Log in
Sign up
See new Tweets

Conversation

Shivon Zilis
@shivon
·
Aug 14, 2017
AI safety, from algorithm development to fictitious data. Great, and important, read.
Dario Amodei of OpenAI. He and colleagues are developing algorithms that can learn tasks by trial and error and also get guidance from human teachers.
nytimes.com
Teaching A.I. Systems to Behave Themselves (Published 2017)
As philosophers and pundits worry that artificial intelligence will one day harm the world, some researchers are working on ways to lower the risks.
2
14
42
NASA DEBUNKER THEBIGLIE!
@Zerstaubung
Replying to
@shivon
Good one for the MadData professor
@KirkDBorne
👍🏻
4:38 AM · Aug 14, 2017·Twitter for iPhone
1
Like