419
AI comes up with battery design that uses 70 per cent less lithium
(www.newscientist.com)
you are viewing a single comment's thread
view the rest of the comments
view the rest of the comments
this post was submitted on 13 Jan 2024
419 points (85.8% liked)
Technology
59419 readers
2916 users here now
This is a most excellent place for technology news and articles.
Our Rules
- Follow the lemmy.world rules.
- Only tech related content.
- Be excellent to each another!
- Mod approved content bots can post up to 10 articles per day.
- Threads asking for personal tech support may be deleted.
- Politics threads may be removed.
- No memes allowed as posts, OK to post as comments.
- Only approved bots from the list below, to ask if your bot can be added please contact us.
- Check for duplicates before posting, duplicates may be removed
Approved Bots
founded 1 year ago
MODERATORS
That's not true at all. AI can in fact generate novel techniques and solutions and has already done so in biotech and electrical engineering. I don't think you understand how AI works or what it is
I think maybe people are running into a misunderstanding between LLMs and neural nets or machine kearning in general? AI has become too big of an umbrella term. We've been using NNs for a while now to produce entirely new ways to go about things. They can find bugs in games that humans can't, been used to design new wind turbine blades (even made several asymmetrical ones which humans just don't really do), or plot out entirely new ways of locomotion when given physical bodies. Machine learning is fascinating and can produce very unique results partly because it can be set up to not have existing design biases like humans do
And the nature of computers is that they are magnitudes better than humans at brute forcing. Machine learning can brute force (depending on the technique, it can be smarter than brute forcing, being more efficient) test many many many more designs and techniques than we could manually do. Sure it'll fail many times, but it's just a numbers game, and it can pump those numbers. It'll try a lot of weird and unique stuff we wouldn't even think to try, with varying degrees of success.
What novel solutions has ai done in electrical engineering?
Name one that wasn't just doing the thing it was told and the users being surprised. You know, the same way that people are surprised when research has results they did not expect using other approaches.
It's a weird way of asking this. Of course it's going to do what's told, the alternative is that it, out of the blue, spits a battery design for no reason. If it were to somehow find a way to make batteries with less lithium in a way that never did before, isn't that an unexpected result using other approaches?
This is not general artificial intelligence, everything we have is narrow AI, focused on solving one specific problem, for identifying birds to understand instructions between drugs.
Yeah, that would be coming up with a battery design.