Back in my days, we had to search things on Google and read the whole damn thing on stack overflow and manually rm rf the system, could you imagine that?!
As opposed to closing Jira tickets, AI requires a mathematical background to truly appreciate what's going on.
I agree, there is no mathematical depth to it. It is simple math plus playing with different ideas to see what works and what does not. There is not much understanding and it is very trial and error dependent. Only big companies can effort these trials and errors.
COLT is one of the few that has decent stuff, but it's still just applied math.
Yeah there are theoretical ML papers that are basically math papers but still there is dichotomy between theoretical stuff and the stuff that people find cool.
If you think it’s boring now just wait until your jobn’t because of it 😄
Do you actually believe this?
Do you actually believe that corporations in a capitalist society are incentivized to pursue anything less? It won’t happen all at once tomorrow, or next year. But over the next 10-20…