brokenmachine

If they could give their AI Agent access, then by definition it wasn't a secure system.

It's a bit of a mouthful... :)

Not buying AI, buying other assets.

The marketing tries to tell us that we're on the brink of AGI, but still in 2026 putting a random laptop reliably to sleep remains an open problem.

Not to jump on you, but it just irritates me when people anthropomorphize LLMs...

LLMs don't "understand".

Don't fall for the marketing and overstate their ability.

The companies won't give up their scams without a fight.

They'll double down with secure boot and encrypted everything.

Can the bubble please pop soon so I can buy the dip before retiring?

In the article it says he was annoyed with Claude because it wasn't solving some other problem that he was asking it about.

Reading between the lines, it seems he may have got annoyed and made a wrongthink joke to a computer and eventually got arrested for it.

A joke with a computer, no matter how distasteful, is not past my reasonable line.

Getting arrested for making a joke in your own home is, however.

Of course, it stands to reason... if he is charged with planning to carry out a shooting, logic dictates they found a gun? Or plans to obtain one? Bullets?

Or anything at all other than an email from an offended slop generating plagiarism machine?

If you don't have a factory, then what's the point.

Yes, but not in the exact, specific, bespoke way that people need it.

I'm pretty ambivalent about AI, but...

Even if some bloated, expensive software that may have contained those functions already exists, bespoke software that does exactly what I want, and only what I want, in the way I want, is still "new" in my opinion.

I liked when it happened to Adobe, but I don't like this.

And then raised the prices of them because why the hell not.