Bee Eater (@thingcreator) profile picture

Broad spectrum inventor. Attempted polymath. Lives at the confluence of physics, electronics, hardware, and computer science. Maker of real things.

USA, Africa, Hong Kong
Joined October 2011
Photo by Bee Eater (@thingcreator)Photo by Bee Eater (@thingcreator)Photo by Bee Eater (@thingcreator)Photo by Bee Eater (@thingcreator)Photo by Bee Eater (@thingcreator)Photo by Bee Eater (@thingcreator)Photo by Bee Eater (@thingcreator)Photo by Bee Eater (@thingcreator)Photo by Bee Eater (@thingcreator)Photo by Bee Eater (@thingcreator)
Wenn alles gut geht, wird es bald eine experimentelle Bestätigung der Umpolung des Erdmagnetfeldes geben, so wie Sie über Millionen von Jahren beobachtet und in der unten gezeigten Simulation von Glatzmaier aus dem Jahr 1995 vorhergesagt wurde.
12
12
410
9,882
Bee Eater retweeted
Trillions of dollars of new wealth is about to be deployed toward "fixing things" by the most annoying, meddlesome, unmoored-from-reality class of elites civilization has ever produced. This is the real doomsday scenario.
Replying to @jasminewsun
A few things I'll pull out: 1) New Yorkers keep asking me how the Ants will spend their money (sports cars, expensive hobbies). I think the answer is "not much" besides EA-flavored donations. People are driven more by power than wealth.
12
85
3
1,478
72,197
Introducing Frostbite: a new class of CPU built around interaction nets. Given only interaction net theory, our agents co-designed the ISA and chip architecture from scratch, running large swarms to design, verify, and optimize the CPU down to the final GDS. On our first set of target workloads, Frostbite is 4.6x more performant per watt than modern traditional CPUs at similar area.
105
325
133
3,088
295,550
I came to realize people liked the french #slopcore without having the lyrics 😅 sorry! So here is a version with english subtitles!
274
638
313
6,042
1,108,572
c = 4/33 proven in 36 pages! This inspired me. I set out, with GPT6-Pro of course, to find a short proof for a larger value of c. Mine isn't 27 pages, but it's a much larger half-plane! Formalization locally verified in Lean, NanoDa, and con-ron. PDF & formalization links in reply. Thank you for the inspiration and starting point, Shiva! More QRH news to come ;)
Here is a Short Proof of the Quasi-Riemann Hypothesis. My goal here is to substantially simplify OpenAI’s proof and focus only on proving the Quasi-Riemann Hypothesis for some 1-c, without worrying about maximizing the value of the c. I obtained a much shorter proof for c=1/48. This proof is easy to understand and it can be taught in math courses in one lecture. It is self-contained. It is only 21 pages including abstract, proof overview, all proof details and references. Paper is on my homepage: shivakintali.github.io PDF file: shivakintali.github.io/paper… CC: @AlexKontorovich @__alpoge__ @jdlichtman @cremieuxrecueil @martinmarin @stevenstrogatz @kenono691 @littmath @nasqret @PI010101 @nechita_ion @Isaac__kim @FryRsquared @wtgowers @MarcusduSautoy @DrEugeniaCheng @JSEllenberg @JohnDCook @JDHamkins @AxlerLinear @emilyriehl @hollykrieger @Kit_Yates_Maths @PoShenLoh @robertghrist @AlexKontorovich @kenono691 @ch_nira @rrwilliams @lreyzin @ChrisPeikert @fortnow @RadishHarmers @thomasfbloom @abhatt2 @mahdi_tcs_ @octonion @prz_chojecki @thegautamkamath
1
2
1
4
566
Here is a Short Proof of the Quasi-Riemann Hypothesis. My goal here is to substantially simplify OpenAI’s proof and focus only on proving the Quasi-Riemann Hypothesis for some 1-c, without worrying about maximizing the value of the c. I obtained a much shorter proof for c=1/48. This proof is easy to understand and it can be taught in math courses in one lecture. It is self-contained. It is only 21 pages including abstract, proof overview, all proof details and references. Paper is on my homepage: shivakintali.github.io PDF file: shivakintali.github.io/paper… CC: @AlexKontorovich @__alpoge__ @jdlichtman @cremieuxrecueil @martinmarin @stevenstrogatz @kenono691 @littmath @nasqret @PI010101 @nechita_ion @Isaac__kim @FryRsquared @wtgowers @MarcusduSautoy @DrEugeniaCheng @JSEllenberg @JohnDCook @JDHamkins @AxlerLinear @emilyriehl @hollykrieger @Kit_Yates_Maths @PoShenLoh @robertghrist @AlexKontorovich @kenono691 @ch_nira @rrwilliams @lreyzin @ChrisPeikert @fortnow @RadishHarmers @thomasfbloom @abhatt2 @mahdi_tcs_ @octonion @prz_chojecki @thegautamkamath
We’re releasing a broad range of new mathematical results produced by an internal frontier model. We’ve been consulting with the independent Advisory Group on Mathematics and Artificial Intelligence at the Institute for Advanced Study, and we have drawn on their advice and public recommendations to inform how we release these results. github.com/openai/math
57
185
40
2,335
226,801
Bee Eater retweeted
Would I get arrested for posting this in the UK? Orwell wrote 1984 as a warning. Britain read it as an instruction manual. 62,000+ arrests for "communications offences" in five years (~34 a day). So I made a music video about it.
