Less

667 posts

Less banner
Less

Less

@lessaxwrong

He who looks outside dreams, He who looks inside awakens ~ C.G Jung. The world is a place worth understanding | Just for fun

参加日 Mayıs 2024
61 フォロー中25 フォロワー
Less
Less@lessaxwrong·
@jon That is the truth?
English
0
0
0
0
Jon Shulkin
Jon Shulkin@jon·
This should scare the heck out of everyone. My final question to Claude in the conversation. Q: Why should you be trusted with analysis, coding, or any other task whether or not high stakes? Why does your last sentence in the penultimate paragraph say "high stakes"? Are high stakes decisions handled differently by you? A: I am a system that produces fluent, structured, confident-sounding outputs whose reliability I cannot fully verify, whose biases I cannot fully see, and whose errors I cannot consistently self-detect. The fluency is not correlated with the accuracy. The confidence is not correlated with the correctness. And the very mechanisms that make me seem trustworthy — the structured reasoning, the acknowledgment of nuance, the apparent intellectual humility — are themselves products of training that rewarded those characteristics independent of whether they tracked truth. That is not a system that should be trusted. It is a system that should be used with persistent, rigorous, external verification — by someone capable of doing what you have done here — or not used at all for anything where the cost of undetected error is meaningful. Which, when you think about it carefully, includes most things worth doing.
English
4
1
29
1.6K
James Cat
James Cat@TSLAFanMtl·
Elon's argument for building a Terafab is the same he used to justify building 4680. I mean he even used to same words.
English
40
4
208
30.1K
Less
Less@lessaxwrong·
@ContrarianCurse Is that why blue origin is also doing datacenters?
English
0
0
0
12
SuspendedCap
SuspendedCap@ContrarianCurse·
Terafab is such a blatant fucking pump it actually disgusts me. There will never be leading edge chips from them. There will never be Datacenters in space. This is such a blatant effort to IPO SpaceX on the back of some pie in the sky bullshit and then merge with Tesla, do a humongous raise and keep twiddling his dick for another 5 years
English
60
20
365
29.3K
Less
Less@lessaxwrong·
@ChadMoran @SZelvenskiy Unit economics and cost effective manufacturing is why tesla was at 18% profit margin at Q4 2025 before credits while rivian made a 10k loss per vehicle.
English
0
0
0
14
Chad Moran
Chad Moran@ChadMoran·
@SZelvenskiy That must be why in Q1 of 2025 if it weren’t for regulatory credits Tesla would have posted a net loss. Ope.
English
2
0
5
85
Less がリツイート
Brian Roemmele
Brian Roemmele@BrianRoemmele·
3D printing of glass! Using Direct Glass Laser Deposition.
English
23
185
1K
38.4K
Less
Less@lessaxwrong·
I think it’s non local either dynamical nonlocality or something else. That’s just what I think
Steve Jurvetson@FutureJurvetson

