So I’ve been studying programming AI bots. And in that warren there is also Machine Leaning/Deep Learning/Machine intelligence.
Further down the warren are the debates about intelligent robots replacing jobs. Which touches on the subjects of economy, deflation and capitalism in general.
So, you know, the A.I. hole goes pretty deep.
Coming back to the surface: having studied more closely programming a bot and how to endow it with some rudimentary intelligence I stand back, into the light of deep learning, and do wonder: Why program a bot at all? If I agree that deep learning is true (I refer generally to the method of reinforcement learning via neural networks), programming a bot “the old fashioned way” doesn’t make sense, and I am just wasting my time on such a redundant and nonsensical mechanical process. The equivalent of punching instructions into ticker tape.
My aim with learning AI bots was to create “apps” that would be useful. I wanted to apply them to Healthcare for example. But why would I continue this path? It would make more sense to work in deep learning and “hyper-program” the bot to think for itself and write the apps itself.
Only, deep learning is pretty heavy stuff.
I was never a great fan of math, and from my first studies of deep learning, well, there’s math. I’m not even great at math. A robot is great at math. Which seems like a conundrum. I need a robot to write the math to write the robot…
Someone is going to have to write the last math ever. Just as some say machine intelligence will be the last human invention ever.
And I ask myself: if robots do take over, and my healthcare contribution is unnecessary, and I am free to do whatever I please — Would I study math?