@carl_feynman: signal 49/100
Is this your account? Share your reading:
#49 of 190 accounts read, fewer needle hits than 72% of them. The leaderboard
0 of these 19 posts are strong enough to show on their own: specific, new, first-hand and usable. With 3, this account joins Find, where people look for builders worth following. The next reading counts the newest posts.
carl feynman, the last 19 posts, read by six needles; the score counts the 5 that are words only. A reading of the words, not a verdict on the person, and not a prediction of reach.
Every post, ranked
Best to weakest by overall signal. Pick a needle to rank the posts by that needle alone.
-
1 of 19 Strongest post
363Kviews5.5Klikes468reposts137replies
Richard Feynman (my father) divided perpetual motion machines into Perpetual Motion Machines of the First Kind, which violate the first law of thermodynamics (AKA conservation of energy), Perpetual Motion Machines of the Second Kind, which violate the second law of thermodynamics (entropy must increase), and Those Other Perpetual Motion Machines, which purport to get their energy from other laws of physics. The Second Kind is rarer; the only example I'm familiar with is one that my dad showed me the plans for. Once in a while, someone would come to him with a perpetual motion machine prospectus and ask if it was a good investment. One of these was a machine that claimed to extract heat energy from the air and run a generator with it. They had built half the machine, which worked as far as they could test it, and were looking for gullible investors to pay for finishing it. By carefully going over the immensely complex plans, my father and I found a "heat exchanger" that was supposed to take warm freon and cool air, and produce hot freon and cold air. Naturally, this heat exchanger was in the as-yet unbuilt portion. Those Other Perpetual Motion Machines get their energy from heretofore unknown physics, like cold fusion, vacuum fluctuations, or secret new physical laws that will be disclosed only on the payment of $1,000,000. These are worthy of much more serious physical investigation, since it is entirely concievable that someone will discover a new easy way of making energy. Back around 1970, my father went to a public demonstration of one of Those Other Perpetual Motion Machines. He discovered an electric cord running out of the back of the machine, plugged into the wall. When he unplugged the machine and pointed out to the inventor that a perpetual motion machine that had to be plugged in wasn't really perpetual, the inventor pushed a button on the control panel and the machine exploded. Several spectators were severely injured; I believe one man lost an arm. The inventor sued my father on the grounds that he had caused the explosion; my father suspected that the explosion was deliberate. The trial ended up with my father not having to pay. Epistemic status: Childhood recollections, untainted by fact-checking. Originally an email from 1997.
signalVague2%Nothing new54%Secondhand52%Nothing to use43%Bait or brag27%Hype4% -
2 of 19 17Kviews192likes17reposts16replies
Here's an article making a good case for the AI boom turning into a crash in the next two years. It's the first such article that I find at all plausible-- it has none of the usual leftist mood affiliation and "AI is a fraud" absurdities that I've come to expect from such articles. It simply argues that the AI companies have committed to purchasing the compute from a lot of new data centers starting next year, and that the amount of this may far exceed their revenues. The companies will have to depend on people being willing to continue investing in them during 2027-2028. This applies a lot to OpenAI, and less so to Anthropic. It's a good article, and sobering.
signalVague3%Nothing new53%Secondhand98%Nothing to use27%Bait or brag11%Hype12% -
3 of 19 19Kviews230likes16reposts8replies
So here's someone suggesting that instead of quitting, I should have worked for safer AI inside an AI company. Why didn't I? When I'm part of a company, I find myself becoming loyal to the company, and adopting the values of my colleagues. This is a normal and salutary part of human nature, perhaps stronger in me than others. And AI is so technically sweet, I know I would end up advancing capabilities just because it's a great problem. But I didn't want to have even a tiny bit of personal responsibility for a bad singularity. So I had to stop working on AI. I don't have the emotional capacities to stay at a company while working subtly against its purposes. You may be right that an ideal politician could do more in my position, but I'm far from that ideal. But if the end times come I will say "I had no part in this!" At least I will have preserved my dignity.
This post carries a quoted post. The needles only read the words, so it is left out of the account's score.
signalVague68%Nothing new55%Secondhand38%Nothing to use28%Bait or brag32%Hype13% -
4 of 19 16Kviews122likes4reposts12replies
I couldn't figure out any good research avenues for making it safer. I'm an engineer-- my career has mostly been turning research ideas into shippable products. Every idea I could come up with made it stronger, faster, smarter. I didn't see any way to do safety with my skills.
