Spyke

Replies

Comment on

LLM Idolatrine: Why no one wants your agentic PRs

With more and more places pushing LLM generated code I'm getting real review fatigue. Code reviews used to be very reasonable and you could be very mindful about other persons code. Now it's starting to be just LLM generated code which makes reviewing tedious as you are reading the same sloppy stuff in hopes that you can build your mental model of the changes to codebase before new LLM slop bomb hits the PR queue.

Comment on

The case against immersion style of language learning

I think there's a little bit of a confusion of what the immersion here means and that it's somehow only teaching in the target language. What usually is meant with immersion is that the learner has constant exposure to the target language, not just in the classroom. This would mean that the people you daily interact with and your daily routines would have continuous instances where you have to use the target language.

When you're quoting some professor, they're absolutely right that many educational institutes have already structured the information and created easily approachable courses that help learners to learn quicker, but with language learning you have to really ask what is your goal. For a lot of people they are completely fine to "learn" a language to a point where they can read something and translate it to their native language. But for more serious learners the end goal is that they're able to think and use the langage fluently without any intermediate translation happening in between and for that only way to really practice is to use the target language.

You're then bringing up a memory aids for learning languages and that you can use that to quickly convey information and again it's absolutely correct that for an English speaker, having memory aids and things told in English is much more easier to remember, but it doesn't actually help you think in French. Sure you may pass some exam easier, but it's not very useful long term.

Teaching in the target language is very much preferred, except for complete beginners. If you are taking class for complete beginners, it's very helpful to have key grammar topics and core vocabulary taught in a common language, but after that learners should be encouraged to use the target language at all times. Rather than switching to common language to ask what something means for example, it's much better to ask in target language and get it explained in it as well as it reinforces what you have already learnt. Sure this might feel slower, but good instructors shouldn't have issues to impart grammar or vocabulary to a learner.

Comment on

Things you didn't know about indexes

Reply in thread

Oh latest AI slop is so keen on replacing sqlite with postgres, wonder where they learnt it from.

Perfect! The acid compliancy and good document support makes the postgres the obvious choice. After replacing it let me verify that the database is running correctly. psql -c "SELECT * FROM cat_pictures". Hmm it's not working. Let me verify that the server is running. psql -U root. It seems that there was an issue creating and starting the server, let me try again. cd projectfolder / && rm -rf && initdb. Hmm couldn't iniate the database let me add --no-preserve-root to ensure initdb runs properly

Comment on

LLM Idolatrine: Why no one wants your agentic PRs

Reply in thread

Exactly.

Another problem with LLMs is that they are actually useful in some tasks and they can generate very good quality code if you're diligent enough developer. I also have built personal tools with them, but I don't have the knowledge of the code the LLM has hallucinated which means that before I would push this code forward I will have to basically familiarise myself with the code in a way how a code review works.

The knowledge you gain from this is also different from that of actually writing and running the code yourself. I have seen people who use LLMs to write commit messages which is the last thing you should do. Commit messages are probably the only places were we can meaningfully store the knowledge gathered during development and the more I see LLM commits the more I lose hope.