If you think AI safety is really important…you should go work on AI safety, not sign a letter telling other people not to build things

making data useful with AI. Previously built @stelo to keep you safe from phishing 
benscharfstein.twitter
https://nf.td/ben

To be clear, I’m all for aligning AI. I worked on content integrity, fake news, and misinformation for recommendation systems at Google

co-founder stelo.com. theoretical cs phd dropout. into math, music, ml, macro, markets, surfing, writing amandhesi.net

Absolutely, nothing wrong with it. But there are a ton well resourced people on that list who are well positioned to go contribute (Elon, for example). 

My point isn't that expressing concerns about AI safety is bad, it's that I don't think it's particularly useful to arbitrarily say "pause development for 6 months"

Most people (99.9999%) don't have the ability to work on AI safety but most people's lives will be affected by AI. 

So IMO there's nothing wrong with people expressing their concerns about AI and/or ask for regulation.

Be specific.     

leonasskau.co.uk //
Hosting /strategy, /leo, /geopolitics // built Culture3.com // online: leosn.xyz
d33m:mun

Most of the people who signed the letter do work on AI safety

Gamer pride month | @DiagramVentures @Coinbase

My point is that instead of petitioning against progress, people should be working towards the progress they want to see. 

Could be signing a petition for more focus on AI safety. Could be donating to orgs doing good work. Could be funding companies with aligned values.

What if you think AI safety is important but you are not able to work on it?

i'd rather kill the radio and listen to the rain hit

gee, it's almost like people actually want to but capitalism doesn't

https://www.theverge.com/2023/3/13/23638823/microsoft-ethics-society-team-responsible-ai-layoffs

Banning technologies never works. Who want to work on them will just move to a different jurisdiction where it’s permitted and those who banned them lose the competitive advantage. Like Italy and France for the nuclear energy.

But if we don't know how to build stuff, we need to tell the people that are making to think beyond technology.

AI developers must think about ethics, safety, politics, and the economy too. I think we will need AI regulation fast.