I am happy. AI is great. I use it. I'm glad that the U.S. is building datacenters. I work for a company involved in siting and building datacenters. But AI is not as scary or dangerous as Dario claims.
Its a mix probably between a new potential and the progress.
Taking it seriuos and then figuring out its not that dangerious is probably a 1000 times better than the opposite.
And the last Ubuntu LTS patch had the most CVEs fixed i ever seen. After Snowden, you can assume a lot more of what i thought would be 'tin foil hat' might just have happened like a massive exploit scheme of CVEs.
Or things like Stuxnet.
You always had to be an expert somehow, now you can throw compute at it. Just imagine a USA, China, russian or israely hackergroup having their uncensored models running in secret datacenters with very basic prompts because they don't care.
It would be weird tbh if state actors/hackers aren't doing that yet which already shows us one big problem with ai:
AI is technology you have to use now and you have to invest. If you are not, someone else will break into your systems.
