Exactly this. There are competing websites that do pay the workers a bit more, and have much better data. I think it is a combination of both of these factors. There is no room left for cheap human data - that is taken by LLMs.

I tried to like GLM-5.2 but it is just... not good. It is at least 6 months behind frontier models, probably more.

For example just yesterday I gave it a prompt to implement a mathematical computation. The code it wrote took over 45 minutes to run. I asked Opus 4.8 the same exact prompt and its code ran in about 15 seconds on the same machine.

Why should you be allowed to resell a game when you have already enjoyed the experience of playing the game? That is like selling your movie ticket after already watching the movie.

"But I was allowed to play it unlimited times"

The economic value of you playing through the game each time is less than the time prior to it. Selling it to a person who has never played it before is stealing.

The whole concept of a "perpetual" licence that includes the ability to buy support and patches for that software in perpetuity is nonsense that was doomed for the start. That said, it is very different to a perpetual licence that does not include said support or patches which can obviously work.

This wouldn't be very useful. The LLM that Google search uses is very small and you can run a much better and smarter model locally on a modern laptop.

Unfortunately they are right on the Constitution here and it is a loophole that the founders didn't consider.

$36 million seems far too low for this data.

I can't say too much but I work in a small firm using models to detect security issues. I have several sources that indicate that Mythos is not as good as hyped. Notably, existing models can achieve many of the same things Anthropic claims Mythos can do. There is some evidence that Mythos is able to achieve the same goals with more context pollution (i.e. it can handle being given a bigger slice of the code and finding a vulnerability in it, rather than being told to look close up at exactly what might have the vulnerability).

Math, physics, chemistry, and computer science articles are extremely good. Politics, history, etc. not so much.

The breacher trade is even more like this simply because there is no other recourse. Reputation is everything.

These groups make money with their reputation. If paying them didn't guarantee deletion, guess how many people will pay them in the future?

3) is the key point. This adds a big stable customer, which is exactly what African businesses need. Guaranteed money. This is a very good thing for Africa, even if MSFT doesn't directly build the new power plants needed.

Gambling has negative expected value. Futures have zero.

I've worked in several places from big tech to HFT. Now I'm at a startup. Basically all I do everyday is create and update agents, supervise them, and review their work. I'm probably 3-5x as productive as I was at other companies.

Why are there so many comments against this?

This is not gambling advertising. This is actually useful information, often more useful than the news articles themselves.

If you think Polymarket odds are wrong, what is it? Are you the smartest person in the world? Otherwise, why has nobody else decided to outsmart the market and bet in accordance with how they are wrong? Of course not. That is silly.