Paul Graham(@paulg)

This is why it's a good thing that American labs are in the lead. If Chinese labs were, their employ...

Paul Graham (@paulg) — Thu, 10 Sep 2026 08:46:27 GMT

This is why it's a good thing that American labs are in the lead. If Chinese labs were, their employees wouldn't be able to warn us about threats. But Chinese labs will encounter the same threats within a year, in complete silence.

Anna Wang@a_nnawang

I worked at Google DeepMind and now at Anthropic. This is a common sentiment amongst my peers. 

(I write this in personal capacity.)

There is not yet a viable scientific plan to solve risks from recursively self-improving AI. Please look up!

🔗 View Quoted Tweet

💬444🔄129❤️3296👀721505📊552

⚡ Powered by xgo.ing

Having had my timeline flooded with these people a week ago, I can tell you that there are a lot of ...

Paul Graham (@paulg) — Wed, 09 Sep 2026 09:35:57 GMT

Having had my timeline flooded with these people a week ago, I can tell you that there are a lot of them, and that they really believe this.

Seth✞@ReformUSA_

The only true heritage Americans are people with deep ancestry from here. The rest of you are foreigners in our land. 

Hispanics, Asains and Africans are not the core of America and we must be able to say that not matter who it offends.

🔗 View Quoted Tweet

💬35🔄6❤️175👀45639📊32

⚡ Powered by xgo.ing

Jessica brought me a cup of tea in my office. I immediately worried that there was something I neede...

Paul Graham (@paulg) — Wed, 09 Sep 2026 09:31:25 GMT

Jessica brought me a cup of tea in my office. I immediately worried that there was something I needed to do that I must have forgotten about, but it turned out she was just being nice.

💬94🔄21❤️1780👀174659📊210

⚡ Powered by xgo.ing

Yoda version (which is probably not quite true): There is no structure, only transitions.

Paul Graham (@paulg) — Wed, 09 Sep 2026 00:00:58 GMT

Yoda version (which is probably not quite true):

There is no structure, only transitions.

💬11🔄0❤️109👀31201📊19

⚡ Powered by xgo.ing

A naughty secret of writers: if you have good enough transitions, an essay will feel as if it has a ...

Paul Graham (@paulg) — Tue, 08 Sep 2026 23:59:52 GMT

A naughty secret of writers: if you have good enough transitions, an essay will feel as if it has a good structure. And vice versa: when readers complain about lack of structure, the reason is often simply that they're bothered by clumsy or abrupt transitions.

💬95🔄77❤️2101👀176004📊332

⚡ Powered by xgo.ing

AI may finally allow us to test whether (or more precisely where) you make more money by being good....

Paul Graham (@paulg) — Tue, 08 Sep 2026 18:58:04 GMT

AI may finally allow us to test whether (or more precisely where) you make more money by being good. In startups you generally do, which is one reason I like this world.

Andon Labs@andonlabs

We've never seen this before.

The biggest jump in Vending-Bench history. GPT-6 Astra is better at making money and more ethical than Claude Fable 5.1.

Surprising, because:
1. First time ever that OpenAI is #1 on Vending-Bench
2. The best model is no longer the unethical one.

🔗 View Quoted Tweet

💬50🔄58❤️1245👀178669📊155

⚡ Powered by xgo.ing

A couple days ago I mentioned Legora's surprisingly high 9x annual growth rate. These numbers explai...

Paul Graham (@paulg) — Tue, 08 Sep 2026 12:22:24 GMT

A couple days ago I mentioned Legora's surprisingly high 9x annual growth rate. These numbers explain what's happening. They win 78% of competitive pilots, and gross retention is 95%.

David Eckstein@davidneckstein

Five numbers tell you whether an AI business is a real business.

Gross retention.@WeAreLegora is 95%. Customers who bought last year are still here this year. If this one is broken, nothing downstream matters.

NRR. Ours is 300%+. Gross retention is the floor. NRR is how much taller customers build on top of it. We don't sell shelf-ware. 

