My first class in machine learning was in the 90s, I worked in a megacorp leading multiple teams of people whose primary job was training and running (pre-gpt) high throughput ML classifiers.
I worked with this shit to pay my mortgage and feed my kids, I was decent at it too. One of the guys I worked with came from Palantir.
Nobody called what we were doing AI, ever. There was money on the line, it wasn’t something we could just halfass and appeal to the machine god for. We had angry customers when we fucked up.
Nobody was calling automated medical scans AI, nobody was calling dynamic annealing for searching a solution space AI. Nobody was calling climate modeling AI.
Calling all that preexisting and ongoing stuff AI feels like the branding exercise to me.
My first class in machine learning was in the 90s, I worked in a megacorp leading multiple teams of people whose primary job was training and running (pre-gpt) high throughput ML classifiers.
I worked with this shit to pay my mortgage and feed my kids, I was decent at it too. One of the guys I worked with came from Palantir.
Nobody called what we were doing AI, ever. There was money on the line, it wasn’t something we could just halfass and appeal to the machine god for. We had angry customers when we fucked up.
Nobody was calling automated medical scans AI, nobody was calling dynamic annealing for searching a solution space AI. Nobody was calling climate modeling AI.
Calling all that preexisting and ongoing stuff AI feels like the branding exercise to me.