You did not provide evidence that UK are "arresting people based on social media posts criticizing goverment..."

I think there are users who view "their AI" as somewhere in the venn-diagram of their relationships.

And it's a spectrum, at one end you got the full-on AI psychosis and at the other "its a machine, I owe it nothing".

Conversational AI is going to be sticky to the extent that you see a switch to a different provider as dropping a relationship.

…just so long as the government remains responsible for clearing up the mess?

Would it not be better to have 100 such tests "Pelican on bicycle", "Tiger on stilts"..., and generate them all for every new model but only release a new one each time. That way you could show progression across all models, attempts at benchmaxxing would be more obvious.

Given the crazy money and vying for supremacy among AI companies right now it does seem naive to belive that no attempt at better pelicans on bicycles is being made. You can argue "but I will know because of the quality of ocelots on skateboards" but without a back catalog of ocelots on skateboards to publish its one datapoint and leaves the AI companies with too much plausible deniability.

The pelicans-on-bicycles is a bit of fun for you (and us!) but it has become a measure of the quality of models so its serious business for them.

There is an assymetry of incentives and high risk you are being their useful idiot. Sorry to be blunt.

It's pretty much over for the human-internet. Search was gamed, its usefulness has plummeted, so humans will increasingly ask their LLM of choice and that LLM will have been trained on the content of the internet.

So when someone wants to know something about the topic that my website is focused on, chances are it will not be the material from the website they see directly, but a summary of what the LLM learned from my website.

Ergo, if I want to get my message across I have to write for the LLM. It's the only reader that really matters and it is going to have its stylistic preferences (I suspect bland, corporate, factual, authoritative, avoiding controversy but this will be the new SEO).

We meatbags are not the audience.