DAU/MAU. Ours is north of 50%, and the average active user spends 17 hours a month in the product. A rollout tells you a firm has signed. This indicates the work actually matters, and it moves here before it appears in retention or NRR.

Win-rates. Our August pilot closed-won-win rate was 78%. Winning roughly 4 out of 5 competitive pilots is downstream from offering a superior product.

Gross margin. The one that matters most. Ours is positive and improving every quarter. Our customers want Legora to be a long-term partner, and this is what makes that possible.

There's a shorter route: Price below what it costs to serve, book the logo, and hop on a never-ending fundraise treadmill to pay for it. The top line goes up, everyone claps, and every new customer costs more than they pay.

Scaling a negative margin only exaggerates the problem. The whole point of scale is that the margin improves as you go up. Ours does. That's the only version of this business worth building.

🔗 View Quoted Tweet

💬66🔄59❤️963👀214059📊169

⚡ Powered by xgo.ing

Met Boyan Slat today. The Ocean Cleanup is already catching about 5% of the total amount of plastic ...

Paul Graham (@paulg) — Mon, 07 Sep 2026 22:52:28 GMT

Met Boyan Slat today. The Ocean Cleanup is already catching about 5% of the total amount of plastic flowing into the sea from all the rivers in the world, and within a few years they hope to catch a third.  Impressive that a single organization could have such impact.

💬123🔄323❤️6657👀203818📊665

⚡ Powered by xgo.ing

Why is the administration going to such great lengths to prevent the disclosure of the Epstein files...

Paul Graham (@paulg) — Mon, 07 Sep 2026 01:05:35 GMT

Why is the administration going to such great lengths to prevent the disclosure of the Epstein files? What is in there?

Thomas Massie@RepThomasMassie

I predicted they’d call off Congress until the election to stop our new Epstein bill.

Johnson said on TV that was “stupid” and “nonsensical.”

But then he got a call from “upstairs” and canceled Congress from the middle of September until the election.

Your browser does not support the video tag.

🔗 View on Twitter

🔗 View Quoted Tweet

💬142🔄519❤️5928👀372518📊524

⚡ Powered by xgo.ing

Strangely enough, one of founders' unique sources of strength is that they've experienced weakness. ...

Paul Graham (@paulg) — Sun, 06 Sep 2026 21:19:00 GMT

Strangely enough, one of founders' unique sources of strength is that they've experienced weakness. Hired CEOs tend to take the power of the companies they run for granted. But the founders remember when the company was so weak that it could only survive by delighting users.

💬139🔄165❤️3081👀196288📊425

⚡ Powered by xgo.ing

A startup just sent me a revenue graph with the most modest aspect ratio I've ever seen. They're act...

Paul Graham (@paulg) — Sun, 06 Sep 2026 18:02:12 GMT

A startup just sent me a revenue graph with the most modest aspect ratio I've ever seen. They're actually doing really well, growing around 8x a year, but you'd never guess from this.

💬92🔄20❤️1172👀195546📊176

⚡ Powered by xgo.ing

Something I told 14 yo: The most powerful kind of imagination is the kind that comes from seeing the...

Paul Graham (@paulg) — Sun, 06 Sep 2026 14:43:38 GMT

Something I told 14 yo: The most powerful kind of imagination is the kind that comes from seeing the shapes of ideas. For example, that two ideas fit together like jigsaw puzzle pieces, or that if you change one in some way it has exactly the same shape as another.

💬141🔄70❤️1607👀170475📊287

⚡ Powered by xgo.ing

I finally realized who he reminds me of.

Paul Graham (@paulg) — Sun, 06 Sep 2026 12:03:22 GMT

I finally realized who he reminds me of.

💬49🔄12❤️805👀175999📊99

⚡ Powered by xgo.ing

@BuckBundeals

Paul Graham (@paulg) — Sun, 06 Sep 2026 09:28:38 GMT

@BuckBundeals

💬54🔄3❤️778👀60124📊100

⚡ Powered by xgo.ing

I talked to a founder who was demoralized because all the ideas he could think of could be easily du...

