Lauren Weinstein on Nostr: The problem isn't "in the know" users who have the time and aptitude to double-check ...

The problem isn't "in the know" users who have the time and aptitude to double-check AIOs and other LLM outputs. It's the 94% or some such who DON'T, and just take the output as correct. They're the ones at risk, they're the ones I care about. And they're the ones being most suckered in dangerous ways by these systems.

Genie back into the bottle? No. Though Big Tech hubris across the board in this sector, including the enormous hate they're creating for themselves with data centers -- bipartisan hate and anger -- could well be a limiting factor, even before the economic bubble bursts -- ideally not taking the entire economy down with it.

The answer is liability. If the AI firms are liable for these first party LLM outputs, they will be forced to make the calculation they're trying to avoid making now through all manner of chicanery. Either be willing to absorb the potentially colossal liability costs associated with these systems -- or don't offer them. That's how most firms operate with whatever it is they build or sell. Big Tech should not be immune. They are acting as if they are -- and using leverage with politicians try keep the game going. But as we see, anger at these politicians regarding AI is rising rapidly in a bipartisan manner, and the playing field could change under the feet of those billionaire CEOs very very suddenly and to their deep detriment.