Although I can’t help but think that a VM is still more convenient, more flexible, and more secure.

That's all the program is telling you. It doesn't matter that it's backwards, but technically it is.

"These things" being the communication between a proxy and the backend.

One of the LLMs replies "If you're curious, I can also tell you how the competitive scene works or how people qualify—it's a surprisingly serious tournament circuit for such a simple-looking game."

Obviously this has to be pure hallucination, since the tournament in question doesn't exist, and not even the fake source has any details about the tournament itself.

Recently one of them asked Gemini a very detailed question about some specific baseball stats and was exclaiming over the quality of the information he got back and how it would have been impossible or at least extremely difficult to find the information via a traditional search.

It wasn't until his cohost asked if he had verified the information that be realized no, he hadn't, he had just immediately taken it at face value.

I recognize this is a single anecdote, but I think it illustrates that there is a tendency to trust what an LLM gives you, when it's stated so factually and with so much detail -- even if you should know better.

This makes me so worried for the future. AI is only useful _because_ it can pull from all these resources which already exist.

    mkdir /some/dir    
    cd !$   
    (or cd )

If you fork that monopolist’s engine, you’re not making any immediate difference to the market. You’ll adopt all their existing behavior, whether or whether not it conforms to spec (and I would guess you would continue to pull in many of their changes down the road).

A brand new implementation is much more difficult, but if it works it’s much more meaningful in preventing a monopoly.

Although I do wish it wasn’t such a black box — I wish the code for the gizmos could be examined, exported, or embedded elsewhere.