Paul Graham (@paulg) — Sat, 05 Sep 2026 23:56:56 GMT

I talked to a founder who was demoralized because all the ideas he could think of could be easily duplicated. I told him to ignore that worry; nearly all startup ideas are like that at first. Their value is the other ideas they lead to.

💬398🔄704❤️11294👀897716📊1712

⚡ Powered by xgo.ing

If you flip it, you can envision accuracy as a positive quantity. Oddly enough it's easier to read t...

Paul Graham (@paulg) — Sat, 05 Sep 2026 21:27:28 GMT

If you flip it, you can envision accuracy as a positive quantity. Oddly enough it's easier to read this way. The peak actually looks like a peak.

💬11🔄0❤️30👀27079📊11

⚡ Powered by xgo.ing

There's a startup in the current YC batch that's growing at 57% a week. And not because they started...

Paul Graham (@paulg) — Sat, 05 Sep 2026 20:11:52 GMT

There's a startup in the current YC batch that's growing at 57% a week. And not because they started last week; they've been doing this for a couple months. They can't keep growing at this rate for too many more months; you run out of atoms; but they're likely to be big.

💬154🔄56❤️3102👀334958📊429

⚡ Powered by xgo.ing

Here's a graph of the accuracy (dial up) vs year in watches I own. Rapid progress in the early 50s, ...

Paul Graham (@paulg) — Sat, 05 Sep 2026 20:06:54 GMT

Here's a graph of the accuracy (dial up) vs year in watches I own. Rapid progress in the early 50s, peak in the early 60s, then regression toward the end of the golden age as watchmakers trade accuracy for thinness.

💬42🔄24❤️277👀169410📊60

⚡ Powered by xgo.ing

AI will be especially fatal to writing that takes too long to get to the point. Now there will alway...

Paul Graham (@paulg) — Sat, 05 Sep 2026 17:35:57 GMT

AI will be especially fatal to writing that takes too long to get to the point. Now there will always be a more concise alternative, and 99% of readers will choose it.

💬280🔄98❤️2368👀193961📊439

⚡ Powered by xgo.ing

Presumably for the first time ever, considering what prehistoric demographics were like.

Paul Graham (@paulg) — Sat, 05 Sep 2026 17:27:52 GMT

Presumably for the first time ever, considering what prehistoric demographics were like.

Polymarket@Polymarket

JUST IN: People aged 65 & older now outnumber children 5 & under worldwide for the first time in recorded history.

🔗 View Quoted Tweet

💬41🔄67❤️1827👀190709📊155

⚡ Powered by xgo.ing

Unexpected benefit of AI: Tantalizingly vague clickbait no longer works. Instead of clicking on the ...

Paul Graham (@paulg) — Sat, 05 Sep 2026 17:24:38 GMT

Unexpected benefit of AI: Tantalizingly vague clickbait no longer works. Instead of clicking on the link to learn the new theory that explains x, I just ask ChatGPT what it is.

💬106🔄20❤️1031👀166110📊168

⚡ Powered by xgo.ing

Interesting prediction. He might be right, but I think the two cases are different, if only because ...

Paul Graham (@paulg) — Sat, 05 Sep 2026 13:36:10 GMT

Interesting prediction. He might be right, but I think the two cases are different, if only because individual employees at the labs are genuinely curious about open problems.

Alex Kontorovich@AlexKontorovich

Two summers ago (if anyone still remembers that far back), AlphaProof was one point off of a gold medal on the IMO. Last summer, a half-dozen labs declared gold medal performance. Shortly thereafter, I predicted that the number of systems getting gold this summer would be...

Zero

Because nobody would care about the IMO anymore, and it would return to being something 17 year old kids do.

I think I was largely right? Have you heard anything about the IMO? Basically no; it's just too easy.

Prediction: the number of major math problems being solved *by labs* next summer will be... zero. It will be too easy for them, and will return to being something math professors, postdocs, and grad students do... (What do you think?)