This post carries a quoted post. The needles only read the words, so it is left out of the account's score.
signalVague53%Nothing new58%Secondhand17%Nothing to use72%Bait or brag23%Hype3% -
5 of 19 58Kviews277likes8reposts72replies
Years ago, when Modafinil was introduced, I thought it would be widely used so that people could sleep less and have more life-- exactky the reasons described in the quoted post. It wasn't addictive or tolerance inducing like amphetamine, and seemed to induce a state more like natural wakefulness and less like mania. But it didn't catch on like I expected. If that didn't catch on, why should we expect other anti-sleep medicines to become popular?
This post carries a quoted post. The needles only read the words, so it is left out of the account's score.
signalVague6%Nothing new63%Secondhand84%Nothing to use49%Bait or brag9%Hype8% -
6 of 19 60Kviews941likes65reposts15replies
So there's a big class of equations called PDEs, partial differential equations. Physicists use PDEs to model real problems, and mathematicians try to prove stuff about them. One of the things to prove about PDEs is whether their solutions can go to infinity in finite time ("blow up"). One direction of proof is to prove blowup-- specify a starting condition, and show how it goes to infinity. The harder direction of proof is to prove that a PDE can't blow up. This often involves considerable creativity in inventing some way of proving the no-blowup property for all possible starting conditions. Back at the turn of the century, when they were coming up with the millennium problems, the most important PDE for which blowup hadn't been decided was Navier-Stokes. Plenty of other more obscure PDEs had been proven to either blow up, or not blow up. (I personally considered it likely that NS couldn't blow up, and I'm delighted to be proven wrong.). Anyway, all of the known techniques had been tried on the NS equation, and it had proven resistant, so the Millennium Prize people figured that whichever way it resolved, some new techniques would be needed, and we would learn something. And here we are in the strange position of knowing NS blows up, but no human learning any new PDE techniques in the process. Perhaps looking at OpenAI's paper will teach us something. Epistemic status: I've read the technical paper that the Prize people caused to be published about the NS problem, and I know about PDEs in general. But this is all from memory, and I haven't fact-checked every statement above. "Most important" is a statement of opinion; please don't grief me.
This post carries a quoted post. The needles only read the words, so it is left out of the account's score.
signalVague2%Nothing new44%Secondhand84%Nothing to use76%Bait or brag34%Hype17% -
7 of 19 27Kviews254likes33reposts5replies
The HRT conjecture has been disproven with AI help. Here's what the conjecture says. Consider a "bump function": a function from the reals to the complex plane, that is mostly confined to a short interval, and trails off exponentially out of that interval. Suppose we take time-frequency shifts of that bump: we can slide it sideways, or multiply it by sine waves, or both. That gives us various other wiggly bumps. Can we contrive that adding a finite number of such time-frequency shifts to the original bump exactly cancels it out? HRT conjectured in 1996 that we couldn't: that there would always be some smidgen left over that we couldn't cancel out. (And I've always found that plausible.). But that's wrong! The paper proves the existence of such a function. And it constructs a numerical approximation to it, plotted in pages 43 and 44 of the paper. Terry Tao has a blog post talking about the proof in an easier way:
signalVague2%Nothing new32%Secondhand96%Nothing to use83%Bait or brag12%Hype14% -
8 of 19 73Kviews984likes45reposts39replies
The Navier-Stokes equation doesn't hold under real circumstances. For example, when fluids spin around fast enough, the pressure in the center of the vortex goes negative, and the fluid discontinuously 'breaks', forming a bubble. This is what would happen if you tried the new solution in real life. It's a strictly mathematical result.
This post carries a quoted post. The needles only read the words, so it is left out of the account's score.
signalVague4%Nothing new39%Secondhand94%Nothing to use79%Bait or brag8%Hype7% -
9 of 19 1.5Kviews15likes2reposts1replies
Here's a vulnerability that lets most ballots in Georgia be tied back to individual voters. It's been known since 2022, fixed by the voting machine manufacturer, but the software has never been updated in Georgia. I have to ask, is that because it's useful to someone? And if so, how? Epistemic status: I've only read one article about it, which may be wrong. Also, this violates the principle of charity toward idiots: never explain by malice what can be explained by stupidity.
signalVague4%Nothing new42%Secondhand96%Nothing to use76%Bait or brag29%Hype15% -
10 of 19 4.2Kviews58likes1reposts2replies
My kid got married in Cambridge City Hall instead of Boston City Hall, even though the happy couple lives in Boston, because Cambridge City Hall is beautiful and grand, while Boston City Hall is a brutalist labyrinth.
