Casey Newton

There's yet another open letter from the AI safety crowd. If they want more people to take them seriously, they need to get more specific

https://www.platformer.news/tuesday-newsletter/

What aren’t the OpenAI whistleblowers saying?

A group of current and former employees say the company has been reckless — but aren’t offering many details

Platformer
Jun 5, 2024 at 12:27amIvory for Mac
Show threadEugen RochkoJun 5, 2024
@caseynewton I think the people behind the letter unequivocally subscribe to the idea that the current LLM technology is a path to AGI. It's important to question that assumption. That's the biggest disconnect for me when reading it. Such an assumption directly serves the interests of OpenAI and other LLM companies.
Show threadCasey NewtonJun 5, 2024
@Gargron great point! would love to see them engage with that more directly
Show threadNow at @[email protected]Jun 5, 2024

@caseynewton @Gargron There's also the vital issues raised by @timnitGebru about the very real risks of racial biases in content produced by LLM.

It's an issue that Google appears to have largely chosen to ignore, and then it has ended up with egg on its face as a result.

Also, there's the underlying issues of the eugenicist origins of the Effective Altruism ideology that underpins a lot of the General AI and AI safety hype.

Show threadSloan StudioJun 5, 2024
@Gargron @caseynewton And there is already plenty of damage these current systems can do…