pvillano

The proofs could be optimized for human consumption. OpenAI could optimize each proof for fewest theorems, low branching, or short dependency chains. OpenAI could have an adversarial network enforce that proofs look human. OpenAI could spend a couple more machine hours per paper with the prompt "edit for clarity".

I think OpenAI wants the outputs to be unreadable. If outputs "too advanced for human comprehension" become the norm, then every interaction requires tokens. If someone can buy a day pass, get what they need, and leave, then there's no recurring revenue stream.

The proofs could be optimized for human consumption. OpenAI could optimize each proof for fewest theorems, low branching, or short dependency chains. OpenAI could have an adversarial network enforce that proofs look human. OpenAI could spend a couple more machine hours per paper with the prompt "edit for clarity".

I have to wonder if LLM reimplementations verified against years of human tests will have holes where original authors thought that tests are not necessary and common sense is enough.

I have to wonder this to preserve my ego as a human. I wonder have to this because I sure as hell am never going to go through all that code to check.

It's a static site.

I am not reading that. XKCD, one of the most popular web comics of all time, is STILL stick figures. There's no excuse not to make the entire thing yours.

Thank you for introducing me. A fun read.

No one asked, but I'll analyze the differences between these two situations. Landis didn't impersonate any living people. It wasn't done against a backdrop of living people losing their jobs. What he made was still art, in that it was both a display of technical skill and worth discussion. Slop is neither. From now on, his forgeries will be clearly labeled as forgeries any time they're displayed. That's easy to do with a physical painting, but impossible when an image has been shared across platforms.

Without AI, if you spend $X of developer time to find the first n most obvious security vulnerabilities, then attackers have to spend a proportionate amount to find one of the remaining ones.

With AI, if you spend $X of tokens/developer time to find the first 2n most obvious security vulnerabilities, then attackers have to spend a proportionate amount to find one of the remaining ones.

The problem is not eating McDonald's for every meal. It's that we've forgotten how to make healthy choices.

You can eliminate some fingerprint by writing in standardized scientific English. One of the benefits of standardized scientific English is to conceal the race and gender of the author for more impartial review.

Nevermind. Banning accounts wouldn't cost the pirates anything if the pirates just do one release batch a month and just burn a new account every month.

Something I'm surprised I haven't seen is spymarking premium digital content to identify the accounts of pirate uploaders.

Why would someone security conscious

1. be okay with sending their prompt to an AI company?

2. need to publish something AI generated?