All of them really. What's typically open source is
1) the code used for training, but never the dataset for initial LLM and never the RLHF (reinforcement learning with human feedback) data used to make a text vomiting LLM into a useful question answering maching.
2) the resulting weights - these are totally uninterpretable.

So it's never fully replicable; even if you had the infra and were willing to burn electricity you don't have a way of going to 2) yourself.

AFAIK that's not just the Chinese but also open-source / weights Llama and Mistral.

This sounds dismissive, but I wouldn't read it as such.

"China doing something first, however, has never been a reliable indicator that the thing will prove durable, economic, or widely replicable. China is large enough to try almost everything."

This has always been true, you can for example read the history of improving iron / steel production during the Industrial Revolution. Either you had existing outfit and capital to try things. Or you raised capital and set-up an outfit to try things. If you had something viable, you made money, if not, the world (and hopefully you) moved on.

And in the end VCs work more-or-less on the same principle (which is why at some point someone was trying to do Uber for xyz).

This is a PR "thought leadership" BS article by Benjamin Riley, Cognitive Resonance, who "provides direct consulting support to organizations to improve understanding of how generative AI works."

This doesn't mean they're wrong but it's probably nothing terribly original (there is a reason why it's not on openreview.net as a submission into one of the relevant AI conferences).

What I don't get is this: let's say they believe that to get to "genuinely useful AI" (generating value in business, automating science) they need 1024x the compute they have now; which I think is ballpark with their spending.

But Moore's law says that number of transistors doubles every 2 years, so if we believe we can keep this going then in 4 years time they'll only need 256x the data centres and in 8 years it's only going to be 64.

So all the chips they put in will be next to worthless in 8 years and they'll only need a fraction of the other infrastructure. So all this massive spend is to be first because? Because whoever is first will be able to use their "genuinely useful AI" to out-innovate everyone to singularity?

The whole thing is a one-way bet on the empirical LLM scaling laws (https://arxiv.org/abs/2001.08361) scaling to something useful, which I don't buy.

This is great but there is still a long way to go. Renewable capacity is not really comparable to fossil fuel power station capacity because the coal / gas ones can run 24/7...

To get a better picture of where we are check out http://grid.iamkate.com/ . Basically in the last year UK electricity was 19% from renewable sources with fossil fuels at 48%.

It didn't "confirm" it was caused by man, as it didn't set out to and doesn't claim to.

Nevertheless the collected data seem to indicate a steady increase in temperature. This has coincided with increased emissions of CO2 (while many other factors remained constant, or more precisely didn't vary enough to allow anyone to claim correlation). This of course does not mean that it's _caused_ by the increased emissions of CO2.

But if my belly starts aching I look at what I ate that others didn't. And if I ate something that others didn't (say a dodgy kebab) and I feel bad and they don't then of course I can't claim I feel bad because of the kebab. But I'm sure not going to have the same kebab next time. I don't wait for a double blind study done on a statistically significant sample to confirm to within some statistical error that the kebab is indeed bad.

Second that!

You'd think that Italian MEPs would have better things to worry about right now, like not being able to borrow money if it wasn't for ECB and possibly not being able to pay their debts.

But no, that c.nt wants to have a way to let the government to spy on everyone.

The whole article is full of strange statements.

Consider the following part:

My school physics may be a bit rusty but I would assume wind turbines produce either pulsating DC or AC and hence the current has to be converted before use by electronics.

They meant to say it's more reliable, because there are fewer parts inside the computer that could fail.

You could argue in the opposite direction: if the main ACDC converter fails, all the computers go down, not just one. Swings and roundabouts I guess.

I'm very happy that majority of users use IE. This makes it still the most attractive target for hackers. In turn that means that they have less time to work on exploits for the browser I'm using. "Security through obscurity" works in this case (though of course the phrase comes originally from open source vs. closed source).

I can see where your arguments are coming from and I think your opinion makes sense.

However anyone can download the code from their website and compile it on the iphone simulator (which, together with XCode is a free download from Apple). You could argue that you still need a non-free os to run it which won't make RMS happy, but then again on a PC you more often than not have a non free BIOS that's needed to run the OS. And pretty much any computer has a non-free hardware.

Because of this, and other reasons, in my opinion there's nothing wrong with selling a GPL program on the Itunes store as long as anyone who bought it can get the source code.

Think about what sort of programming you would like to do. Remember that writing software is only a way of solving some wider problem, so think what should that be? What would you love doing? To get anywhere as a programmer you have to undersand and enjoy working in the wider field. Very few people end up writing OS kernels or other "programming for the sake of computers". The rest of us make tools for other people (Word processors, painting programs, CAD, banking software, databases, you name it). Think which of these above would you enjoy working on.

Once you've decided on that, choose any language you like, and start working on some small relevant project. Once you have something put the link on your CV. At our company, once we get someone's CV and it looks good, we go to google and find what the guy has done. Publications, software projects etc. Then we decide whether to interview. During the interview we'll talk about that. We don't care whether the person can do Java or C++ if they're graduates. (even though we prefer C++).