this guy got a turing award for backprop and like 20.other things posted:

A clever deflection. But can you please say if you think learning disabled people understand some things even though they do not understand others. This should be an area in which you actually have some relevant expertise.

Geoff

Jose Valasquez posted:

Designing Data-Intensive Applications is my go to recommendation for system design interview prep. It's a great book, but not really a way to practice though.

http://highscalability.com is another good resource, it has a section of real life designs, so you can do something like pretend the question is "design instagram" and see how close you get to their example. Ironically the site can be very slow at times.

I haven't used it, but I've also seen this recommended https://github.com/donnemartin/system-design-primer
Ultimately system design interviews are going to be more about the back and forth conversation to uncover requirements and address them, so it's hard to practice them in the same way you can grind out leetcode

bob dobbs is dead posted:

ml publication always looks like that tho

the fuckery is literally not publishing any replication anything. so the science is at a complete end and they shouldnt prolly be accepted to any conferences anymore

Xarn posted:

So if your job is primarily doing the same old thing we've done and documented over and over, you have an issue.

IShallRiseAgain posted:

and then it turns out that Glaze is stealing code
https://www.reddit.com/r/StableDiffusion/comments/11sqkh9/glaze_is_violating_gpl/

ultrafilter posted:

ultrafilter posted:

The Fool posted:

speaking of contaminated data sets

Sivart13 posted:

hey artist, before posting pass every single one of your artworks through this magic box that will confuse computers for three or four months

Jabor posted:

It's fine to say you want more money, you just might need to phrase it diplomatically. Remember, you're worth the amount you're asking for - the reason you're leaving your previous employer is because they're not willing/able to pay you what you're worth.

Are you desperate for any job at all, but getting paid more would be a nice-to-have? Or is getting paid more a hard requirement for you to make the jump? If you're in the latter situation, you don't need to worry about companies who can't afford to pay you more rejecting you out-of-hand - in fact, it's a good thing if they do, because you wouldn't have taken that job anyway and now you don't need to waste time on the interview loop with them.

Ensign Expendable posted:

Usually I say I capped out the career growth at my current company and I'm looking for a position that isn't attainable without changing jobs.

StumblyWumbly posted:

"I've enjoyed working at my current company, there are neat things to work on, but I think my compensation has fallen behind and I've tried to fix it there but it hasn't really worked out" is an answer that shouldn't turn anyone off.

oliveoil posted:

Do you usually have meetings with your lead when your project with an estimated 3 month timeline slips 2 or 3 days to figure out how to get it back on track?

I was super tired for a few days and worked pretty slowly. There's no "blocker" or work that suddenly needs to be parallelized or need for me to get some expert to explain how the system works to speed things up. It doesn't really make sense to me.

It's the same for individual sprint estimates. You say you're finishing early and that's fine. You say not everything is gonna get done this sprint and suddenly it's meetings on how to get you back on track. You're gonna over or underestimate the task length almost 100% of the time so about half the time you're gonna be off. I've never had these meetings about why something isn't done yet and how to get it "back on track" in 8 years till I got on this team so it seems a little unusual.

It feels demoralizing though.

oliveoil posted:

My estimate failed because I was unexpectedly weirdly tired for a week and got about half as much done as I expected.

I'll try to commit to less though.

Steve French posted:

Am I crazy for thinking that for a 3 month project the idea that you can measure whether you’re on track with 2-3 day precision is absurd unless you’re at the end of the project? I guess OP didn’t mention how far along the project is

George Wright posted:

If you find yourself getting weirdly tired again maybe go see a doctor, too.

Edly posted:

Do you have a sense of how much you get done compared to the rest of your team? Have you ever gotten feedback before around velocity?

oliveoil posted:

Yeah, I do feel like other people churn out code much faster than me. I've only been able to match the volume I've seen from others by taking Adderall. I also got feedback from my manager on velocity when I first joined this team as well, when I was under this same TL.

To be frank I've always wondered if there's something wrong with me since I pretty commonly find myself tired seemingly at random and repeatedly forgetting what I read immediately after reading it, forgetting what a section of code or a function does immediately after reading it while I try to understand how it fits into another part of the codebase, etc.

Like supposedly normal people can keep 3-7 things in mind at once, but I randomly end up being at 0-2 things in mind at once. And I spend a lot of my time at 0 or 1 where I just repeatedly forget what I'm reading or doing while I'm still doing it and have to start over or at 1 where I forget what I was just reading or doing as soon as I have to do one other small task or read one other small thing.

I've never found a doctor who would take it seriously though so I don't know if I'm just imagining it. I do feel like it's rare that I can just sit down and code at the same level as my peers though.