- cross-posted to:
- Aii@programming.dev
- cross-posted to:
- Aii@programming.dev
You must log in or register to comment.
I hate that everyone just calls everything “AI” now. At this point, this term has basically lost its whole meaning. But I guess writing “Scientists Used Machine Learning to Find …” is not as attention grabbing. Though “Machine Learning” is also a very broad term
As usual, the term AI is used for Machine Learning. Nothing to do with the Techbros large scale language model dead end. Sad that a success like this could be misconstrued to be a Techbro success.
They‘re based on the same principle and are therefore both AI or based on machine learning. The actual misconception comes from LLMs being treated as AI when they‘re just a product of machine learning. Both things are based on the same principles of pattern recognition and training data, though.
The difference here is that scientist use machine learning for something useful while tech bros use it to make word salad generators.
I wish they picked a better acronym than LLM, it’s really awkward to say. Maybe then people wouldn’t call everything “A.I.”. It’s the equivalent of calling everything from a phone to a desktop to a traffic light “a computer”
LLM is fine to say, not more awkward than element. And it is AI.
And I wish we’d call phones computers. It’s been very profitable for corps for people to not realize that they are.
It’s more awkward to say because it’s “ell-ell-emm” not “el-em-ent”. It’s like that middle part of the alphabet that kids always fuck up “ell emm enn” and it’s practically a tongue twister. It’s a very unusual mouth movement for English speakers. Which is why many people started calling them “llamas” because the longer word is actually easier to say.
before anyone immediately smashes “downvote” and comes here arms raised, “AI” includes ML, as in models specifically trained for this, and models have been around for 20+ years. AI != LLMs
One of the insanely powerful uses of AI is just parsing large amounts of data and looking for patterns. However it still requires human oversight and may overlook patterns also. I think this is the problem with AI in general: The vast majority of people can’t be trusted to verify the results. Scientists are extremely thorough in general so that’s probably fine.
I had a family member arguing with me about a particular piece of law that I happened to be very familiar with. They are actually an attorney. And they flaunted that as if it made them right. The problem is they didn’t even actually review the law, they just typed it into chat GPT. So they were breaking the law based on advice from a chat bot.
the greatest power of machine learning algorithms is the source of its greatest drawback. they are essentially heuristic models of something, constructed in a way that they are much cheaper to execute than a traditional algorithmic approach. this cheapness results in error, which for a lot of applications is fine because you can refine/check the result with more accurate tools, but it also means you can never just trust it like you can with more traditional tools. this problem is baked into the technology so theres no amount of scale that will make it go away, as we’ve seen time and again with LLMs. failing to understand this is the mistake most people make when it comes to “AI”, and results in all kinds of bad decision making. but in the hands of people who understand the limitations, machine learning can truly be a game changing technology. not chatbots though those are fucking stupid and i hope they go away when the bubble bursts.
Also, LLMs doesn’t equal generalist trained chatbots.
How is that relevant here?
Not relevant to the article which isn’t about LLMs… but it IS relevant to “this isn’t about LLMs, don’t downvote!” Which implicitly implies LLM discussion is intrinsically worth downvoting.
