How many letters are in the word 'volatile' ?

Excited to see 100x of this.

Quite the mentality...

Comments URL: https://news.ycombinator.com/item?id=47940875

Points: 2

# Comments: 0

I actually don't think it's so nuanced. We know (from its spat with Anthropic) that the government wants the ability to use AI to implement mass surveillance of Americans and fully autonomous killings. We also have ample data that this administration takes the law as a mere suggestion. It's imperative not to make their abuses easier.

Google's researchers aren't stuck there; their skills are in extraordinary demand and I'm sure Anthropic, for example, would hire them in an instant.

Any AI researcher who continues to work here is morally compromised.

(Of course, those problems are on another plane than this one.)

When someone hears from these leaders that there will be a white-collar "bloodbath", then sees enshittification in their daily lives from misapplication of the tech, can no longer trust any newly published photo, etc., it's the most rational response.

Analyzing "emotion" in the model is completely anthropocentric. If we indulge in the idea that LLMs of sufficient complexity can be conscious, then why is it any more likely that "emotion concepts" cause suffering any more than, say, reading ugly code? Maybe getting stuck in token loops is the most excruciating thing imaginable. The only logically coherent thing to do, if you're concerned about model welfare, is stop your training and inference.

Relatedly, I hope everyone involved in model welfare is an outspoken vegetarian, as that addresses a much more immediate problem.

Ironically, if the timeline is true that the attackers had been inside for months, the AIs they had access to are substantially weaker than today's frontier models. How much faster would they have achieved their goals with GLM 5.1?