trotsky

116 posts

trotsky

trotsky

@LeonHowqua

Bergabung Eylül 2023
73 Mengikuti3 Pengikut
trotsky
trotsky@LeonHowqua·
@extliqprovider @DevelopmentsAI @teortaxesTex Yes. Math, science and code skills follows automatically from model capability. Improve the ability to further scale pretaining, and posttrain/RL/scaffold better reasoning, will lead to higher model capabilities. Better model propose more optimisations, and you get a flywheel
English
0
0
1
21
Teortaxes▶️ (DeepSeek 推特🐋铁粉 2023 – ∞)
Teortaxes▶️ (DeepSeek 推特🐋铁粉 2023 – ∞)@teortaxesTex·
I really don't think OpenAI is going to let this slide. I've been saying it for a long time, the real inflection was when they reached 5.2. I have no clear insight on what they currently have internally, but if they haven't made a Mythos/Fable yet, it was *a choice*.
Andrew Curran@AndrewCurran_

The internal boost from Mythos-assisted development since February is just too big. Anthropic is pulling away from the pack for the first time, and at the same time they are also speeding up. The race legitimately feels like it is changing for the first time in years.

English
17
9
452
46K
trotsky
trotsky@LeonHowqua·
@extliqprovider @DevelopmentsAI @teortaxesTex At this current point in time, no fancy new math/science is really needed to improve the LLMs. It's just more efficient training code, architectural experiments, scaffolding, data generation, RL environment building etc, which is achieved with better coding capabilities
English
1
0
2
47
efe
efe@extliqprovider·
@LeonHowqua @DevelopmentsAI @teortaxesTex isnt coding just a tool to implement your research ideas? how can rsi be achieved if this model is only good at coding and mid tier at maths/science/etc? not assuming mythos is bad at math but coding is just one vertical
English
1
0
0
54
trotsky
trotsky@LeonHowqua·
@DevelopmentsAI @teortaxesTex Recursive self improvement, that's how take-off happens. For now seems like coding is the way for that to happen
English
1
0
0
95
Latest Developments in AI
Latest Developments in AI@DevelopmentsAI·
@teortaxesTex OpenAI and Gemini models are stronger at math. Maybe it will turn out that math is more important than coding.
English
1
0
0
1.5K
trotsky
trotsky@LeonHowqua·
@Meta_Logos_35 Not sure about neurological, but phenotypical definitely not. There's higher phenotypical diversity on the mainland
English
1
0
0
69
Patrick Eizo Omori (大森英三)
Patrick Eizo Omori (大森英三)@Meta_Logos_35·
Japanese are the most high variance East Asians: Japanese people have more neurological and psychological variance than continental East Asians More phenotypical diversity than continental East Asians Greater sexual dimorphism than mainland East Asians
English
5
4
19
1.5K
trotsky
trotsky@LeonHowqua·
@robertwiblin @akrugs94 Relax the Nvidia, TSMC, ASML etc. bans, and we'll see. You can't expect them to win the boat race when you ban them from buying boat engines
English
0
0
0
23
Rob Wiblin
Rob Wiblin@robertwiblin·
@akrugs94 Because they're in large part improving their models by copying/distilling US models. If the US models stop improving, progress on Chinese models will also slow down greatly.
English
7
1
15
2.1K
Rob Wiblin
Rob Wiblin@robertwiblin·
I predict that the water skier will not successfully overtake the boat even if the boat slows down.
Rob Wiblin tweet media
English
28
9
261
36.7K
trotsky
trotsky@LeonHowqua·
@teortaxesTex Is that good news for Deepseek then? I think there were news that Ascend chips are not good at training but quite usable for inference. If inference scaling is (almost) all you need, then that's good news for the DS-HW tech stack
English
0
0
3
235
trotsky
trotsky@LeonHowqua·
@dsonoiki Well the art director is taking a risk now on her reputation to get higher payoffs
English
0
0
0
477
Gene Parmesan
Gene Parmesan@dsonoiki·
this attitude explains a lot of the Bernie style attitudes towards wealth they simply do not understand the concept of risk these folks wouldn’t agree to ONLY get paid if the movie makes a profit they wouldn’t agree to some sort of clawback where they have to pay back some of their income if the movie flops producers put capital at risk and it paid off these same producers will make other movies and some of those movies will lose money or barely turn a profit that said, it would be a nice gesture to pay a bonus of sorts like how Taylor Swift paid bonuses to her tour staff but we can’t all be expected to be Taylor Swift there’s only one Taylor Swift
That COMICBOOK Guy@Culture3ase

‘OBSESSION’ Art Director calls for industry reform after getting paid less than $7k as film passes $200 Million globally, becoming Focus Features’ Top movie of all time.

