HackerTrans
トップ新着トレンドコメント過去質問紹介求人

bayareapsycho

no profile record

コメント

bayareapsycho
·2 か月前·議論
I'm doing OK, I can get phone screens / interviews at big tech. Can't manage to get any from AI labs.

My current company is highly abusive and is doing a bunch of saber rattling over layoffs and I'm honestly fed up at this point.

They want us to work 7 days a week and are like "ohh maybe we'll lay half the company off, we'll see". And all of the reorgs are making my life fucking hell. (It's a large company that's run by an asshole. Guess which one)

The hard part is just getting the time to prepare for interviews properly. I'm not even bothering interviewing at the more leetcode heavy companies right now.
bayareapsycho
·3 か月前·議論
> Being spoon-fed information isn't the same as learning, to me

It's like it distills it for you. I feel like you're thinking of an example like trying to learning operating systems by reading wikipedia articles (i.e. it gives you a high level summary but nothing more).

The way I see it, code says a lot, but it takes time to scroll through it and cmd+click back and forth. But if you just ask the AI "where's x thing happening around this file" it will just point you right to it. So I feel like less cognitive energy is spent dealing with the syntactic quirks of code and more is spent on the essential algorithmic task.

I don't really like using it to summarize natural language written by one author or group, like a paper for example, that just feels like laziness to me.
bayareapsycho
·3 か月前·議論
But how do they learn to do their respective task? How is the information disseminated?

The capability is there for robotics to handle these kinds of repetitive tasks from a long term view. They're just statistical processes on a fundamental level.

In general, a lot of this shit that we do can be represented this way. It's just a question of where the incentives are to apply it first and how many economic cycles it'll take to get there.

Also, who controls the training data will matter a lot more. I.e. the sort of "ancestral knowledge" within different enterprises and how they deliver on respective business goals.
bayareapsycho
·4 か月前·議論
I don't really feel this kind of friction. In deep coding sessions it usually follows this loop of

1. cook up design

2. coding

3. compiling+running it

4. view the logs, figure out what broke

5. back to 2

AI just makes 2 and 4 happen faster. Which frankly just makes things easier. I don't have to worry about how "modern" my C++ is because it already does it for me. And for debugging, it just does the cmd+clicking through the codebase for me.

So it really just makes it less fatiguing, especially for moving around large bodies of code and renaming stuff. I can spend more time on the essential problem.

I think having multiple of these loops running at once (or having an agent just iterating on its own) is kind of dumb tbh and I don't use them that way. I think having 100 agents running at once or whatever the fuck these people are saying is bullshit. Just using it to speed up 2 and 4 is good enough for me (and also using to explain what the code is doing for building my mental model).

Usually step 3 takes a long time as well, so if claude alone vs claude+me is less accurate, that gets amplified. Another reason why I don't like to let it run all by itself
bayareapsycho
·7 か月前·議論
I don't think the "bubble will burst" in the sense that all of this tooling will go away. Maybe companies shift the focus on running them more efficiently.

I don't think companies will suddenly turn over a new leaf and start hiring juniors like they were in 2020. Even if they tried, within 2 or 3 years they won't be able to as easily.

Fewer people will major in a field where there's some shrieking news headline every 5 seconds about how it's going to get automated out of existence. And the candidate quality will go down, since why bother leetcoding or reading a bunch of textbooks to upskill if there's no ROI associated with it.

If I was an undergrad right now, I totally wouldn't. And I wouldn't have read all of the books that I suffered through to develop understanding. Choosing a major is just betting on the future.
bayareapsycho
·7 か月前·議論
There will be a cliff in the number of people in this industry. There will be a bunch of senior people floating around the job market and no more junior talent. There will be fewer new grads and the pipelines will dry up.

If they go down the route of automating as much as possible, it'll destroy the social pipelines that allow companies to reproduce themselves.

I don't think that'll actually happen, but it'll be interesting to watch
bayareapsycho
·7 か月前·議論
This program was so good, I was most of the way through it before it managed to help me land a better job. But now I have no idea if I'll ever finish it because it's pretty time consuming

The best two classes are AOS and HPC imo. Very grateful to Profs Ramachandran and Vuduc

AOS (and its prerequisite) gives a really strong foundation for working on infrastructure.

HPC pushed me farther than any other class I've done, it's very unique, helped me land my current gig
bayareapsycho
·9 か月前·議論
> Seeing it as a book about psychology that revolves around theological themes, when it is a book about theology that revolves around psychological themes.

I don't think Dostoevsky's Christianity is genuine. It feels about as genuine as Hegel's Christianity. To both of them it's just a convenient prop where their actual ideas take center stage

Every nihilist main character that he writes follows this pattern where they do something really bad, then destroy themselves as some kind of act of penance. This is the only way that conversion happens in his books. But in this case, it's obviously just a way of processing guilt (and reenacting the author's trauma from near execution most likely)

Maybe I'm psychologizing religion too much, but I don't think religious belief is genuine if it's rooted in some kind of (obvious) psychological trauma

The one thing that stands out in Dostoevsky is the psychological depth of the characters, especially in Demons
bayareapsycho
·9 か月前·議論
The funniest thing about events like this is that they're targeted pretty much exclusively to men, but there are plenty of women who'd be more than happy to fish in a pod full of rich guys

So the poly chicks who are running this thing are really just gatekeeping a bunch of rich simps. The "serious" monogamous women have to be kept out or they'll lose their clientele pretty quick
bayareapsycho
·9 か月前·議論
I agree with the sentiment of this but I don't think the root issue is state involvement

The root issue is that the state is involved in the wrong way. It refuses to fund childcare, so marriage is economically incentivized.

But then it makes divorce super financially one-sided and easy to get, so now high earners will try to avoid marriage. But high earners are the ones who make marriage worthwhile, per the above point

Then it has LGBT acceptance in urban centers now, so there's an easy, convenient, and legally protected way for people to avoid the whole marriage pipeline

The only real fix we have for this is to just import more people, who knows whether that will pan out

In general though, I agree with the sentiment that for high earning men it's the shittiest deal imaginable and even the extreme option of dying alone is better
bayareapsycho
·9 か月前·議論
What's the source on this?

I know it's a thing in Canada, Australia, and Washington state (if you cohabitate for some number of years, they can claim that it's a "de-facto relationship" and get property division). Apparently the current UK government is trying to push it

I looked up for Germany and found nothing
bayareapsycho
·10 か月前·議論
It used to be more spread out with a strong valley presence, but over the past couple years they've been force reloing people to Arkansas. It really started to ramp up at about the time I left
bayareapsycho
·10 か月前·議論
In my old org of 80ish, like half of them were from Telegana. All of management was from there. In total, at least 80% of the org was south asian. I guess it's just a coincidence ¯\_(ツ)_/¯. And I can promise you, at least half of them were completely useless. I mean, like so useless they couldn't even figure out how to use generics without 30 minutes of handholding

Also, WMT is not "in tech". Global tech is WITCH tier. The business side is run by the same type of MBA personality running Boeing

They're also forcing like half the company to move to Arkansas at the moment, so a bunch of people are trying to gtfo. I wouldn't advise anyone going there, startups are probably a better option