ppppo

> Wombat dissections show that cubes are formed within the last 17 percent of the intestine.

the article itself is paywalled, I wonder how did they get the wombats to dissect.

I'm sure they have infinite resources and time to make it available in every language ever including braille.

But seriously, it's obviously meant to have a wider audience than just those countries you mentioned. Not even mentioning the elephant in the room that they are funded by many other eu countries (~23) that have their own languages as well. What else besides the most international language in the world should they have chosen?

I do see people already creating this new niche, as we have airport security and unnecessary bureaucracy (Germany as an example) we will probably have something like "AI supervisors" who basically chat with an llm for a few minutes a day for the state or local municipality and make a career out of that.

Thanks, I see. I guess the real issue is the reach. For now at least I don't see ai having that sort of reach that would make me uneasy, like interacting with kids in video games.

I can see how it should have guard rails there on what to say. However, since llms are approximations of the same neural nets as humans, I wonder if it will ever be 100% safe. Even with humans who we assume to be for sure good have bad intent sometimes (eg doctors, teachers abusing kids).

Serious question, what is there to ai safety? So far I think it's just pretty helpful when coding certain things, but most of what it gives is low quality. What are the real risks involved?

Maybe it's just me but so far I don't see risks, maybe once there's GPT 42o.