Skip to content

Comment on Ask HN: Isn't Anthropic currently doing "security through obscurity" for Mythos?

Comments

"a bunch of companies" have "probably" fixed "some issues"? There are huge security risks for government IT systems around the world, giving attackers tools cannot be understood to be a direct solution and is not viable without more buy-in.

It's been more than 2 months since Mythos was released to a select group of companies to prevent this scenario. I wouldn't mind if Anthropic were to delay the release as many months as it takes, as long that at some point, they release the full model to everyone.

Because what are you implying otherwise? "There are huge vulnerabilities everywhere, it'd be bad if people were to find them!" How is this suddenly the acceptable take towards cybersec? Wouldn't you WANT them to be found and then fixed? If anything, today it's mythos, in 2 years it's OpenMythos, in 6 years it's DeepMythos... All AI converges towards progress anyways

Everyone keeps warning against the huge vulnerabilities as an excuse to not release the model, wouldn't you want to live in a world where every piece of code actually survives the model instead?

AboutSource Built by g1lg1l

Hackerly is an independent reader for Hacker News, built on the public HN API. Not affiliated with Y Combinator.