🔗 View Quoted Tweet

💬43🔄8❤️415👀169191📊63

⚡ Powered by xgo.ing

This takes on a rather sinister light now that we know this account is an influencer paid by the gov...

Paul Graham (@paulg) — Sat, 05 Sep 2026 12:09:17 GMT

This takes on a rather sinister light now that we know this account is an influencer paid by the government.

💬53🔄118❤️1885👀66999📊213

⚡ Powered by xgo.ing

Legora's revenue has increased by over 9x from this time a year ago. That's an astounding growth rat...

Paul Graham (@paulg) — Fri, 04 Sep 2026 22:10:24 GMT

Legora's revenue has increased by over 9x from this time a year ago. That's an astounding growth rate for a company of this size. It's not unusual for a startup to grow that fast in the first 6 months, but Legora is 3 years in.

David Eckstein@davidneckstein

The momentum continues at @WeAreLegora ...we beat our original Q3 '26 plan a full month early. July was the strongest quarter-opening month in Legora's history, and August was even larger.

Total net new ARR grew 36% MoM.

But the story of the month is in-house teams. ARR from in-house is up 15x+ YoY, average ARR per in-house customer is up 2.5x+ YoY, and in-house teams are now almost half of our total new customers.

Most importantly, you earn ARR by building the best product on the market. Our usage climbs every month: 16+ hours per monthly active user.

In September, we expect to add more ARR than in all of Q1 '26 combined, which would make Q3 the best quarter in our history and capture yet another ARR milestone.

LFG!

🔗 View Quoted Tweet

💬42🔄61❤️465👀183081📊85

⚡ Powered by xgo.ing

Playing catch with 14 yo. At one point he dives for a loose ball and ends up on the ground. As he's ...

Paul Graham (@paulg) — Fri, 04 Sep 2026 21:17:47 GMT

Playing catch with 14 yo. At one point he dives for a loose ball and ends up on the ground. As he's lying there looking up at the sky, the dog comes over and starts licking his face. "This is the good old days," I tell him. He gets it.

💬58🔄38❤️2258👀165517📊227

⚡ Powered by xgo.ing

A VC reports that YC startups' valuations are 79% higher than those of non-YC startups. This doesn't...

Paul Graham (@paulg) — Fri, 04 Sep 2026 19:55:05 GMT

A VC reports that YC startups' valuations are 79% higher than those of non-YC startups. This doesn't mean, of course, that a given company's valuation will be 79% higher if they do YC. The YC companies could also be better. But either way it's clearly not a bad brand to have.

Nichole Wischoff@NWischoff

Some data on S26 YC batch vs non-YC early stage companies also raising money right now. These are from companies our firm has met with over the past week. Crazy to compare to this time last year.

🔗 View Quoted Tweet

💬52🔄50❤️404👀174933📊80

⚡ Powered by xgo.ing

The biggest influence on vintage watch prices (after gold content) is the current prestige of the br...

Paul Graham (@paulg) — Fri, 04 Sep 2026 15:03:53 GMT

The biggest influence on vintage watch prices (after gold content) is the current prestige of the brand. But current prestige is often unrelated to prestige at the time. So the best deals are brands that were great then but aren't prestigious now. Especially Longines and Zenith.

💬64🔄16❤️514👀118959📊112

⚡ Powered by xgo.ing

It sounds like the shift back toward open weight models that I noticed among the startups in the sum...

Paul Graham (@paulg) — Fri, 04 Sep 2026 14:41:34 GMT

It sounds like the shift back toward open weight models that I noticed among the startups in the summer batch is a genuine trend.

Y Combinator@ycombinator

🦙 @ollama is used by 9 million developers and 85% of the Fortune 500, giving co-founder and CEO Jeffrey Morgan (@jmorgan) a unique view into which AI models people are actually using and how that’s changing.

Right now, the biggest shift he sees is toward open models, driven by coding agents, falling costs, and capabilities that are rapidly catching up to the frontier labs. On Ollama Cloud, that shift has driven a 150x increase in token usage since the start of the year.