English
86
124
2.6K
203.5K
trotsky
trotsky@LeonHowqua·
@GreenwoodNoddy @balababada 很显然这套（严格的企业内部尊卑体系）不管在中国，台湾还是新马的华人社会都并不常见，是日韩特有的。不排除韩国企业文化本身源自日本影响的可能，朴正熙时代官僚和财阀很多都有殖民政府、关东军等背景
中文
0
0
1
216
Noddy
Noddy@GreenwoodNoddy·
@balababada 這不就是中國傳統文化嘛，長幼有序是指內部，外部看地位
中文
3
0
18
16.1K
trotsky
trotsky@LeonHowqua·
@deanwball Because they understand ultimately the current US lead over China in AI is policy supported. Specifically US policies to block access to ASML EUVs. There's no secret sauce other than massive compute
English
0
0
0
25
Dean W. Ball
Dean W. Ball@deanwball·
The salient thing is that nobody, absolutely nobody, in the Washington DC strategic class describes “capitalism” as existing in competition with the Chinese system. It is always “democracy.” They think capitalism already lost and that we have to become like China, with state-led public/private enterprise (“strategy”). So the word we use to describe, waves hands, “our way of life vis a vis their way of life” is “democracy.” Their defense is of a political order. On average they seem to feel no particular attachment to the economic order. This is bipartisan and is why the feeling of many in DC was “of course the government can tell anthropic to do whatever they want!” Civil-military fusion, the death of the private and its assumption into politics, they ultimately cheer these things on. They love them. They only decry them when social media is framed as the driving agent behind these trends (eg “polarization” discourse). Then it is bad, but that’s just because that framing makes them feel not in charge.
English
5
5
147
10.7K
Dean W. Ball
Dean W. Ball@deanwball·
I think part of it, at least vis a vis US/China competition, is that US and western chattering classes find it hard to believe that the market-driven outcome of frontier AI could possibly be right. They basically believe, in their hearts, that the Chinese system, with its “industrial strategy,” has eclipsed capitalism. So they harbor the same inferiority complex toward the Chinese system that many Americans once harbored toward the EU’s system. Their heuristic is that the industrial strategists of China have grasped the whole picture of the technological competition in a way that US industrialists, with their “profit maximizing incentives,” could not possibly have matched. And so any outcome in the economy that is not the result of “strategy” is therefore prima facie worse than what the “strategists” have concocted. They also believe the Chinese strategists possess awesome powers of foresight and the ability to evade all tendencies of financial and economic gravity, due of course to “strategy,” really it’s almost a kind of orientalism. Meanwhile the U.S. industrialists are making new advances in math and science, and the fastest-growing businesses in history, by spending hundreds of billions of dollars on high-margin chips whose legacy is in rendering video games, cramming them underneath tents if need be, and investing generational capital into new energy generation technologies as they do it, and perhaps even colonizing space as an instrumentally convergent result. But none of that is “strategy,” you see.
Dean W. Ball@deanwball

I find it so interesting how persistently unable the strategic classes of free society are to analyze AI well. So many keep getting stuck in these basins of delusion. I was at a conference where it was not just asserted but taken for granted that Chinese models have dominant global inference market share. The 2024/early 25 version of the delusion was “mode collapse/data wall” (even after reasoning models!), then it was “AI is plateauing and a bubble” for most of 2025, now it’s “Chinese OSS is good enough.” The share of people in the strategic classes who think this is gradually declining, but it is still sufficiently common that you can attend a prestigious conference and encounter a room principally filled with basin-dwellers.

English
53
89
721
353.6K
trotsky
trotsky@LeonHowqua·
@lianda_edu Well, is the statement true? If it is not true explain that. If it's true (which I expect it is), it's the same as many other descriptive social science where models help us understand the world but is not directly applicable to policy or business.
English
0
0
0
19
Hayashi Heikichi
Hayashi Heikichi@lianda_edu·
A physicist friend mocked economics like this: "Imagine a simple counterfactual. If the knowledge of Heisenberg's matrix mechanics and Schrödinger's wave mechanics no longer existed, human civilization would suffer a massive regression. But if Arrow-Debreu general equilibrium theory disappeared tomorrow, there would be almost no material loss to humanity."How should I respond to this sarcastic jab?
English
1
0
4
572
trotsky
trotsky@LeonHowqua·
@tekbog 0 instincts for self preservation. The first slap nearly missed, and my little guy here looks visibly scared but still doesn't move out of the way before the inevitable next slap comes
English
0
0
1
20
Jawwwn
Jawwwn@jawwwn_·
Many people forget that @elonmusk is a self-taught aerospace engineer. Elon explains why he thought he was qualified to be the CTO of SpaceX: “I read a lot of books, and talked to a lot of smart people.” Via @60Minutes
Jawwwn@jawwwn_

.@elonmusk on SpaceX’s IPO in 2012: “SpaceX will go public at some point, as I think it should ultimately be owned primarily by the public.”

