ML has been used for eg. medical reasons years before ChatGPT dropped. It’s only called “AI” now because that sounds fancier. Consumer ML (“AI”) is useless for anything professional - generating a 10 paragraph text filled with emojis and em dashes, and even more factual errors won’t solve anything. Continuation of training ML models to eg. detect virus DNA similarities, simulate vaccines, create realistic simulations of weather, climate, vulcans etc., which - again - has been done since more than a decade or so, is actively being done. It’s just that it’s not as simple as feeding a few million Reddit and Facebook posts into the learning process to get factually correct results in a useful format.
ML was called AI before too, but since the consumer didn’t interface with it directly usually, it was just part of the marketing, not the entirety of it.
Had a fancy ML algo power a small part of your solution? Your software is now AI powered.
Some years ago, I wanted to implement machine learning to get rid of the pigeons on my balcony. The other birds were fine for me, but those little fuckers crapped all over the place. I thought about an automated water gun.
We could be using Ai to solve the world’s problems, advance humanity, and grow as a human race… Yet here we are.
Imagine how many of the world’s problems would be solved by now without AI if that were ever the goal of the people in charge.
ML has been used for eg. medical reasons years before ChatGPT dropped. It’s only called “AI” now because that sounds fancier. Consumer ML (“AI”) is useless for anything professional - generating a 10 paragraph text filled with emojis and em dashes, and even more factual errors won’t solve anything. Continuation of training ML models to eg. detect virus DNA similarities, simulate vaccines, create realistic simulations of weather, climate, vulcans etc., which - again - has been done since more than a decade or so, is actively being done. It’s just that it’s not as simple as feeding a few million Reddit and Facebook posts into the learning process to get factually correct results in a useful format.
ML was called AI before too, but since the consumer didn’t interface with it directly usually, it was just part of the marketing, not the entirety of it.
Had a fancy ML algo power a small part of your solution? Your software is now AI powered.
Some years ago, I wanted to implement machine learning to get rid of the pigeons on my balcony. The other birds were fine for me, but those little fuckers crapped all over the place. I thought about an automated water gun.
Future alien archaeologists studying our ancient cities and remains of our civilization … “I wonder what caused them to all die out so suddenly?”
Why don’t you quit bitching and go solve some of the world’s problems then?
Maybe I will
I’m actually kind of interested in:
How they got it to give them all 5 toes.
How they got it to make millions and not to 20 or 30
me: make me a picture of loads of cats
AI: 2 cats
me: no, like millions of cats
AI: 7 cats
me: no literally, half a billion cats.
AI: 5 cats and an Okapi
Have you taken a closer look?
This 25 minute Veritasium video covers just that: https://youtu.be/P_fHJIYENdI
Obviously not an exhaustive list of all things that need doing, but proteins and protein folding are important for a lot of fields.