[lhnz]

I think this is very unfair. In the few months that I was looking, I put in more than 300 hours of study. It wasn't superficial and I didn't attempt to memorize anything — I wanted to understand fundamental techniques, and put a considerable amount of effort in to doing this, treating it as if it were my job.

Nobody owes me anything but that doesn't mean that I'm not allowed to point out that this is the weakest I've ever seen the market in my decade or so of experience of it. It's very tough for those that have been laid off. Particuarly those have to support families as I did -- I am a sole earner and have two children.

You could benefit from practicing a bit of empathy and not lying to yourself that the market has merely moved back to "sanity" without gaining a bit of recent experience in it.

[ctvo]

> I put in more than 300 hours of study. It wasn't superficial and I didn't attempt to memorize anything — I wanted to understand fundamental techniques, and put a considerable amount of effort in to doing this, treating it as if it were my job.

If you landed a job, do you think this new ability to implement a CS 101 data structure would let you keep it? Help in your performance in a measurable way? This is what I mean by superficial. The same applies to the LLM work. You studied enough to where you could buzzword it in a 45 minute interview.

My response, which I'll be more clear with: Maybe it's not that bad that we're more careful and these superficial techniques are no longer adequate to get someone a job in this industry.

> I'm not allowed to point out that this is the weakest I've ever seen the market in my decade or so of experience of it.

Don't think anyone is preventing you from doing anything.

[lhnz]

I did land a job making applications for an HFT firm and while I don't think that being able to implement data structures or write my own algorithms is necessarily what will allow me to keep this job, recently I somewhat regularly use the skills I pick up doing competitive programming exercises in my workplaces. Therefore, my answer to your question about whether it will help my performance in a measurable way is: yes.

[jonnycoder]

Leetcode is not superficial when nearly all companies are using leetcode as a barrier to entry these days. Many of us with 10+ years of experience did not need to practice leetcode to get hired originally, but many would fail miserably today due to the time constraint and necessity to socialize/speak the problem while solving it in 45 min. Some practice problems took me all day at first, but after a while it became a skill of quickly identifying the type of data structure to use and solving it. It's great practice.

As for LLMs, it's probably the best thing to learn right now, especially by building something. I personally believe it will lead to an explosion in potential jobs. It's hard to describe but imagine a software engineer learning how to create a rest api for the first time after doing soap. And then integrating various apis to harness saas products, cloud deployments, etc back in the 2000s.

Then again I also talk to an ex-googler I met on Hacker News who built an LLM app and is now spending most of his time selling his saas product because he can't find a job either. That's a very positive go-getter attitude.

[hasty_pudding]

Are you seriously saying practicing data structures and algorithms to improve your skills is 'superficial'?