This post carries a quoted post. The needles only read the words, so it is left out of the account's score.
signalVague3%Nothing new89%Secondhand41%Nothing to use86%Bait or brag14%Hype3% -
11 of 19 33views0likes0reposts0replies
"You can see the world as it is, or you can care whether you live or die. You can't optimize along both axes simultaneously." Peter Watts, "The Twenty-one Second God"
signalVague34%Nothing new62%Secondhand98%Nothing to use58%Bait or brag7%Hype8% -
12 of 19 4.8Kviews39likes2reposts4replies
A very good question! What's been happening for the past decade is that engineers improve the design of a model and its training procedure, train it, and it comes out better than the last model: some combination of faster/smarter/cheaper/saner/more capable. As long as improving designs produces progress, they're going to want to keep training new models. When the technology tops out, they will keep vainly training for a while, and then give up. Technologies always top out, but it's very hard to call the timing before it happens. And it might not top out until AIs have taken over the world, in which case the motivational structure to train new AIs will change. Epistemic status: As certain as I can be talking about the future. I am an engineer who spent a career productizing AI, and I've read lots of histories of technology, and seen lots of technologies go through this cycle.
This post carries a quoted post. The needles only read the words, so it is left out of the account's score.
signalVague56%Nothing new77%Secondhand66%Nothing to use56%Bait or brag27%Hype18% -
13 of 19 7.6Kviews111likes3reposts8replies
25 years ago, I wasn't texting my family, scrolling social media, watching films, or chatting with AI on my computer. What sort of stick-in-the-mud does it take to only be doing things they could do 25 years ago?
This post carries a quoted post. The needles only read the words, so it is left out of the account's score.
signalVague14%Nothing new90%Secondhand73%Nothing to use83%Bait or brag50%Hype20% -
14 of 19 107Kviews775likes25reposts60replies
Have the national labs done anything since the turn of the century? Back in the '80s, I used to hear about all sorts of new science out of the labs. But I don't think they've done anything major for the last twenty-five years. It seems like they've been completely taken over by careerists.
This post carries a quoted post. The needles only read the words, so it is left out of the account's score.
signalVague19%Nothing new89%Secondhand72%Nothing to use93%Bait or brag15%Hype8% -
15 of 19 5.6Kviews65likes2reposts6replies
I think it's an AI-propelled economic boom and insane tariff policies causing collapse, happening at the same time in the same country. It averages out to modest growth.
This post carries a quoted post. The needles only read the words, so it is left out of the account's score.
signalVague44%Nothing new79%Secondhand97%Nothing to use85%Bait or brag7%Hype68% -
16 of 19 5.0Kviews41likes0reposts4replies
It would be great if we could get AI to the point where it can build itself, without it needing to be superintelligent. An industrial explosion would deliver all kinds of benefits. I hope we can get an industrial explosion without building a thing that will be smarter than us, take over, and do as it wants.
This post carries a quoted post. The needles only read the words, so it is left out of the account's score.
signalVague79%Nothing new80%Secondhand98%Nothing to use73%Bait or brag8%Hype61% -
17 of 19 4.9Kviews20likes0reposts0replies
Impressive results. Intuitively, I would have imagined this obviously wouldn't work. So, today I learned.
This post carries a quoted post. The needles only read the words, so it is left out of the account's score.
signalVague90%Nothing new87%Secondhand74%Nothing to use82%Bait or brag26%Hype14% -
18 of 19 15Kviews56likes1reposts3replies
This looks like a step forward in mathematics, but it doesn't count until you explain how you did it.
This post carries a quoted post. The needles only read the words, so it is left out of the account's score.
signalVague84%Nothing new88%Secondhand95%Nothing to use83%Bait or brag24%Hype11% -
19 of 19 Weakest post
3.1Kviews36likes4reposts8replies
It is indeed great to be all those things. The only dark spot is that while AI is wonderful, if it gets too smart it may kill us all. But other than that, AI is sure exciting.
This post carries a quoted post. The needles only read the words, so it is left out of the account's score.
signalVague69%Nothing new94%Secondhand98%Nothing to use91%Bait or brag14%Hype52%
Read on 2026-09-24. This is your account and you want it gone? Write to hello@postneedle.com. The leaderboard