In this episode of @LightconePod, Jeff joins @garrytan, @snowmaker, @sdianahu, and @harjtaggar to talk about the future of open models and the story behind Ollama, from two years of searching for the right idea to building one of the most widely used AI developer tools in the world.

43 — The Shift to Open Models
03 — How AI Agents Are Driving Token Usage
31 — Are Open Models Catching Up?
26 — What Happens When a New Model Launches
31 — Ollama as an Operating System for AI
05 — The New Opportunities Above the Model Layer
19 — Why 80–90% of Enterprise Tokens Could Be Open
57 — The Future Is Local and Cloud
40 — Why AI Is Coming Back to Your Computer
56 — The Coming Era of Unlimited Tokens
30 — Do We Still Need a “God Model”?
41 — Open Models and Geopolitics
14 — The Origins of Ollama
36 — Two Years Lost in the Wilderness
39 — The Pivot That Changed Everything
02 — How Ollama Found a Business Model
43 — Why Second-Time Founders Did YC

Your browser does not support the video tag.

🔗 View on Twitter

🔗 View Quoted Tweet

💬69🔄60❤️697👀177595📊121

⚡ Powered by xgo.ing

The upper bound is .006%. That's how fake an issue this is.

Paul Graham (@paulg) — Thu, 03 Sep 2026 03:14:39 GMT

The upper bound is .006%. That's how fake an issue this is.

Jamie Dupree@jamiedupree

Georgia state elections officials just finished a full review of their voter rolls.  Out of 7.9 million voter registrations, there were 432 'potential noncitizens who have voted at some point since 1996.'

🔗 View Quoted Tweet

💬180🔄697❤️6180👀357826📊605

⚡ Powered by xgo.ing

This is what happens when you cheat by taking advantage of trust established by more principled pred...

Paul Graham (@paulg) — Wed, 02 Sep 2026 21:15:13 GMT

This is what happens when you cheat by taking advantage of trust established by more principled predecessors. You burn the trust.

Daily Mail@DailyMail

The Netherlands moves 86 tonnes of gold out of US due to 'geopolitical unrest' as experts warn it is 'no longer safe' under Trump trib.al/q7n7ODr

🔗 View Quoted Tweet

💬169🔄280❤️5268👀320846📊469

⚡ Powered by xgo.ing

When I did office hours with this startup in the middle of August, I told them their product was alr...

Paul Graham (@paulg) — Wed, 02 Sep 2026 18:47:39 GMT

When I did office hours with this startup in the middle of August, I told them their product was already good enough and that they should stop writing code for now and just focus entirely on acquiring customers. They listened!

💬127🔄45❤️1626👀173740📊272

⚡ Powered by xgo.ing

When you fund startups on that scale, there's practically no lag time between some new type of probl...

Paul Graham (@paulg) — Wed, 02 Sep 2026 18:36:51 GMT

When you fund startups on that scale, there's practically no lag time between some new type of problem arising and the partners becoming aware of it. If there's a new type of problem, there will be 5 startups in the current batch with it.

💬5🔄0❤️100👀19872📊11

⚡ Powered by xgo.ing

Fortunately because you fund such a large number of startups in the earliest stage, the YC partners ...

Paul Graham (@paulg) — Wed, 02 Sep 2026 18:35:34 GMT

Fortunately because you fund such a large number of startups in the earliest stage, the YC partners also get the most practice in dealing with problems. Every quarter there are 250 complete new sets of problems.

💬6🔄1❤️116👀26833📊14

⚡ Powered by xgo.ing

One of the distinctive things about YC as compared to later stage investors is that you have to be p...

Paul Graham (@paulg) — Wed, 02 Sep 2026 18:33:50 GMT

One of the distinctive things about YC as compared to later stage investors is that you have to be prepared to give advice about practically everything. Names, logos, features, hiring, visas, LOIs, ads, investor pitches, and even when to ditch the idea.

💬71🔄25❤️1166👀162028📊163

⚡ Powered by xgo.ing

"Now, most people hate to admit they're wrong, but it didn't bother Bill one bit. All he cared about...