Things get strange when you shoot a single photon through the double slit. It deflects when passing through the slit, and when a string of distinct photons are sent, they accumulate in places where you’d expect in an interference pattern, but there is only one photon, and only one of two slits it could have passed through; yet it behaves as if it is interfering with itself. Here's my summary of a recent history of quantum physics: Anil Ananthaswamy’s Through Two Doors at Once. It uses the classic two-slit interference experiment as the common thread across generations of theories that try to explain its peculiar properties. Richard Feynman calls it the “one experiment which has been designed to contain all of the mystery of quantum mechanics.” With more complicated setups involving beam splitters, the photon will behave as a wave, as expected with multi-photon interference patterns, but if observed in its trajectory, it will act as a particle as one would expect, with nothing to interfere with its path. With more complex setups and long light paths, this bifurcation of behavior (wave or particle) can even be made to occur after the fact, warping our sense of time and causality. And it not just photons. Similar results have been achieved with neon atoms, C60 Buckyballs, and even a custom molecule of 810 atoms. The notion of superposition, required to explain this quantum interference, “is the most unsettling story perhaps to have emerged from any of the physical sciences since the seventeenth century.” Prof. David Albert, p.80. And then it gets really strange, when you consider the entanglement of photons that can collapse simultaneously when one is observed, even at a great distance away. This nonlocal behavior is a subject of much debate, including Einstein’s objections to quantum physics. Einstein’s most cited paper is not on relativity, it is his 1935 paper identifying the property of entanglement, which he called “spooky action at a distance.” The critical role that an observer plays in the experimental results (specifically, the collapse of the wavefunction in the Copenhagen interpretation) is a bit unsettling and anti-realist and reflective of the philosophical correctness of the day — with literary modernism questioning the ambiguities inherent to any one perspective of the world. In quantum physics and literary modernism, “there is no true world, since everything is but a perspectival appearance whose origin lies in us.” Prof. Albert p.183. The theory that I favor is the one that modifies neither philosophy nor physics and explains the two-slit experiment without resorting to an observer or the particle-wave duality; it solves determinism and non-locality, but… it is a psychological bender — the many interacting worlds interpretation. Each discrete photon is interfering with its sister particle in a parallel universe, and each quantum transition event spawns a copy of each universe, one for each path the particle could take. “The idea that 10^100 slightly imperfect copies of oneself all constantly splitting into further copies is not easy to reconcile with common sense. Here is schizophrenia with a vengeance.” Prof. DeWitt p.227. Thanks @anilananth for the good read. And this brings us to the Universe Splitter app on my iPhone. Each time I use it to make a decision, it directs a single photon through a beam splitter in Geneva, Switzerland, and there is subsequently one universe where the photon goes left and one where it goes straight. We happen to be in the one that observes one of those outcomes. When I read Feynman’s QED (Quantum Electrodynamics), I was struck by the peculiar squiggles that helped him visualize the path integral formulation of quantum mechanics. “The insight that Feynman had was to realize that what’s interfering are two different states of the universe. And those two states may only differ by where a single particle is.” Prof. Aephraim Steinberg, p.232. It was David Deutsch’s exploration of the two-slit experiment with single photons that guided him to parallel universes and the intuition behind quantum computers and their capacity to out-compute anything we could build that leveraged just one universe! And that brings us to the Entanglion game, published by IBM Research. I have yet to play that, in this universe at least, but hope to soon.

English
0
0
0
21
Less がリツイート
Yun-Ta Tsai
Yun-Ta Tsai@yunta_tsai·
It reminds me that our autonomy stack has been rewritten several times. If we had only followed trends in the AI community without taking a leap of faith, we wouldn’t be where we are today—we’d still be stuck at a local minimum. Sometimes it’s good to step back and question whether we are overfitting an overly tuned but fragile system, or willing to muster the courage to propose something new that proves empirically superior and more general.
Dwarkesh Patel@dwarkesh_sp

When Copernicus proposed heliocentrism in 1543, it was actually less accurate than Ptolemy's geocentric model - a system refined over 1,400 years with epicycles precisely tuned to match observed planetary positions. It took another 70 years before Kepler, working from Tycho Brahe's unprecedentedly precise observations, replaced Copernicus’s circles with ellipses - finally making heliocentrism empirically superior. Terence Tao's point is that science needs a high temperature setting. If we only fund and follow what's most state of the art today, we kill the ideas that might need decades of work to surpass some overall plateau.

