Why Embodied AI Is the Red Line We Cannot Cross

chrbutler.com

4 pointsdelaugust3 comments

Seems pretty straightforward that AI wiping out humans is a bad thing.

Article needs to address the premise more clearly. Why can't we embody AI?

The author seems afraid that AI would replace and wipe out humans.

1) Is that a bad thing? 2) the author is literally anthropomorphising AI - giving it traits of humans (aggression, desire for conflict, genocide) just because they might look human.

These are not encessary truths and the article leaves them totally unaddressed. There may be valid evidence and support for 1 and 2 being bad/true but since they're not covered the article is just unsubstantiated fearmongering.

The real topic is "how do we ensure our bad traits aren't passed on to AI made in our image".

I'm not sure if you've read the AI 2027 report, but that's what this post is responding to. It's not the author anthropomorphizing AI. That's the entire premise of the AI 2027 forecast.

Reply on news.ycombinator.com