2026/01/04

Newest at the top

2026-01-04 19:18:29 +0100confusedalex(~confuseda@user/confusedalex) confusedalex
2026-01-04 19:17:49 +0100 <gentauro> monochrom: xD
2026-01-04 19:17:00 +0100merijn(~merijn@host-cl.cgnat-g.v4.dfn.nl) (Ping timeout: 245 seconds)
2026-01-04 19:16:53 +0100confusedalex(~confuseda@user/confusedalex) (Remote host closed the connection)
2026-01-04 19:12:15 +0100merijn(~merijn@host-cl.cgnat-g.v4.dfn.nl) merijn
2026-01-04 18:45:55 +0100karenw(~karenw@user/karenw) (Ping timeout: 264 seconds)
2026-01-04 18:44:24 +0100 <humasect> (=
2026-01-04 18:44:06 +0100cyphase(~cyphase@user/cyphase) cyphase
2026-01-04 18:40:13 +0100 <EvanR> a compiler is really really narrow expert system xD
2026-01-04 18:34:57 +0100nek0(~nek0@user/nek0) nek0
2026-01-04 18:34:07 +0100merijn(~merijn@host-cl.cgnat-g.v4.dfn.nl) (Ping timeout: 240 seconds)
2026-01-04 18:29:19 +0100merijn(~merijn@host-cl.cgnat-g.v4.dfn.nl) merijn
2026-01-04 18:29:15 +0100machinedgod(~machinedg@d75-159-126-101.abhsia.telus.net) machinedgod
2026-01-04 18:22:31 +0100dolio(~dolio@130.44.140.168) dolio
2026-01-04 18:20:37 +0100 <humasect> =)
2026-01-04 18:17:53 +0100merijn(~merijn@host-cl.cgnat-g.v4.dfn.nl) (Ping timeout: 256 seconds)
2026-01-04 18:14:52 +0100 <monochrom> We don't have to be pessimistic about AI. But we can be rightfully pessimistic about how humans use AI. >:)
2026-01-04 18:13:09 +0100dolio(~dolio@130.44.140.168) (Quit: ZNC 1.10.1 - https://znc.in)
2026-01-04 18:11:17 +0100merijn(~merijn@host-cl.cgnat-g.v4.dfn.nl) merijn
2026-01-04 18:06:36 +0100housemate(~housemate@202.7.247.155) housemate
2026-01-04 18:04:27 +0100 <haskellbridge> <Liamzee> and you don't get to C / Algol by putting more labor / clock cycles on Fortran
2026-01-04 18:03:45 +0100 <haskellbridge> <Liamzee> the problem right now, is that AI codegen is not even at the C or Algol level of maturity
2026-01-04 18:03:24 +0100 <haskellbridge> <Liamzee> still needed, but very few people use it because it's so labor inefficient
2026-01-04 18:03:20 +0100 <humasect> slave work
2026-01-04 18:03:06 +0100 <haskellbridge> <Liamzee> i think traditional programming as is right now is becoming like assembly
2026-01-04 18:02:45 +0100housemate(~housemate@202.7.247.155) (Quit: I am over the moon, although it is day time so I am probably on the other side of the world, and upside down which would make me the right way up.)
2026-01-04 18:02:34 +0100 <humasect> +prolog
2026-01-04 18:01:36 +0100 <haskellbridge> <Liamzee> gentauro: I think something in the lean / agda / haskell tradition will survive and become dominant after the AI bubble pops
2026-01-04 18:01:22 +0100 <humasect> x_x
2026-01-04 18:01:07 +0100 <haskellbridge> <Liamzee> ncf: think of AI tokens as unityped =3
2026-01-04 18:01:06 +0100 <gentauro> only hope I have is that MS LEAN takes over the (AI) world …
2026-01-04 18:00:20 +0100CiaoSen(~Jura@2a02:8071:64e1:da0:5a47:caff:fe78:33db) (Ping timeout: 245 seconds)
2026-01-04 18:00:18 +0100 <gentauro> far from it :(
2026-01-04 18:00:15 +0100 <gentauro> I spent many months last year doing that. It's not good
2026-01-04 18:00:07 +0100 <haskellbridge> <Liamzee> I'm just saying it'll be how software development will survive, a switch from imperative or concrete programming to declarative and verified programming
2026-01-04 17:59:59 +0100 <gentauro> Liamzee: At some point, when the bubble pops, somebody will have to "fix/maintain" all the AI-slop.
2026-01-04 17:59:49 +0100chromoblob(~chromoblo@user/chromob1ot1c) chromoblob\0
2026-01-04 17:59:46 +0100 <haskellbridge> <Liamzee> i mean natural language is bad at specifying what you want the AI you want to do
2026-01-04 17:56:15 +0100cyphase(~cyphase@user/cyphase) (Ping timeout: 240 seconds)
2026-01-04 17:46:49 +0100jmcantrell_(~weechat@user/jmcantrell) (Ping timeout: 264 seconds)
2026-01-04 17:43:01 +0100CiaoSen(~Jura@2a02:8071:64e1:da0:5a47:caff:fe78:33db) CiaoSen
2026-01-04 17:42:00 +0100 <ncf> i mean you can think of a horse as a unicorn if you like, but you don't need to tell us
2026-01-04 17:41:37 +0100 <darkling> c
2026-01-04 17:40:14 +0100jmcantrell_(~weechat@user/jmcantrell) jmcantrell
2026-01-04 17:38:44 +0100 <humasect> or supersearch
2026-01-04 17:38:27 +0100 <haskellbridge> <Liamzee> you can think of AI as a "supercompiler" albeit one that's extremely bad at what it's supposed to do
2026-01-04 17:37:51 +0100 <haskellbridge> <Liamzee> gentauro: you don't need to be so negative
2026-01-04 17:30:35 +0100humasect(~humasect@dyn-192-249-132-90.nexicom.net) humasect
2026-01-04 17:30:35 +0100merijn(~merijn@host-cl.cgnat-g.v4.dfn.nl) (Ping timeout: 240 seconds)
2026-01-04 17:26:02 +0100merijn(~merijn@host-cl.cgnat-g.v4.dfn.nl) merijn