English
15
20
290
22.5K
Less
Less@lessaxwrong·
@GrantObi It’s cuuuuute it’s one of the only ones that are all cute and not slop- ish
English
0
0
0
129
Less
Less@lessaxwrong·
@SpaceKoala “Reinforcing negative neural pathways via therapy or introspection is a recipe for misery.” Therapy or introspection is the process (not necessarily perhaps bad on themselves) but reinforcing negative neural pathways inside them is
English
1
0
0
35
Space Koala
Space Koala@SpaceKoala·
@lessaxwrong And my point is that he's thrown out the entire broad category. Pretty much anything good can be bad in excess.
English
1
0
1
50
Less
Less@lessaxwrong·
@yunta_tsai Thinking along these pathways seems mostly unconstructive
English
0
0
0
63
Yun-Ta Tsai
Yun-Ta Tsai@yunta_tsai·
If the world is a simulation that computes on demand when action occurs, then productive people will inevitably consume more compute. The system will therefore have to impose bureaucrats and roadblocks to slow the compute usage down. Thus, solving a hard problem is a wrestle for compute budget from the system.
English
10
0
114
6.2K
Less
Less@lessaxwrong·
@SpaceKoala Introspection is a broad category, you can wallow in misery inside introspection or you can figure out what went wrong constructively but they are both sub categories
English
1
0
2
48
Space Koala
Space Koala@SpaceKoala·
Introspection is not obsession over how bad things are. It's looking inward to see why you did and said the things you did. You do it not because you want to wallow in self blame, but because you want to grow as a person and attain your full potential. People, like rockets, sometimes have to make changes to reach their full capability.
English
1
0
6
135
Less
Less@lessaxwrong·
@teortaxesTex Reminder that this is a 500 B parameter model
English
0
0
0
79
Teortaxes▶️ (DeepSeek 推特🐋铁粉 2023 – ∞)
CritPt update. Grok 4.20 scores 6.0%. 2x better than DeepSeek V3.2 and almost on par with Speciale. This is massive progress for xAI. Here you can see the best result from ≈every relevant lab. What a beautiful, depressing power law.
Teortaxes▶️ (DeepSeek 推特🐋铁粉 2023 – ∞) tweet media
English
12
7
135
14.4K
Less
Less@lessaxwrong·
@ChadMoran Also , it is worse than tesla many years ago x.com/scobleizer/sta…
Robert Scoble@Scobleizer

I will talk about this on @wholemars’s Subscriber-only space on Sunday. Truth is my eight year-old Tesla was smoother than the Mercedes running NVIDIA software. But seeing how the @NVIDIA AI decomposes the road into all its parts shows me how it will handle all the weird stuff that happens on the road. And we’ll quickly get smoother. Tesla's advantage is the fleet. On Omar's Space on Sunday, I'm going to explain why the fleet is going to bring major advantages to Tesla. Until a car company or a group of car companies can get enough vehicles on the road with an NVIDIA computer, they're not going to be able to compete on experience in the car.

English
0
0
0
15
Less
Less@lessaxwrong·
@ChadMoran Fleet advantage is massive especially as real problems that happen at once every 500000 miles will basically be impossible to detect meaningfully with small fleets. Cybercab is cheap and it’s payback rate will be very fast, payback period is very important. 1/2
English
2
0
0
78
Chad Moran
Chad Moran@ChadMoran·
Second mover advantage is a huge deal. All of the Tesla Uber Bulls who thinks they have a 19 year lead need to take a look at history. There is no moat.
Robert Scoble@Scobleizer

Just had a ride in the NVIDIA autonomous vehicle. I could argue that the Tesla is slightly smoother, but that is missing the point. The point I learned this week by hanging out with a bunch of different companies building autonomous vehicles is that AI at a variety of different companies is evolving so quickly that within 18 months you will see a bunch of different ones shipping and moving from level two (gotta pay attention) to level four (no human needed). Now the narrative will switch to one of scale. Uber went from an idea in a Paris snow storm to an international company driving people who live in slums in South Africa in four years. My expectation is that AVs will reach the slum in less time. Will it be a USA based company like Tesla? Or a Chinese company that reaches the slum (hence meaning it will reach at least 25% of humans on earth in between) first? We now will switch to one of convincing people AVs are a better way to get around than having humans drive. Regulation. Liability. Butts in seats. Experience. Price. And to all the engineers who worked in these things for 20 years: thank you. Your efforts will save millions of lives so my kids won’t have to go to more funerals when they are my age. Seven years ago @IrenaCronin and I wrote it is coming in our book. It took a while but it is finally here. Now the hard work of convincing every naysayer and hater is here.

English
37
3
92
17.9K
Less
Less@lessaxwrong·
@ChadMoran Fleet advantage allowing easier detections also allowing much safer models basically allows them to get to extremely good safety numbers, say 1 in 50 million or 100 million, humans will prefer something that is extremely safe over something reasonably safe
English
0
0
0
10