Honestly in day to day programming I find data types & associated APIs are so so much more important than algorithms.

I would rather work with a flexible data type with suboptimal performance than a brittle data type that maybe squeezes out some extra performance.

Your example of in-place array mutation feels like a good example of such a thing. I feel like there should be a category of interviewing questions for "code-safety" not just performance.

I would rather work with persistent data structures, the least brittle of all, which would also in many cases trivially allow me to parallelize the work, but as far as I can see all the leetcode problems are low level mutation based problems with no clue about functional data structures. Clueless interviewers look to these problems as if they alone epitomized great programming, while they are often inflexible single core stuff, that may not even be appropriate for this day and age any longer.

In case 5: Photos of Yourself in Different Eras

The output just looks like a clearly different person. Its difficult to production-ize things that are inconsistent.

Personal opinion:

MCP for data retrieval is a much much better use case than MCPs for execution. All these tools are pretty unstable and usually lack reasonable security and protection.

Purely data retrieval based tasks lower the risk barrier and still provide a lot of utility.

Are the tags themselves AI generated? they are quite funny in how specific they get e.g. "gaming culture" as the tag.

Also some tags include quotes and other don't haha.

But yeah you should respond to the other comment asking someone to use an LLM to show the trend in AI topics.

> Are the tags themselves AI generated?

Yes. The full explanation's on my blog and the code's public in my GitHub repo btw.

The main nuisance is that it will use inconsistent synonyms, so an article might be "ai" or "artificial intelligence" or "llms" without obvious differences in the topic.

When I get around to the stat-tracking I might do a Show HN or something, although it feels a bit frivolous for that.

> Also some tags include quotes and other don't haha.

Yeah, that kind of thing is a nuisance too. I'll add some more scrubbing rules to clean questionable characters out when I get back.

Repo: https://github.com/dcminter/hntags.com

Interesting! Good to know other people's experiences, yeah maybe my sampling is just off

I still pretty much use it daily-ish but its more to verify something from Gemini than to find entirely new information.

YES, this is the way to make AI tools pretty much a strictly positive productivity tool on large codebases.

After bouncing off of Unity the first few times, I finally decided to commit to it after trying out some of the JS game frameworks (like Phaser)

Absolutely worth it, have made lots of games in Unity just for myself that feel pretty polished, there are just so many systems to make a game work.

The advice around game engines kind of seems like "to learn how to write programs first create the compiler.

Not to say all games should be made in engines but it certainly helps.

In a non-smug kind of way sometimes I just wonder if they types of problems I work on are just harder (at least for an LLM) than a lot of people.

Currently working at a FAANG on some very new tech, have access to all the latest and greatest but LLMs / agents really do not seem adequate working on absolutely massive codebases on entirely new platforms.

Maybe I will have to wait a few years for the stuff I'm working on to enter the mass market so the LLMs can be retrained on it.

I do find them very very useful as advanced search / stack overflow assistants.

I agree there are good examples of 90% being good enough but what you purposed doesn't sound like a good one.

This assumes that AI can't also introduce new bugs into the code causing a negative.

A case of 90% being good enough sound more like story boarding or giving note summaries.