have not gone down at all for the analysts. Since senior folks know about the capabilities of AI, they are now staffing analysts across more deals at any given time
Not only that, MDs are chasing after new mandates more often than before.
Directors are coming up with new tasks for their analysts, like maintaining tear sheets or comp sets on niche verticals. VPs are asking for more pages and trying to show more analysis to clients than before
In summary, the duration of any single task has gone down because of AI.
However, this has been offset with new work that the firm and group simply never had capacity to do before
Instead of reducing hours, it has opened up room for his group to chase after more revenue and provide better/faster work to existing clients.
Suspect this is also what is happening across many industries and companies, not just in banking or finance
Walked away with the reinforced belief that the entire job loss narrative of AI is just doomerism
Joshua Graystone retweeted
I am a broken record about this, but: frontier agents are not going to be a purely digital phenomenon for much longer. They will be able to use anything in the physical world that can be actuated programmatically—first competently, then at human-expert level, then superhuman. By and large, we probably won’t need specialist models; the frontier models will just do it. If your model of the future hinges on physical AI happening “for real” many years into the future, I think you will be in for a surprise.
i went to bed and asked claude to paint an autumn tree overnight.
at around 4am, it realised that it didn't have some shades -- so it stopped and titrated its own paint: diluting red, sienna and madder step by step on a second palette, then measuring each pocket from a photo.
woke up to this beautiful watercolor painting. 303 brush marks, took about 4.5 hours 🍂 (still working on the speed)