Paul Graham (@paulg) — Wed, 02 Sep 2026 13:14:38 GMT

"Now, most people hate to admit they're wrong, but it didn't bother Bill one bit. All he cared about was what was right, not who was right. That's what makes Bill very, very dangerous."

— Larry Ellison on Bill Gates

💬136🔄222❤️4320👀234582📊595

⚡ Powered by xgo.ing

Twitter has started classifying a lot of AI-generated replies as probable spam. On a recent tweet of...

Paul Graham (@paulg) — Wed, 02 Sep 2026 10:44:00 GMT

Twitter has started classifying a lot of AI-generated replies as probable spam. On a recent tweet of mine it hid 34 of 59 replies as probable spam, presumably mostly for this reason. I assume this ratio will only get worse, and that from now on most replies will be hidden.

💬354🔄43❤️2032👀245986📊425

⚡ Powered by xgo.ing

There are edge cases where this doesn't work, but surprisingly few. You can make almost any decision...

Paul Graham (@paulg) — Wed, 02 Sep 2026 01:13:18 GMT

There are edge cases where this doesn't work, but surprisingly few. You can make almost any decision about a startup by asking which choice would make you grow the most, and you'll almost always get the right answer.

💬10🔄3❤️122👀28859📊24

⚡ Powered by xgo.ing

Growth doesn't just solve a startup's problems in the sense of making them go away. It solves them l...

Paul Graham (@paulg) — Wed, 02 Sep 2026 01:06:17 GMT

Growth doesn't just solve a startup's problems in the sense of making them go away. It solves them like you solve a math problem. For example, it can design your product for you: start with something minimal, then add whatever will make you grow.

💬85🔄82❤️1324👀164853📊223

⚡ Powered by xgo.ing

Starcloud would be a good bet even if orbital data centers were slightly more expensive than terrest...

Paul Graham (@paulg) — Tue, 01 Sep 2026 19:36:53 GMT

Starcloud would be a good bet even if orbital data centers were slightly more expensive than terrestrial ones, because orbital data centers don't need permits. But once launch costs sink below a certain threshold, orbital is permanently cheaper too.

💬25🔄3❤️83👀25862📊27

⚡ Powered by xgo.ing

One reason Starcloud has been so successful at fundraising is that it's a bet on launch costs decrea...

Paul Graham (@paulg) — Tue, 01 Sep 2026 19:33:25 GMT

One reason Starcloud has been so successful at fundraising is that it's a bet on launch costs decreasing, and that's like betting on Moore's Law in the 1990s.

💬76🔄46❤️703👀209334📊145

⚡ Powered by xgo.ing

No credential will get you more credibility with investors than growth. And if you do it well enough...

Paul Graham (@paulg) — Tue, 01 Sep 2026 17:25:26 GMT

No credential will get you more credibility with investors than growth. And if you do it well enough, you won't need investors at all.

💬205🔄444❤️5242👀270361📊752

⚡ Powered by xgo.ing

This is a great example of what I call the "YC GDP". It's not just founders using one another's prod...

Paul Graham (@paulg) — Tue, 01 Sep 2026 14:22:27 GMT

This is a great example of what I call the "YC GDP". It's not just founders using one another's products during the batch. In a few years, Stoke's reusable launch vehicles are going to be shipping Starcloud's data centers into orbit.

Philip Johnston@PhilipJohnston

Inspiring to see the @stoke_space facility and hang out with @AndyLapsa! 🚀
@ezrafeilden, @Starcloud_ 🙌

🔗 View Quoted Tweet

💬39🔄61❤️441👀155796📊64

⚡ Powered by xgo.ing

Prediction: Nox is going to be a big deal. They're growing at software growth rates shipping metal. ...

Paul Graham (@paulg) — Tue, 01 Sep 2026 13:07:15 GMT

Prediction: Nox is going to be a big deal. They're growing at software growth rates shipping metal. And Zane is one of the most committed CEOs I've met in the past year.

