14 Comments

Well…Elon Musk said AI would overtake the human race by 2025 and is more dangerous than nukes. But then again, he said my in-laws would have Internet in rural Michigan by 2020, so…

Expand full comment

Couldn't be much worse than our current governance

Expand full comment

I for one welcome our Boston Dynamics Deadly Robot Dog overlords.

Expand full comment

In all seriousness, I think a lot of AI research won't amount to anything useful.

A lot of what we need is pretty low-fi. Housing, doctors, mass transit, alternative energy. None of this is helped much by better software.

I really wish our smart people worked on non-software tech. I've heard China has an initiative like that.

Expand full comment

Housing, doctors, mass transit and other infrastructure. Totally agree these are pressing needs that don’t seem very high priority to our leaders. Still….I want a Rosie the robot from the Jetsons before I’m dead. Forget overlords, we could use robot servants.

So… but what is China doing?

Expand full comment

Basically moving their smart people from software to other tech, believing that the former can produce a lot of private profit but won't do much long term growth. Just heard it on a podcast though I don't have a good source

Expand full comment

Interesting. Am always interested in what they’re up to.

Expand full comment

Oh! Forgot to mention my favorite part! Had to re-read this before I got it:

"Unlike professional baseball players . . ."

LOL!

Expand full comment

I'm learning a *lot.* But I find, here and there, a lotta home truths from the business sphere:

"As with machine learning, the difficulty of AI does stems not from implementing the algorithm but from understanding the problem."

Seems like this is great general rule of thumb that applies in business as well as a lotta aspects of any person's life.

"However, almost all real-life implementations face time constraints,"

Wouldn't life be easier if this weren't the case. A lotta people plan as if it wasn't.

"In all cases, creating and optimizing the AI requires a lot of human judgement."

Ah, yes. And, as AI goes mainstream, I believe one-a the highest paid occupations will be going to those who can tell if an AI prediction is reasonable and fairly accurate. Because, at first, a lotta people will just wanna take whatever comes out, "because the computer 'said' it was right."

Expand full comment

I updated all the typos. Thank you for that! I can't believe I published this one as is.

Expand full comment

Naw, it was my pleasure!

Expand full comment