English
Steve Chu
1.2K posts
@jaredrg12 @ProudSocialist Fastest growth in wealth for top 1% occurred under Obama. Second faster occurred under Biden. China has more wealth disparity than US. The record is clear, voting socialists creates disparity en masse, ignorant bitch.
English
70% of Americans are struggling to afford groceries, housing, and healthcare, and there is a genocide in Gaza and Lebanon.
Fuck the Met Gala.
English
🇺🇸 An American young man:
I am now in Japan. And when I say that I am American, people look at me with contempt and disgust. And wherever I go in the world and say that I am from America, people look at me with hatred.
We are now in a state of humiliation. This is disgusting and shameful, simply.
English
Deepseek v4 pro (high) scores 46.5% on WeirdML, which is way lower than I'd expected. It does not exceed deepseek 3.2 speciale, although it does match it with far fewer tokens.
I ran this through fireworks AI, and the outputs look reasonable, but the results being this weak makes me suspect something is wrong.
I have a separate run with "max" reasoning setting ongoing right now.
Håvard Ihle@htihle
WeirdML v2 is now out! The update includes a bunch of new tasks (now 19 tasks total, up from 6), and results from all the latest models. We now also track api costs and other metadata which give more insight into the different models. The new results are shown in these two figures. The first one shows an overview of the overall results as well as the results on individual tasks, in addition to various metadata. The second figure shows cost vs performance and shows a clear scaling with better results for higher costs. We also have a very varied pareto frontier with 11 models from 6 different companies having the best accuracy for a given cost for at least some of the cost range. Grok 3, Claude Opus 4 and GPT 4.5 are the ones that underperform for their costs, while Gemini pro and o3 pro have the best results at the highest costs. Qwen3 30B3A, grok 3 mini and deepseek R1 also each represent a good chunk of the pareto frontier.
English
@alex_avoigt I have a few fucking water bottles in my trunk, enough for 1000km. lol Cope
English
The BMW iX3 and iX are examples of how little the company still understands about BEVs.
A very large battery means that in 99.99% of cases, when long distances aren't being driven, an unnecessary amount of weight is being carried around, which is inefficient and expensive.
Since a large battery naturally takes longer to charge, even with fast charging speeds, the supposedly advantageous feature of a long range is worthless, because you simply end up spending more time charging and if you take shorter charging stops you could have been more efficient with a smaller battery.
To everyone claiming they want to drive 1,000 km without a break: studies show that people with such a strong bladder are rare. And if your family is coming along, forget it. Such cases are so rare that it makes no sense to develop a vehicle for them.
Both vehicles are fundamentally flawed in their design and are only suitable for people who, as ICE owners, have been falsely instilled with fear for years that BEVs supposedly don't have enough range or that there aren't enough charging stations—both of which are lies perpetuated by the oil and gas industry.
English
@jbulltard1 That fucking dick and all his cult are fucking shilling for Space fucking X
English
@moonsewage @jbulltard1 I can fucking sleep on the floor. Btw that fucking dick is buying a 300 million house in Florida. You are dumb AF. Sucker.
English
@jbulltard1 "For himself" is extremely misleading. He sleeps on the floor of his companies, rarely takes a holiday, and his companies generally pursue goals to unlock abilities he sees as essential for humanity. Whether one agrees with them is irrelevant. More money is to attain them faster.
English
Elon Musk called these people "the best analysts in the world".
They are projecting that robotaxi will have 6x more revenue the entire world ride-hailing, taxi, and mass transit markets combined.
Their assumption is that it will not only swallow all those markets but somehow also increase them by 6x.
Mind you that these "best analysts in the world" also projected just a few years ago that Tesla would deliver 18 million vehicles this year.
Tesla would be lucky to deliver 2 million vehicles this year.
There's no other way to say this. These people are either paid shills or morons. No way around it.
ARK Funds@ARK_Funds
The robotaxi market could hit $34 trillion by 2030. Who captures most of that value? @TashaARK outlines ride-hail economics in our Q1 webinar. Watch now!
English
Breaking news: Japanese Prime Minister Sanae Takasugi kneels at the Nanjing Massacre Memorial Hall... oh, no, she's in Australia.
During WWII, Japan never actually landed in Australia. Aside from some prisoner of war atrocities, Japan committed relatively few war crimes against Australia—nothing compared to the horrors inflicted upon China, Korea, and Southeast Asia. Yet, a Japanese prime minister has never knelt in Korea, China, or Southeast Asia nations, instead, Japan has a lot of historical grievances with these countries till today. Absurd enough, right?
It is obviously because Australia is a white country and the Japanese have been completely tamed by white people. Japanese is a nation that has been conquered from men's cerebral cortex to women's womb. This self deprecating race of course has no future.
English
@Jewish_Story @mokhanhim Fucking jew, fucking dickhead, fucking ass licker, fucking gay
English
@mokhanhim Sounds like he succeeded. Your actions had consequences. Welcome to adulthood.
English
1 year ago today, Dave Portnoy forced me to go to Auschwitz and apologize to the jews
I refused — and Dave tried to ruin my life
I was expelled from university, my post-grad job offer was revoked, I lost my “best friends”, was removed from my fraternity, received death threats, had my house broken into, and even got banned on Cash-app and PayPal
These are the scare tactics jews will use to frighten you into submission
Never bend the knee, never apologize, and always double down
1 year later, I don’t regret a thing.
English
@SputnikInt This post is misinformed. The refinery was build in conjunction with India not China!
Thank you for your attention to this matter.
