What if?

Easy to "what if" on an unfalsifiable claim.

On the other hand, Grok and GPT finish these tasks in <5min with no issues, and significantly better output.

I've been using them to test my infra/devices as well as reverse engineering.

Not giving legitimate people cyberdefense capabilities with the safety excuse is irresponsible

You just simplified most of the problems people work on down to cancer complexity. Ironic, isn't it?

That's also simply not the case, most people are building CRUD apps with some frontend code and some accessory stuff like build systems etc., which while complex, can still be expressed in very plain, easy to understand language for anyone who's a bit technical.

Does not excuse the Claude slop.

If someone is doing that, it's classified as either a true threat or an incitement, both of which have been clearly established as not covered by the 1st amendment by SCOTUS and thus illegal - and can be stopped for no more reason than the fact that it's illegal.

On the other hand, online content moderation rules often get politically weaponized to silence dissent that is inconvenient but fully legal (and perfectly fine, for as much as you might disagree with it / hate that others get to hear ideas you don't like).

Going from "I voiced my opinion on the internet" to "people are inciting gassing the jews" is quite the stretch, crosses a well defined line, and is just nonsensical.

> Since LLMs can give different answers to the same question, each question was run five times. That means, each LLM was tested 600 times, and in total over 10,000 questions and answers were assessed.

> All models were given the same zero-shot format. They were not given worked examples, previous conversations, hints or an opportunity to correct their answers. This is to make it as similar as possible to a response to a question from consumers.

As for the evaluation itself:

> Responses were checked against this (using an LLM-as-a-judge), and was only given a pass if every element was met; otherwise it was assessed as a fail. This all-pass approach was intentionally strict, so that the score measures whether an answer is complete enough to meet the expert legal standard, rather than how many individual points it gets right.

It's just AI slop and it should be taken with a mountain of salt.

While simultaneously announcing: "btw we're releasing the ever more powerful and more dangerous and more benchmaxing model next week, get our $200 sub asap!"

"We literally prefer it that way" might apply to you and your bubble, but there's more outside of it.

As opposed to non-religious ideologies, whose fertility is decaying and leading them to self extinction.

You have to fork() yourself eventually.