Latest post in my blog on popular science:
Artificial intelligence or computer algorithms?
https://populscience.blogspot.com/2021/11/artificial-intelligence-or-computer.html
Regards,
It might call for some further precision in the definition of “intelligence”, though. Certainly computers are doing–or at least it’s becoming more widely known that they’re doing–things we haven’t seen until pretty recently. For some reason, all the examples I’m thinking deal with image processing, but Apple’s Photos app does a pretty good job of recognizing faces–not just that it’s a face, but whose face it is. Facebook does a fair job of describing the contents of a photo. Astronomy software is able to look at a picture of the sky, and determine exactly where the instrument is pointing based on the arrangement of the stars.
Are any of these AI? No, in all cases they’re computers following the instructions they’ve been given, and all of it has been, in theory, possible for a good long time. But the combination of better digital imaging and faster processing means that they’re able to do it with greater speed and accuracy, so that it’s now actually useful. But still, it’s (if we’re being reductionistic) nothing more than computers following algorithms.
So what is AI, then? Can we define it in a way that is clearly distinct from “computers following algorithms”?
El sáb, 13 nov 2021 a las 17:42, Dan Brown via SpareOom (<
noreply@talk.spare-oom.com>) wrote:
So what is AI, then? Can we define it in a way that is clearly distinct
from “computers following algorithms”?
One definition I’ve read somewhere:
AI is all that we still don’t know how to do with a computer.
Regards,
…and that definition pretty much means we can never have AI.
For me, one example of AI would be the ability to translate properly from one language to another, as well as a good human translator. I have found machine translation to be useful in helping with translations (i.e. finding good solutions to particular words, or even phrases), but it will invariably break down at some point because it cannot ultimately understand the MEANING of what is being translated – it can only rely on probabilities, as far as I can tell – which can often lead to very funny renderings, sometimes the opposite of the actual meaning.
Dimitry