sam0x17

I remember when I was ~14 at a private day school with a 1:1 laptop program in the early 00s, we were on AD. There was essentially one mac user in the IT department and the rest was windows, and inadvertently his mac created a hidden folder, I think it was "FindByContent" or something, that allowed full permission to all users in AD. Kids quickly noticed this, made a new AD group that was just us and locked it down a bit more, and for years interesting things started building up in that directory, pirated games (halo 1 for pc), artwork, test answers, even some porn if I remember correctly. Years later I worked for two years in that IT department and it was still there, with a whole decade of built-up content. We decommissioned it but it was super interesting to see how it had been used over the years.

Notoriety also has value, not that that is OPs motivation, but worth noting

1. Make a new model

2. Add "Flash" to the end

3. ?

4. Profit

What they mean is "what was considered frontier a few months ago"

I thought it meant 15 million underground tunnels

> GPT 5.6 Sol is the best "vision" model OpenAI ever released

I mean I should hope so, as it is also the latest one

I didn't say fix, I said context. As in "you are using an established thing the wrong way" which is most questions people ask. Edited

> I can talk to Claude myself

OK but can you talk to my specific claude code conversation that has 6 months of context on the mechanism you are struggling with?

> The only thing a ban would do is protect the American market from further downward price pressure on inference, protecting VC investors in the short term

It won't even do this. Streisand effect will probably draw even more attention to the open weight models

The entire marketing narrative in AI already operates this way --- "GPT 2.0 is too dangerous to release, oh noooo!" etc

Mark Twain

I wonder what kind of result you'd get with floats

What I like is it's still hard, but you can do things like "prove this using game theory" or "find the optimal value for this and a proof" without having to know game theory or deep math really

There used to be a lot of research into using deep NNs to train decision trees, which are themselves much less of a black box and can actually be reasoned about. I wonder where that all went?

Sounds like a vibe-coded feature if I ever heard of one