- You could also try to replicate something like the Monster 6502: https://monster6502.com/
It's not lithography, but you can build a working processor out of small surface mount chips, and you can solder these chips with lead-free solder. That seems very achievable for a motivated engineer, and probably involves much less toxic chemicals?
- It's probably true that free enterprise helps a lot, but China has that in large part. Even though the CCP calls itself communist, China is very capitalist in a number of ways. But I guess China is showing us that capitalism can exist without democracy.
- "Windows 3.1 primarily used palette-based color modes, common modes included 16 colors (VGA/EGA) and 256 colors (SuperVGA)"
- China has their own fabs. They are behind TSMC in terms of technology, but that doesn't mean they don't have fabs. They're currently ~7nm AFAIK. That's behind TSMC, but also not useless. They are obviously trying hard to catch up. I don't think we should just imagine that they never will. China has a lot of smart engineers and they know how strategically important chip manufacturing is.
This is like this funny idea people had in the early 2000s that China would continue to manufacture most US technology but they could never design their own competitive tech. Why would anyone think that?
Wrt invading Taiwan, I don't think there is any way China can get TSMC intact. If they do invade Taiwan (please God no), it would be a horrible bloodbath. Deaths in the hundreds of thousands and probably relentless bombing. Taiwan would likely destroy its own fabs to avoid them being taken. It would be sad and horrible.
- I had a 386 PC with 4MB of RAM when I was a kid, and it ran Windows 3.1 with a GUI, but that also had a VGA display at 640x480, and only 16-bit color (4 bits per pixel). So 153,600 bytes for the frame buffer.
- The reason NaN exists is for performance AFAIK. i.e. on a GPU you can't really have exceptions. You don't want to be constantly checking "did this individual floating-point op produce an error?" It's easier and faster for the individual floating point unit to flag the output as a NaN. Obviously NaNs long predate GPUs, but floating-point support was also hardware accelerated in a variety of ways for a long time.
That being said, I agree that the way NaNs propagate is messy. You can end up only finding out that there was an error much later during the program's execution and then it can be tricky to find out where it came from.
- You can buy old rackmount servers on ebay for relatively cheap and used them as desktop PCs AFAIK (though they can be on the noisier side).
- I've had a similar experience with Gemini ignoring things I've explicitly told it (sometimes more than once). It's probably context rot. LLM give you a huge advertised number of tokens in the context, but the more stuff you put in there, the less reliably it remembers everything, which makes sense given how transformer attention blocks work internally.
- I think it's because they're all trained on the same data (everything they could possibly scrape from the open web). The models tend to learn some kind of distribution of what is most likely for a given prompt. It tends to produce things that are very average looking, very "likely", but as a result also predictable and unoriginal.
If you want something that looks original, you have to come up with a more original prompt. Or we have to find a way to train these models to sample things that are less likely from their distribution? Find a way to mathematically describe what it means to be original.
- From my experience they quickly fail to understand anything beyond a superficial description of the image you want.
- Also how do you avoid billionaires worldwide? Not everyone lives under your government. Even if you could, how do you know for a fact that some people don't secretly control hidden assets? Is Xi openly a billionaire? China is a "communist" country on paper. How does he hold so much power?
The sad reality is that the world has a nonzero percentage of power-hungry narcissists. We need governments that are more democratic and robust. We all know that the current government processes are broken and corrupted.
- It might also be something like: people who experience relatively little psychological distress are optimistic, and also live longer.
We don't like to talk about it, but there is a biological component to how happy or anxious someone is. Not to say that everyone is doomed to be a certain way because of their genetics, but I've known people who are basically never sad, and it's not because they've found some kind of secret of happiness. I have a friend who has the temperament of a golden retriever.
- 2. New state-funded joint venture: EuroNV, pronounced euro-envy.
- Will definitely be entertaining to watch if it happens.
- I guess it makes sense. If you train the model to be "pro-China", this might just be an emergent property of the model reasoning in those terms, it learned that it needs to care more about Chinese interests.
- Google maps is probably a big moat that's very hard to replicate. You can't as easily just crawl all of that data. It's not easy to generate directions. The average user doesn't want to use your search engine for one thing and Google for everything else, they just want a one stop shop for search.
- Don't hold your breath. Chip fabrication is extremely sophisticated. It's basically nanotechnology. Not something that could practically be done at home anytime soon.
- Genetics do factor. It's not just a question of genetics affecting metabolism. People literally don't feel hunger with the same intensity as one another. It's like sex drive. There are both very horny people and asexuals out there. There are also people who routinely forget to eat. For many people though, the notion of "forgetting to eat" seems completely alien, because those signals are much stronger for them.
- They're very expensive drugs so it would be one lobby against another. However given that they're so expensive, I would think that broke, uneducated alcoholics just won't have access to them, so those profits are safe...
I get that it's tempting to say "we no longer have to program game engines, hurray", but at the same time, we've already done the work, we already have game engines that are relatively very computationally efficient and predictable. We understand graphics and simulation quite well.
Personally: I think there's an obvious future in using AI tools to generate game content. 3D modelling and animation can be very time consuming. If you could get an AI model to generate animated characters, you could save a lot of time. You could also empower a lot of indie devs who don't have 3D modelers to help them. AI tools to generate large maps, also super valuable. Replacing the game engine itself, I think it's a taller order than people realize, and maybe not actually desirable.