ialyos

I don’t think that’s true! IIUC the protections from that act only apply to a subset of humans, including us military and nato military members. It’s in the article you linked! So if you see Putin at your local Safeway go ahead and citizens arrest him :)

Is the goal here to make money? Engineer #25 will get at most .3% equity. At a billion dollar valuation that's 3 million, which sounds like a lot, but if you see that exit after 5 years, that's more like 600k a year. Senior eng at Fang makes 400k+, and is much more attainable than first 50 at big company. You'll have a much better chance climbing FAANG ladder if you want to make a bunch of money

Thanks for the nuance. I completely agree with how you've framed the situation.

The article is simply wrong.

I know this because I worked as an ML engineer at an extremely successful company that automated medical coding using deep learning.

The confusion stems from conflating a "perfect solution" with a "human augmented" one.

90% of coding cases are trivial, have low value and can be done by a model. 10% are really subtle and need human expertise.

That's fine. You can make a billion dollar company on low hanging fruit. I think it's best not to conflate the perfect solution with a very good solution.

Did you watch the video you linked?

Become a machine learning swe. It's an amazing blend of stats / match and raw engineering. It's a ton of fun

The phrase "in practice" doesn't really work when you're referring to highly finicky strategies like RL and adversarial training

I don't have a nuanced thought to add. I think it sucks these videos were demonetized and I agree there's something weird going on when 3d bust test prints abound. Sorry you have to fit so hard to get people on the same page

This comment is underrated

Isn't this what protos are for?

I don't think that's true. As a consumer I pay Uber, that internally splits its revenue between unionized and non unionized labor. Both parties want money, and the both charge me for it.

I've built one and so has someone else, it's a journey, but gets you a reasonably competent arm at a reasonable price.

This is at a high level how one of our Google internal black box optimizers behaves. The term used is "infeasible region" but it's the same idea as using a nan.

I would caution against using nan to always mean infeasible. Instead users should catch experiments outside the feasibility region and return a special infeasible value. This will increase visibility into the behavior of the optimizer, because it leaves nan to be used for values inside the region of constraint that are still problematic (due to bugs, numerical instability, etc)

Work for Google. Work on the same feature for a year. Deadlines are fungible and no one cares if you aren't killing it