>And the game is even on an off-the-shelf game engine, they possibly don't even employ game engine experts at Embark Studios.
Perhaps, but they also turned off Nanite, Lumen and virtual shadow maps. I'm not a UE5 hater but using its main features does currently come at a cost. I think these issues will eventually be fixed in newer versions and with better hardware, and at that point Nanite and VSM will become a no-brainer as they do solve real problems in game development.
I think this almost has to be the future if most compute development goes to AI in the next decade or so, beyond the fact that the proposed API is much cleaner. Vendors will stop caring about maintaining complex fixed function hardware and drivers for increasingly complex graphics APIs when they can get 3x the return from AI without losing any potential sales, especially in the current day where compute seems to be more supply limited. Game engines can (and I assume already do) benefit from general purpose compute anyway for things like physics, and even for things that it wouldn't matter in itself for performance or would be slower, doing more on the GPU can be faster if your data is already on the GPU, which becomes more true the more things are done on the GPU. And as the author says, it would be great to have an open source equivalent to CUDA's ecosystem that could be leveraged by games in a cross platform way.
>subsequent removal to obtain the supposed "original, non-derived" form
Also called a "back-formation". FWIF I don't think the existence of corrupted words automatically justifies more corruptions nor does the fact that it is a corruption automatically invalidate it. When language among a group evolves, everyone speaking that language is affected, which is why written language reads pretty differently looking back every 50 years or so, in both formal and informal writing. Therefore language changes should have buy-in from all users.
Yeah, I saw this critique show up a few months ago and now I'm seeing it everywhere, even in major financial news sites like Bloomberg.[0] It's certainly worth discussing, but people are taking it as a gotcha to prove the AI boom is fake. However, all the AI companies have to buy from Nvidia anyway. And Nvidia has tons of cash, in fact it has 4x the cash on hand now than it did in 2023, despite all the investments.[1] So yes, if they think the AI market will grow then of course they will buy into it. If all of Nvidia's deals went bad, their stock would plummet, but not because they lost a few tens of billions, rather because that would mean the AI market is going down in general. There is a great counterexample to the "AI is propped up by circular funding" argument in Google, which uses its own TPUs and builds its own AI, and integrates it into it's own end-user products, no circular deals needed. If AI is propped up by anything it is investors and companies thinking it will give them a huge return. Circular deals are a result of that: cash is going everywhere into that market, it's that simple. The AI boom may be a bubble, but not due to circular deals in particular.
These OCR improvements will almost certainly be brought to google books, which is great. Long term it can enable compressing all non-digital rare books into a manageable size that can be stored for less than $5,000.[0] It would also be great for archive.org to move to this from Tesseract. I wonder what the cost would be, both in raw cost to run, and via a paid API, to do that.
Not always, you can improve the loop by putting something real inside, like, a code execution tool, a search engine, a human, other AIs or an API. As long as the model can make use of that external environment its data can improve. By the same logic a human isolated from other humans for a long time might also be in a situation of going crazy.
Practical example - using LLMs to create deep research reports. It pulls over 500 sources into a complex analysis, and after all that compiling and contrasting it generates an article with references, like a wiki page. That text is probably superior to most of its sources in quality. It does not trust any one source completely, it does not even pretend to present the truth, it only summarizes the distribution of information it found on the topic. Imagine scaling wikipedia 1000x by deep-reporting every conceivable topic.
>what's really funny here is how absolutely horrified people are by the suggestion a single company which has a monopoly shouldn't also define the web platform
They don't. In general browser specs are defined via various standards groups like WHATWG. As far as I know there is no standard for what image formats must be supported on a web browser,[0] which is why in this one case any browser can decide to support an image format or not.
>They barely have their toes in the middleware game anymore.
Well they do have Steam Audio but yeah I agree. I think Epic is much better in this space, even though its only source available in practice they do a lot to support engine modifications and also accept external PRs. I think Valve has a lot to gain from open sourcing Source 2 and they should realize how important modding was to their initial success. The issue is now they can just print money with Steam so there is no need to invest in modding support.
The weighting of this study is strange. The difference of number of years of education maxes out at 1 point, while being raised in different locations and different school types are each given also 1 point. It seems unreasonable that going to school in London vs New York should be given a point here despite the average educational quality in both cities potentially being the same. This also means that someone with 4 years more education but from the same city is considered educationally similar, and it is impossible to achieve the "educationally dissimilar" metric (ED DIFF > 2) without one of the other two points. I feel therefore like there is some wordplay being done here by the term "educational differences." I think some readers will assume that "educational differences" means "educational quality," but only one metric out of the 3 is directly correlated to this. This said there does seem to be some correlation and that is interesting, as we would expect no difference between location, yet there does seem to be one. In my opinion the different location variable is likely to be measuring something aside from/in addition to education. Some education types would seem to be better than others eg. boarding school. Also worth noting that the "very educationally dissimilar" group is n = 10. This said, the authors do admit that "certain level of inference is involved with comparing pedagogies and curriculum" and "Readers are encouraged to re-score and re-analyze the data in additional ways not done here." I would try weighting location much less and not cap number of years of education at all, instead studying how the differences change as the number of years increases.
Because commenters outside Japan may end up buying products containing chips made in Japan. If it was built in let's say France people would be thinking less about potential invasions. Just as "obviously Japan is going to want to develop lucrative manufacturing within Japan," obviously people outside of Japan are going to want manufacturing that is not liable to be shut down or taken over in some way. Not that I think Japan and China will actually go to war any time soon myself.
>geopolitical narrative fed to them by the US state department
Just this week Japan and China have been getting into a fight over the current PM's comments over Taiwan. China has canceled some flights to Japan and complained to the UN, announcing it will defend itself from Japan.[0][1] I'm not sure what point you are trying to make here. Are you saying major disputes between China and Japan don't exist and are invented by the US state department? Or that thinking about it in this context is the result of the commenters being fed by the US state department?
The PRC and Japan is not a remotely comparable situation to the PRC and Taiwan.
The most the PRC could do is potentially sabotage production in Hokkaido, but if they can sabotage production in Hokkaido, they can sabotage it in Arizona.
I don't think China wants to go to war with Japan. I just mean to explain why people are focusing on geopolitical tensions. And the answer is that those tensions do exist, and is partly why some countries are trying to become more self-sufficient to begin with. So discussion of it is valid, that is my main point. Now once we get into those discussions, they might not be as high quality or informed as in let's say a pure technology article, but that is to be expected.
More like a new plant in Iceland, after the PM of Iceland said any attack on Greenland would be a survival-threatening situation for Iceland.
To be clear I think the comments about "geopolitical stability" or whatever term we use are not as interesting as new chip plants itself. Or at least they are a bit tired by now. I also wish Japan the best and I think they are fully capable of building such a factory and I hope they do so. But to claim that the geopolitical considerations are invented is wrong. And in fact one of the reasons the Japanese government is investing in local fabs to begin with is due to national security, as mentioned in the article:
>Securing control over chip manufacturing is being seen as a national security priority, both in Japan and elsewhere, as recent trade frictions and geopolitical tensions between China and Taiwan raise concerns around the risks of relying on foreign suppliers.
So yes, viewing the entire story through a geopolitical lens is understandable.
The built in Steam DRM is very weak. Of course that can change at any time, but at least the current catalog of Steam DRM-only games are not really tied down to steam except via law/licensing.
Perhaps, but they also turned off Nanite, Lumen and virtual shadow maps. I'm not a UE5 hater but using its main features does currently come at a cost. I think these issues will eventually be fixed in newer versions and with better hardware, and at that point Nanite and VSM will become a no-brainer as they do solve real problems in game development.