Bear in mind these types can explain things like why word-alignment matters and train themselves into being net productive within a few weeks.

There’s also a role called being an algorithms engineer in standard tech companies (typically for lower level work like networking, embedded systems, graphics, or embedded systems) but the lack of an engineering background may hamstring you there. Engineers working in crypto also use a fair bit of algorithms knowledge.

I do low level work at a top company, and you only use algorithms knowledge on the job a couple of times a year at best.

Rumor has it they ran an internal poll on whether their employees would purchase AI glasses which is their first step when developing a new product.

Then it’s not for them. You don’t see me complaining about advice for plumbers bc I’m not a plumber. The advice is for the 1%. Anecdotally I know numerous people, including myself, who have been in this situation.

The use case of transferring ALL data over every time is obviously misusing the GPU.

If anyone’s ever tried running a model that’s too large for your GPU you will have experienced how slow this is when you have to pull in the model in parts for a single inference run.

https://web.archive.org/web/20190906050523/http://www.strike...

https://www.ribbonfarm.com/2009/10/07/the-gervais-principle-...

My personal takeaway was that much of our work involves complete BS. Either join a useful, innovative startup and make a difference or play your part in the farce and take your paycheck.

I still work hard in some roles, but now it’s to get a raise/bonus, or because I find the work fun. It’s not for the sake of working hard anymore. If you want to work hard, go dig a hole.

We basically bought out a ton of A- to A+ players from the rest of the industry and there are many high performers developing entire modules of core functionality themselves with a team of juniors and contractors supporting them. I also have a hunch they make $350-$750k.