HardikVala

Check out epoch.ai. You won't get a direct answer but it'll give you directional, and useful data points.

I think the frameworks became complex over time but were initially simpler because of the more managable set of AI workloads.

There might be but instead of thinking about it as a binary, think about it as a continuous conviction curve. Ask yourself, if this idea is true, what's the smallest testable prediction you can make? It can mean running a small experiment that doesn't give you statistical significance or would pass scientific scrutiny, but can be a gating function for a slightly more sophisticated experiment. And then you rinse and repeat until you prove/disprove the idea.

Not the first product category you think of when it comes to AI consumer hardware but how about smart glasses (eg. Meta Raybans). Extrapolating to the future, AR glasses are probably going to be heavily dependent on AI.

Great recommendation. Based on the ToC, its similar to AIMA.

Andrew Ng's course is great for the learning NN's from scratch, but not understanding how NN's fit in the broader discipline of AI.