English
🇨🇳🇳🇬 China’s industrial muscle turns Nigeria into net exporter of refined petroleum
👉 For decades, Africa’s biggest oil producer couldn’t refine its own fuel — until China helped flip the script.
💰 The catalyst was the $20 billion Dangote Petroleum Refinery, built leaning on Chinese engineering, procurement, and construction contractors.
🟠 Operating at ≈ 94% of its 650,000 barrels-per-day capacity, the Lagos-based refinery covers domestic demand, with excess sent abroad
🟠 Roughly 44,000 barrels of gasoline per day were exported in March alone; a single shipment of 317,000 barrels reached Mozambique—its first delivery to East Africa (historically supplied by the Middle East)
🟠 Output is projected to hit 1.4 million barrels per day within three years, which would make it the largest single-site refinery globally
English
@SamRo If he had held the S and P 500 index in the same period he would actually have done better. Quite making the BRK boys into some sort of financial genius.
English
@jimstewartson Then I guess it’s on the involved people to do good? He’s just a figurehead but this is some doomer stuff trying to get engagement. The sun will destroy earth, we do have to leave eventually. Do the math or shut up.
English
No matter how insane you think the SpaceX IPO is, it’s so much worse.
Jim Stewartson, Decelerationist 🇨🇦🇺🇦🇺🇸@jimstewartson
The SpaceX IPO isn’t a normal company going public; it’s the birth of a “network state”—a corporate dictatorship devoted to “exit.” Musk’s $2 trillion fraud is fascist science fiction to con gullible investors into paying for a financial führerbunker. mind-war.com/p/birth-of-a-n…
English
@airkatakana income inequality in korea is lower than in japan now.
big companies and strong labor unions working well.
eg, sk hynix union getting workers average worker bonus of ~us$1mm this year.
English
when your country has 4 companies which make up 40% of your gdp, metrics like gdp/capita become completely divorced from the situation on the ground
100 people are living in +7sd levels of abundance while everyone else is wondering if they can afford starbucks or not
English
@airkatakana But this is a super gay comparison. Their turnover is not 40 % of gdp. It is their market caps that are. While GDP basically is a countries turnover.
English
AC does not bring in outside air
It just rotates existing air in your house and cools or heats it
Magnus@sungamma
@levelsio aircon?
English
A lot of people objected to, were even outraged, that restricting the revenue analysis to American software workers.
But let’s think about how much Anthropic revenue could be international. Excluding China and Russia (Claude not available), the US makes up about 20% of global devs.
But one thing to keep in mind is, to get to $44 billion ARR, the only thing that really matters is devs paying $1k/month. Even Max subscribers don’t move the needle at that scale.
Realistically, devs outside America aren’t expensive enough for this to make financial sense. Median software engineer salary in India (largest international dev market outside US and China) is $30k, is it reasonable believe that Indian firms are casually accepting a 50% cost hike on their software engineers to give them API scale token budgets? Again I believe many, if not most Indian devs are using Claude. But (like myself) my guess is the vast majority are using the much more cost effective subscription and managing limits.
Same story in most other countries. Median software engineer salary in Europe is $60k, in Japan $50k, in Brazil $30k, in Britain $65k . I’m not saying there is zero API usage in these countries, but I think it’s unrealistic to think these countries have anywhere near the casual API spend that US software engineers have.
I think a conservative upper bound on international API revenue is maybe 50%. If we go back to US devs being 20% of the global market, then assuming they have 2X Claude spend intensity seems reasonable.
So at most, the original analysis has a 2X on the denominator. Doesn’t really fundamentally change anything in the order of magnitudes. To the extent software work is driving the Anthropic ARR, either near majorities of devs must be spending thousands per month, or there must be significant super users who are blowing through tens or even hundreds of thousand per month.
Doug Colkitt@0xdoug
I’m really struggling to see how the back of the envelope math on this works out… There are generously 4 million characterized “software workers” in America. That’s pretty broad and includes a lot of people who aren’t really classical engineers don’t produce that much code. That comes out to nearly $1k per month of average Claude spend across every dev in America. Yes, there’s some international usage, but it can’t be that much. Yes there is some non software Cowork usage, but that doesn’t use that many tokens. Yes, some non engineers are using Claude to vibe code, but I really doubt many are spending hundreds per month on. Even if we assume 50% of all software workers are using Claude, that comes out to $2k spend per month per Claude user. Thats 10X more than the highest tier Max subscription. So almost all of Anthropics revenue has to be API billing So the only explanation is that something like 20%+ of software engineers are not only Claude users but on API billing and regularly spending thousands per month. At $5/m Opus tokens that means the average API user has to be going through something like 25 million tokens per day. *OR* the other possibility is API revenue is heavily power law dominated. Maybe there’s just something like 100k super users who are making up the majority of the revenue. For that to work the typical super user would have to be spending on the order of $50k/month and guzzling nearly 1 billion tokens per day.
English
@Iam_No_One____ @TheAhmadOsman It means “what I have to say is more important than what anyone else has to say”. See @naval
English
@_ueaj @alkimiadev @scaling01 Anthropic inference stack's margin is 70% and keeps increasing (per SemiAnalysis report)
English
@alkimiadev @scaling01 we don't know the closed source cost is what I'm saying, so we don't know what the closed source cost/performance frontier actually is and thus we don't know the algorithmic gap
English
chinese models are ~8 months behind and are falling further behind
Séb Krier@sebkrier
DeepSeek V4’s capability lags behind leading U.S. models by about 8 months. nist.gov/news-events/ne…
English