Newest at the top
| 2026-01-27 23:59:22 +0100 | tomsmeding | shakes fist at Rembane's shaking fist |
| 2026-01-27 23:59:10 +0100 | Googulator39 | (~Googulato@2a01-036d-0106-030a-3891-da7f-f3f3-f997.pool6.digikabel.hu) |
| 2026-01-27 23:58:48 +0100 | tomsmeding | too |
| 2026-01-27 23:58:40 +0100 | Rembane | shakes fist at Google |
| 2026-01-27 23:58:39 +0100 | <jreicher> | Yep, I've been using them for searching. But as soon as I feel like I need to learn an underlying concept that might be new for me, I follow the link to the source. |
| 2026-01-27 23:57:57 +0100 | <tomsmeding> | whereas putting "@Int" in a search engine is guaranteed to fail |
| 2026-01-27 23:57:42 +0100 | <tomsmeding> | LLMs are an amazing search engine: if you want to know what weird GHC extension gives you `foo @Int` syntax, for example, it's essentially guaranteed to be able to tell you "that's TypeApplications", and then you can look up the docs for that |
| 2026-01-27 23:57:14 +0100 | shr\ke | (~shrike@user/shrke:31298) shr\ke |
| 2026-01-27 23:57:14 +0100 | shr\ke | (~shrike@user/paxhumana) (Changing host) |
| 2026-01-27 23:57:14 +0100 | shr\ke | (~shrike@user/paxhumana) paxhumana |
| 2026-01-27 23:57:05 +0100 | shr\ke | (~shrike@user/shrke:31298) (Read error: Connection reset by peer) |
| 2026-01-27 23:56:46 +0100 | <jreicher> | I'll concede that's possible, but I think I'd like to see it. |
| 2026-01-27 23:56:29 +0100 | <tomsmeding> | it can't make it better, but it can make it more personalised |
| 2026-01-27 23:56:18 +0100 | <jreicher> | Yes, but summarising well written teaching text can't make it better. It was already well written. It's only going to make it worse. |
| 2026-01-27 23:56:10 +0100 | <tomsmeding> | as you can see, I also don't agree that learning just from an LLM is a good strategy, but "teaching is not what is does" is slightly overly reductive |
| 2026-01-27 23:55:42 +0100 | <tomsmeding> | there's teaching content on haskell and other FP topics out there, and that's part of the "language that exists" |
| 2026-01-27 23:55:18 +0100 | <jreicher> | LLMs model language that exists. Teaching, by definition, requires modelling ignorance. |
| 2026-01-27 23:54:41 +0100 | merijn | (~merijn@host-cl.cgnat-g.v4.dfn.nl) merijn |
| 2026-01-27 23:54:18 +0100 | <jreicher> | Guest41: chatgpt won't teach you anything. That's not what it does. |
| 2026-01-27 23:54:07 +0100 | <ski> | one could view it as taking "separate internal machinery from UI presentation" to a higher degree |
| 2026-01-27 23:53:22 +0100 | <Guest41> | yeah lots of unlearning, separating IO for example is very weird :laugh: |
| 2026-01-27 23:53:07 +0100 | <Guest41> | so far I've made conways' game of life, snake, a simple sat solver, and now i'm working on a stochastic parrot |
| 2026-01-27 23:53:04 +0100 | <tomsmeding> | keep thinking |
| 2026-01-27 23:52:53 +0100 | <tomsmeding> | then again, learn as you wish, as long as you have fun it's probably fine :) |
| 2026-01-27 23:52:41 +0100 | <ski> | you will have to unlearn old habits |
| 2026-01-27 23:52:37 +0100 | <Guest41> | I always try to make small little projects that increase in complexity and that's how I learn to program in new languages |
| 2026-01-27 23:52:31 +0100 | <ski> | some things will be different, sometimes very different. some things will carry over. but it's better to defer comparisions until you've got the basics covered |
| 2026-01-27 23:52:17 +0100 | <Guest41> | yeah maybe just having a student mindset instead of trying to blaze thru would help |
| 2026-01-27 23:51:57 +0100 | <tomsmeding> | as you get a little experience with functional programming, you'll start to see the connections to what you've been doing so far and you'll be able to use all your knowledge |
| 2026-01-27 23:51:52 +0100 | <ski> | coming at it with a mindset that you don't know programming yet, helps, yes |
| 2026-01-27 23:51:25 +0100 | <tomsmeding> | you should expect to feel like a programming noob, at least for a little while :) |
| 2026-01-27 23:51:08 +0100 | <tomsmeding> | having much programming experience is of course very helpful, but still, programming in haskell requires a rather different mindset than for more traditional imperative languages |
| 2026-01-27 23:51:02 +0100 | <ski> | most of those are "mostly the same" (not counting Verilog), when comparing to Haskell |
| 2026-01-27 23:50:48 +0100 | <Guest41> | Huh actually <https://www.inf.ed.ac.uk/teaching/courses/inf1/fp/> is the course for the textbook I do have, hooray! |
| 2026-01-27 23:49:55 +0100 | mange | (~mange@user/mange) mange |
| 2026-01-27 23:49:52 +0100 | <Guest41> | but you're right I can't just google all the output and hope to double check everything, the small hallucination risk sucks there |
| 2026-01-27 23:49:30 +0100 | <Guest41> | been programming for a long long while in C,C++,python,verilog,etc. so I can "kind of" tell when it's lying |
| 2026-01-27 23:49:28 +0100 | <ski> | when you're just learning, you can't |
| 2026-01-27 23:49:19 +0100 | <ski> | when you know more about an area, you can (more) tell what is nonsense, and what isn't |
| 2026-01-27 23:48:48 +0100 | <ski> | when you're learning something, you don't want to deal with the risk of hallucinations |
| 2026-01-27 23:48:36 +0100 | <tomsmeding> | (you don't know yet which of its output is useful and which is not) |
| 2026-01-27 23:47:47 +0100 | ski | wouldn't learn anything using an LLM chatbot |
| 2026-01-27 23:47:43 +0100 | <Guest41> | but maybe I should go along it with the course as well :-) |
| 2026-01-27 23:47:31 +0100 | <Guest41> | I have this one : Introduction to Computation: Haskell, Logic and Automata |
| 2026-01-27 23:47:22 +0100 | <lambdabot> | <https://github.com/byorgey/haskell-course>,<https://www.seas.upenn.edu/~cis194/spring13/lectures.html> |
| 2026-01-27 23:47:22 +0100 | <ski> | @where CIS194 |
| 2026-01-27 23:47:18 +0100 | <ski> | or, perhaps, a course, like |
| 2026-01-27 23:46:42 +0100 | <ski> | how about learning from a textbook ? |
| 2026-01-27 23:46:06 +0100 | EvanR | (~EvanR@user/evanr) (Remote host closed the connection) |
| 2026-01-27 23:45:50 +0100 | <Guest41> | Does anyone else find haskell hard to learn using chatgpt, I can get working code of cute snippets and programs but I feel like if OpenAI went bankrupt I would be useless lol, or would take forever to type down |