English
244
594
8.2K
1.1M
trotsky
trotsky@LeonHowqua·
@Noahpinion What I've argued before on Lesswrong against the so called FOOM scenario. Physical world operate at rates limited by physics, biology and social structures. AI even if superintelligent can't just FOOM its ways out of bottlenecks in an afternoon
English
0
0
1
223
trotsky
trotsky@LeonHowqua·
@racistbilbo Fuck the CCP, but wrong is wrong Gandalf, they teach you that in Wizard school
English
1
0
0
24
gandalf the racist
gandalf the racist@racistbilbo·
@LeonHowqua Im not going to argue the point with you because your entire account exists to argue the point. I am right and no amount of CCP shill propaganda will change that. Stop spending money on online propaganda, and build some fucking infrastructure for your people instead.
English
1
0
1
31
gandalf the racist
gandalf the racist@racistbilbo·
lol I never even got a single notification for the dozens of chinese and idiotic western tankies that decided to comment on this post. Im guessing they reported it into oblivion. What I said is 100% true despite what some chinese propaganda "person" will tell you. China bribed the UN and the WHO around 2018 to stop reporting on this 100% true, highly embarrassing, fact. When they say "running water" now, they mean the village has a hand pump that technically provides water, or a giant water tank that people can siphon from. Its not a definition of running water that would be recognizable to a westerner, which is clean, potable water INSIDE the house, available via the twist of a tap. Just look at that house. You think someone retrofitted that place with PEX and put a toilet and a sink in there?? Come on lol Outside their potemkin villages, which even ill admit, they have quite a few now, this is still what most of china still looks like and how most chinese live.
gandalf the racist@racistbilbo

An interesting fact is that even though china is now the 2nd largest economy in the world and has many large, technologically impressive cities, 50% of all chinese still live in homes without running water. You know what percentage of americans dont have running water? 0.2%.

English
2
2
32
943
trotsky
trotsky@LeonHowqua·
@racistbilbo If you said 30%, that's still wrong but a more defendable assumption using urbanization rates as proxy
English
1
0
0
31
trotsky
trotsky@LeonHowqua·
@racistbilbo LoL imagining getting this worked up. Your post came up on my timeline because X algo works that way, some rando Chinese accounts replied to you and I replied to rando Chinese accounts before, therefore the algo thought that I might be interested. Anws, 50% claim is still wrong.
English
2
0
0
40
trotsky
trotsky@LeonHowqua·
@Web3Educator_ @AFpost > No clean environment , burden about taxes , no job security, no work - life balance. None of this is recent, it has always been this way in the past 2 decades. Thus it's not the main driver of the fall in fertility
English
0
0
0
104
Dear_Myself
Dear_Myself@Web3Educator_·
@AFpost No clean environment , burden about taxes , no job security, no work - life balance. Meanwhile the government announces 10k,20k for 3rd & 4th babies. We're paying taxes and you guys are giving back to us. Use it for developments without corruption and commission.
English
2
2
93
15.2K
AF Post
AF Post@AFpost·
India’s fertility rate has fallen below replacement for the first time in the country’s history, declining from a TFR of 2.3 to 1.9 in just a decade. Delhi’s fertility rate now sits at 1.2, lower than Finland’s. Follow: @AFpost
AF Post tweet mediaAF Post tweet media
English
1.8K
2.4K
14.3K
27.1M
trotsky
trotsky@LeonHowqua·
@Porkchop_EXP The people discussing embryo selection and the people discussing pro-life are totally different and non-overlapping demographics
English
0
0
0
25
Porkchop Express
Porkchop Express@Porkchop_EXP·
Just two weeks ago this platform was vigorously but calmly discussing purely the utility of selecting an embryo with a supposed IQ in the 98th percentile, and there was barely any outrage at the fact that the graph showed the couple creating more than a dozen embryos for pure eugenic selection purposes, but if someone decides to terminate one embryo with a demonstrated severe disability all hell breaks lose (probably the same people).
English
2
0
55
4K
Porkchop Express
Porkchop Express@Porkchop_EXP·
American society is absolutely schizophrenic: You are allowed to create a dozen healthy embryos only to screen them for something like eye color or IQ and kill the rest (actual eugenics), while another huge part of the population equates early stage abortion in cases of severe disability to murder. In Europe the near-universal ethical consensus is that screening and abortion in severe genetic disease or disability cases is allowed, while also keeping strict restrictions on any kind of eugenics selections starting with even gender. The slippery slope is not so slippery.
English
38
27
441
25.1K

Jelajahi

@extliqprovider @DevelopmentsAI @teortaxesTex @Meta_Logos_35 @robertwiblin @akrugs94 @dsonoiki @GreenwoodNoddy