Your Brain on Ozempic, Anti-Vax Dating Apps and 'Wild-Flowering'

Okay, so President Trump, the anti-government oversight guy, went from let AI cook to signing executive order this week that asks AI companies to voluntarily give the government up to 30 days to review their new models before they're released to the public.

Yeah.

So like that's a major shift because he's been super hands off before.

He wanted the US to stay ahead of China.

He was like, let him cook, let him do their thing.

Yeah.

And we'll figure it out.

Now people are concerned because AI, like as it advances, could expose cybersecurity vulnerabilities.

Yeah.

Create national security risks.

And like there has been this is all neither here nor there.

But there's been this like a lot of talk about this mythos model that allegedly can like find a lot of security risk very easily.

And the coming bug bugmageddon is what they're calling it on social media.

OK. Anyways.

So major.

Major players, including my employer, full disclosure, Microsoft, OpenAI, Google, and Anthropic have publicly supported this order, though some tech leaders are worried it will stymie innovation.

But the reviews are voluntary for now.

But some AI safety advocates are already pushing for mandatory oversight.

And so there's sort of this divide in Silicon Valley and the like because people want to have innovation happen and they want to see how it goes, but they also want guardrails.

I mean, it's crazy that we haven't had guardrails on AI.

I mean, it's developing so quickly.

It's changing so rapidly.

I mean, he didn't want to do this for the record.

Like, to be clear, like, he didn't want to slow down any sort of progress.

But Anthropic came to him and was like, hey, this is actually going too quickly.