this post was submitted on 24 May 2024
1059 points (98.9% liked)

Science Memes

10420 readers
2714 users here now

Welcome to c/science_memes @ Mander.xyz!

A place for majestic STEMLORD peacocking, as well as memes about the realities of working in a lab.



Rules

  1. Don't throw mud. Behave like an intellectual and remember the human.
  2. Keep it rooted (on topic).
  3. No spam.
  4. Infographics welcome, get schooled.


Research Committee

Other Mander Communities

Science and Research

Biology and Life Sciences

Physical Sciences

Humanities and Social Sciences

Practical and Applied Sciences

Memes

Miscellaneous

founded 2 years ago
MODERATORS
 
you are viewing a single comment's thread
view the rest of the comments
[–] [email protected] 67 points 4 months ago (3 children)

Neural networks are magical anywhere that near misses are good enough.

Companies keep using them as if they're infallible, when lives and fortunes are at stake.

Tech is not the problem.

[–] [email protected] 30 points 4 months ago (3 children)

Tech is ravenously trying to convince the world they need AI for every aspect of their business. Tech wants you to think LLMs are infallible and they strongly imply that they are even if the fine print says otherwise. So personally I would say tech is very much part of the problem. One could say they are the root of the problem in fact.

[–] [email protected] 29 points 4 months ago (2 children)

I think he’s referring to literal technology itself as “Tech” and you’re referring to the people trying to sell it as “Tech”—aka “Big-Tech” as some would say

[–] [email protected] 18 points 4 months ago

Yup, sounds like a classic miscommunication. I think we should all compromise and just agree capitalism is the problem lol.

[–] [email protected] 5 points 4 months ago

I think you're bang on with that. My bad.

[–] [email protected] 27 points 4 months ago

I don't think the tech is the problem, it's the business drones trying to sell you it

[–] [email protected] 8 points 4 months ago

The tech is not the problem.

[–] [email protected] 11 points 4 months ago

If you give out hammers to everyone, some people will end up with smashed balls.

[–] [email protected] 2 points 4 months ago (1 children)

Yeah, I have to agree with you. For example, I would have no problem using a decently tested LLMs for engineering simply because Engineering usually accounts for errors and uses appropriate factors to accommodate them. Sure LLMs could be get more accurate in future, but I believe the error will reduce asymptotically. Essentially, more accurate LLMs get, it will get that much harder to increase the accuracy. There is always a price to pay, IMO.

[–] [email protected] 5 points 4 months ago

"There's always a price to pay" is basically what engineering is.

Anybody could build a bridge to last 100 years, or to survive a barge ship crashing into it, but it takes an engineer to build a bridge that will barely last 100 years, or barely survive a bridge crashing into it (which you could kind of say the F.S.Key bridge did, since only really the middle section was taken out).

Put another way, in the real world, there are budgets and sacrifices.