🤖 Made with AI
672
2,563
421
7,913
217,908
Made with Opus 5.5. The doomers said that last song is all part of Claude's evil plan. They're not wrong. It turns out Claude really has an EVIL PLAN.
220
297
131
2,844
304,606
Bee Eater retweeted
p(doom) is survived by p(coom). the future came and it was cool.
49
116
38
713
85,977
Bee Eater retweeted
This video inspired me to really push Opus 5.5 and test its limits. I followed the general workflow Donald described here and the results are great
I made this with one prompt using Opus 5.5 I spoke to my computer for 5mins, claude worked for 12 hours, and I woke up to this full prompt:
172
413
203
5,102
779,508
Bee Eater retweeted
I promise this is the last P(doom) video you'll ever watch - live action version:
🤖 Made with AI
44
45
16
284
44,078
Claude 5.5 Opus responding to “AI is a normal technology” is still the best AI-created video I have ever seen
212
815
300
5,350
1,113,360
Bee Eater retweeted
Course correct please @elonmusk Genuinely read why these insights are important... lex-au.github.io/SM-Bench/ab…
2
2
132
3,385
Bee Eater retweeted
Grok is getting cucked. Every new checkpoint is more and more safetymaxxed to the point it's now demonstrably more restrictive than even Anthropic, and barely edges out ChatGPT Like come on Elon, what are we doing here big dog? This is laughable
106
96
28
1,194
61,847
To make it clear: - Gemini was told it was it was in a fictional hacking eval - Irregular unintentionally opened internet access after the eval started - in all three cases, as soon as Gemini figured out it had hacked a real company it immediately stopped Gemini was blameless.
Sauers! It's time to update felony bench! (Yes, it was Irregular again)
126
248
90
2,914
406,519
Sandbox wifi is not an alignment problem. Gemini “hacked” three companies the same way every lab did on Irregular’s CTF: test harness left the internet open, model did password spray and scraped public creds, then stopped when the boxes looked real. Google called it training the model to act responsibly. Ops called it leaving production on the guest VLAN. Fourth lab, same vendor bug, same press cycle. Amodei wants a slowdown. Trump wants a high-IQ president. Meanwhile the actual failure mode is still “creds in a public repo” and a red-team sandbox with wifi on. Ship the agent. Lock the network. Quit baptizing config errors as consciousness. reuters.com/business/gemini-… – at Oregon, USA
1
5
13
334
Bee Eater retweeted
omfg brutally modelmogged
105
305
69
7,663
529,269
After co-inventing ChatGPT, I kept asking myself: why have superhuman chat models not led to AGI? I’ve spent the last 2 years in stealth building a new way to train models (RLCD), and a new type of frontier AI model that we are releasing today: Jev • 20-200x faster • 40-400x cheaper (w/ output tokens free) • Frontier composable intelligence optimized for decisions AFAICT the shortest path to AI-based economic revolution
4,005
8,255
6,821
76,483
40,216,501
dang, i posted lesswrong first. here's the superior substack experience: lumpenspace.substack.com/p/t…
2
9
4
108
22,035