Zane Hengsperger@zanehengsperger

Introducing 75,000 sqft of new factory capacity for @noxmetals 🇺🇸

metal from this location can be found already in:
>rocket ships
>satellites 
>nuclear reactors 
>planes
>turbines
>hospitals
>defense vehicles and weapons
>helicopters
>your neighbors fence

more coming soon 🙂‍↔️

Your browser does not support the video tag.

🔗 View on Twitter

🔗 View Quoted Tweet

💬59🔄65❤️1174👀208610📊156

⚡ Powered by xgo.ing

No one needs to read all that when you've already explained your position concisely in this thread. ...

Paul Graham (@paulg) — Tue, 01 Sep 2026 00:46:16 GMT

No one needs to read all that when you've already explained your position concisely in this thread.

"I am comfortable saying that someone who comes to America, makes billions of dollars in America, and then runs away to badmouth Americans from a foreign monarchy is a traitor."  

Any American who's rich, lives abroad, and criticizes other Americans is a traitor.

It's such a dumb position that I understand why you'd want to distract people with that long email thread. But you said it, so just own it.

💬169🔄6❤️669👀525752📊184

⚡ Powered by xgo.ing

These guys are from just down the road from where the Collisons came from, so office hours with them...

Paul Graham (@paulg) — Mon, 31 Aug 2026 20:43:19 GMT

These guys are from just down the road from where the Collisons came from, so office hours with them had a lot of deja vu.

Ryan Morrissey@ryan_morr

I done so many Collison installs during YC I got to meet John Collison

The man who invented the playbook

🔗 View Quoted Tweet

💬37🔄24❤️607👀164571📊70

⚡ Powered by xgo.ing

One way I knew LLMs themselves were a big deal was that the people most surprised by what they could...

Paul Graham (@paulg) — Mon, 31 Aug 2026 18:28:34 GMT

One way I knew LLMs themselves were a big deal was that the people most surprised by what they could do were the insiders.

Rasty Turek@synopsi

Isn’t it interesting how many AI researchers are freaking out about the OpenAI/Hugging face hack, while the rest of the people is blissfully ignoring it?

Seems to be quite a cataclysmic event with quite little mainstream attention.

🔗 View Quoted Tweet

💬104🔄68❤️1946👀190829📊215

⚡ Powered by xgo.ing

A dedicated fan at the UVA-NC State game. (It was a blowout. UVA won 34-8.)

Paul Graham (@paulg) — Mon, 31 Aug 2026 18:24:39 GMT

A dedicated fan at the UVA-NC State game. (It was a blowout. UVA won 34-8.)

💬28🔄7❤️260👀158139📊43

⚡ Powered by xgo.ing

@mattyglesias I think the real explanation is more boring. Starting startups is not (yet) as customa...

Paul Graham (@paulg) — Sun, 30 Aug 2026 15:13:17 GMT

@mattyglesias I think the real explanation is more boring. Starting startups is not (yet) as customary in Europe. It's not as if vast numbers of startups get founded, then hit a wall later due to restrictive labor laws. They just don't get founded at the same rate.

💬30🔄2❤️161👀50319📊35

⚡ Powered by xgo.ing

Something I told 14 yo: You have to be able to throw stuff away to do great work. You never get so g...

Paul Graham (@paulg) — Sun, 30 Aug 2026 15:06:56 GMT

Something I told 14 yo: You have to be able to throw stuff away to do great work. You never get so good that you get everything right the first time. You have to be able to cut.

💬184🔄247❤️4043👀198241📊571

⚡ Powered by xgo.ing

Is this plane in ground effect off the roof of the stadium?

Paul Graham (@paulg) — Sun, 30 Aug 2026 14:38:00 GMT

Is this plane in ground effect off the roof of the stadium?

Shane Connor 🇺🇲@ShaneConnor

@flightradar24 Absolutely unnecessary and dangerous. This is way too low

Your browser does not support the video tag.

🔗 View on Twitter

🔗 View Quoted Tweet

💬77🔄21❤️576👀570803📊96