Comment

Comment on AI-powered CRISPR could lead to faster gene therapies, Stanford Medicine study finds

I am asking that all the people doing this watch Jurassic Park at least once.

I’ll let experts do their thing, but this kind of thing worries me. I never had “there is no singularity, AI just engineers a generic disease by accident and kills is all” on my bingo.

source

Sort:hotnew top

echodot@feddit.uk ⁨13⁩ ⁨hours⁩ ago
Of course the actual moral of Jurassic Park is have a well staffed IT team and not just one random guy who you under pay.

source
Telorand@reddthat.com ⁨21⁩ ⁨hours⁩ ago
This is actually an excellent use case for AI. Physics and chemistry as scientific disciplines are lots of complex pattern recognition and manipulation. AI is just a pattern recognition and generation engine, despite what the tech bros and apologists like to tell us.

What these engines generate will ultimately be vetted by experts before it even goes to trials. Scientists don’t just take things on blind faith simply because a robot or even another expert comes up with something; their entire deal is to understand their particular field of study in great detail, after all!

source
- Cybersteel@lemmy.world ⁨20⁩ ⁨hours⁩ ago
  Back in the day, I used to fold proteins at home.
  
  source
  - echodot@feddit.uk ⁨13⁩ ⁨hours⁩ ago
    There’s something similar that you can do with categorising stars. Basically they want to map as much of the universe as we can see. We’ve got the photos the problem is is that when you zoom in on a 1 mm by 1 mm patch of sky 1,100,000 new stars appear. It’s going to take a while, even with AI.
    
    You never know you might find a Dyson sphere one day.
    
    source
  - Telorand@reddthat.com ⁨15⁩ ⁨hours⁩ ago
    Same
    
    source
Reverendender@sh.itjust.works ⁨20⁩ ⁨hours⁩ ago
Eh. Read the book first.

source