What is Pete Davidson doing to that poor soul?

Then please explain to me one simple thing - how do you implement sanctions when they can be circumvented by setting up a single company?

No, it’s not new or strange. It’s a normal component of sanctions, and it’s fundamentally how they’re implemented. Otherwise you could circumvent them by setting up two companies.

It becomes impossible to predict which companies and services may be suddenly impacted.

It’s pretty easy to predict. Do you do business with a sanctioned country? Then you’ll be impacted. Easy enough.

I’m all for the EU sanctions against Russia, and consequences for those entities breaching them. But Microsoft didn’t breach the sanctions, and should be used as a tool to punish those that do.

Are you under the impression that Microsoft is being punished in any way? They aren’t, they’re simply not allowed to do business with companies acting against sanctions if they want to keep doing business in the EU.

I don’t know why you’re acting like this is such a strange thing.

Nayara supplies & operates in a sanctioned country. The EU doesn’t want companies supplying companies that do so. If Microsoft wants to keep operating in the EU, they aren’t allowed to keep supplying companies that do so.

Of course there is an indication that Microsoft was legally obligated to suspend their service in this case:

In this instance, the cutoff was sought by the European Union (EU), in an attempt to pressure Russia to back off its assaults on Ukraine.

If they wish to operate in the EU, they have to follow some of the EU’s demands.

It’s like getting the power company to cut your electricity because you have unpaid parking tickets - It’s probabkly a great way to get parking offenders to pay what they owe, but it undermines trust in general, yes?

It’s more like “getting your accounts frozen because you operate in a country that has sanctions against it”. Which is a totally normal thing to do. Companies cutting off other companies that operate in countries which attack other countries doesn’t undermine my trust - companies continuing to operate in such countries undermines it.

It’s the way it should work. A private company can only be compelled to enforce a government demand under due process of the applicable jurisdiction. Ensures trust through transparency.

They are compelled to enforce a government demand under due process of the applicable jurisdiction. For a multinational corporation, the applicable jurisdiction are all the jurisdictions they operate in. Since multinational corporations exist to funnel profits into their host country, that country has the ability to compel them under due process in other countries.

You might argue that it’s not good for companies to be this large, and I’d agree. You might also argue that specific sanctions aren’t good, and I’d agree. But the idea that a companies ToS should supercede jurisdictions and that they shouldn’t be curtailed by the governments under which they operate is fundamentally corrosive to the concept of statehood.

Sanctions exist to restrict trade with other countries. This can’t work if companies can just ignore sanctions, and I don’t want e.g. european companies to ignore sanctions against Russia.

“Sorry government, I can’t enforce your sanctions, my ToS don’t allow me”

Do you really think this works?

The sky color is part of the training data. How did the LLMs include the training data before it existed?

My god.

There are many parameters that you set before training a new model, one of which (simplified) is the size of the model, or (roughly) the number of neurons. There isn’t any natural lower or upper bound for the size, instead you choose it based on the hardware you want to run the model on.

Now the promise from OpenAI (from their many papers, and press releases, and …) was that we’ll be able to reach AGI by scaling. Part of the reason why Microsoft invested so much money into OpenAI was their promise of far greater capabilities for the models, given enough hardware. Microsoft wanted to build a moat.

Now, through DeepSeek, you can scale even further with that hardware. If Microsoft really thought OpenAI could reach ChatGPT 5, 6 or whatever through scaling, they’d keep the GPUs for themselves to widen their moat.

But they’re not doing that, instead they’re scaling back their investments, even though more advanced models will most likely still use more hardware on average. Don’t forget that there are many players in this field that keep bushing the bounds. If ChatGPT 4.5 is any indication, they’ll have to scale up massively to keep any advantage compared to the market. But they’re not doing that.

But really the “game” is the model. Throwing more hardware at the same model is like throwing more hardware at the same game.

No, it’s not! AI models are supposed to scale. When you throw more hardware at them, they are supposed to develop new abilities. A game doesn’t get a new level because you’re increasing the resolution.

At this point, you either have a fundamental misunderstanding of AI models, or you’re trolling.

I’m supposed to be able to take a model architecture from today, scale it up 100x and get an improvement. I can’t make the settings in Crysis 100x higher than they can go.

Games always have a limit, AI is supposed to get better with scale. Which part do you not understand?

It’s still not a valid comparison. We’re not talking about diminished returns, we’re talking about an actual ceiling. There are only so many options implemented in games - once they’re maxed out, you can’t go higher.

That’s not the situation we have with AI, it’s supposed to scale indefinitely.

If a new driver came out that gave Nvidia 5090 performance to games with gtx1080 equivalent hardware would you still buy a new video card this year?

It doesn’t make any sense to compare games and AI. Games have a well-defined upper bound for performance. Even Crysis has “maximum settings” that you can’t go above. Supposedly, this doesn’t hold true for AI, scaling it should continually improve it.

So: yes, in your analogy, MS would still buy a new video card this year if they believed in the progress being possible and reasonably likely.

I’m gonna disagree - it’s not like DeepSeek uncovered some upper limit to how much compute you can throw at the problem. More efficient hardware use should be amazing for AI since it allows you to scale even further.

This means that MS isn’t expecting these data centers to generate enough revenue to be profitable, and they’re not willing to bet on further advancements that might make them profitable. In other words, MS doesn’t have a positive outlook for AI.

Fucking propaganda. It seriously enrages me how people like you have become so programmed that they’ll attribute everything to an organization you’ve been told to hate. Don’t you ever stop and see how you’re being used?

Why are you bringing this up like it hasn’t been discussed thousands of times? Like it hasn’t been stated (and cleared) in this very post?

Where do you see it working? I see the result:

There were no results matching the query.

The extension mentioned in the post is supposed to:

return the relevant quote and inference for the user, along with links to article and quality signals

I don’t see any relevant quotes, or links to articles, or quality signals.

I tried with your comment: https://en.wikipedia.org/wiki/Special:Search?go=Go&ns0=1&search=I+can+currently+right-click+and+then+click+"Search+on+Wikipedia"+in+the+context+menu.++I+believe+this+works+in+both+FF+and+chromium+browsers.

Why doesn’t this work? If your complaint were valid, this should work.

Legally it’s only cheese if it comes from the “Formaggi” region in Italy. Otherwise it has to be called “bad milk product”.