Conversation
Notices
-
Embed this notice
it's interesting to think about. When farming took 99% of time, all people did was farm. Maybe they complained that the machines weren't as good and yield and quality would drop?
What can the computer chuds start doing when finally not every little stupid thing takes 2 weeks? Where to spend effort?
There is this relationship in the real world, where very professional people tell you "make it work first, make it good later" / do a vomit draft / just tape a cardboard tube to a widget as a handle to see if you like it, etc.but with software, common wisdom was if you don't get the fundamentals right, you will end up in trouble. there never was an identical draft to polish path for coding. reimplementing your entire UI in a better-suited lib took ages, for example.
AI might align effort more with the Pareto curve
Of course maybe those kinds of people were always wrong and you should "just get better at writing code" and then a rewrite wouldn't have been that big of a deal, but even if so, maybe AI will help people realize that.
RT: https://shitposter.world/objects/2cccb614-758f-4263-a10b-feb95cc9d8e0
-
Embed this notice
@WandererUber there's two different categories. one of them is the programs where the demo you shit out is good enough to be the final program and MAYBE you bolt on enough professionalism on like, basic security. This fits it, it's a personal program so it only ever runs on my localhost or maybe on a web server locked behind a passkey so I can use it on my phone.
AI has absolutely unshackled my wrists and enabled me to pound out a hundred little itch-scraching programs that make my life measurably better. I can't even fathom someone looking at that and still being like "show me the useful software"
I hated every japanese learning program I used. so I wrote my own that works the way I like, and it's just so...liberating of a feeling.
-
Embed this notice
@sun lain talks a lot about this type of software that is just for him.
that's an "unscratchable itch" for commercial software because they have to do one-size-fits-all every single time. Complete non-issue now!
-
Embed this notice
@WandererUber between this and, AI being able to reverse-engineer proprietary software so you can modify and interoperate with you, I do strongly consider AI to be liberation technology when used correctly.
-
Embed this notice
@sun @WandererUber Any technology can lead to a golden age if placed into the right hands and a dark age if placed into the wrong ones. Nothing escapes morality and philosophy, no matter how hard people try to make the world a secular or amoral place.
People treat technology like it's a cold, objective, clinical thing, divorced from the human element, but ultimately all technology is just a bunch of people working together on stuff based on previous stuff a bunch of other people worked on in the past. All technology is sociology at the end of the day.
-
Embed this notice
@LivingSpaceStudios @WandererUber I am quite worried about the companies that are making the AI but the product itself is insanely powerful
-
Embed this notice
@sun @WandererUber I agree on both points.
>I am quite worried about the companies that are making the AI...
That too, is a problem of philosophy and sociology more than science or engineering. Everything comes back to who's in charge and why.
-
Embed this notice
@WandererUber @sun one big issue used to be that 'doing it right' took exactly as long as becoming wise about your problem domain, so total rewrites were mostly impractical. that's not the case now, a total rewrite of a system that took 10 years to build can be done in a few weeks.
karpathy said the LLMs show that there were always two kinds of programmers, 'builders' and 'coders'. the builders now use the new tool, they usually didn't dislike programming, but if they can make superpleroma 2 without writing a single line, they'll gladly do it. the coders really care about the coding itself. i sort of agree with that split.
'just get better' really doesn't work here, because a deep understanding of the problem space only comes from actually engaging with the issue. there's no 'just plan ahead moron', humans can't handle 'make no mistakes' either.
i think that's also something you see with projects like opencode: before, the issue would have been to write the code for it. now, the problem is to curate the feature set and to figure out where you actually want to be in that space. so you might have 1000 PRs open, each with a good feature, but still not want to merge a single one.
-
Embed this notice
@loonycyborg @WandererUber @sun the way you derive that would also make it impossible to ever work with another human without personally checking everything they are doing
-
Embed this notice
@lain @WandererUber @sun
I understand that LLMs are very good at parsing natural language. But the thing is the only reason I got into coding in the first place is that I don't like specifying problems in natural languages. That's whole reason I think that "vibe coding" is fscked up idea. It fully replicates the sad experience of working with another human being using same inexact language. Maybe an LLM can shorten the time spent to just get overview of some area but you need to do full research based on upstream data anyway to actually finalize things any reliably.
-
Embed this notice
@lain @WandererUber @sun I'm currently rewriting a 10+ year old application from big ball of mud into elixir wirh proper DDD and instrumentation and telemetry and all the buzzwords. Or more accurately, I built out the infrastructure and factory line so that I could effectively dupe this app using context from as many angles as possible. A few weeks is pretty generous (on speed), if you actually need to human validate an application this large then it might take longer. but in principle? I'm literally doing it rn
-
Embed this notice
@loonycyborg @lain @sun >But the thing is the only reason I got into coding in the first place is that I don't like specifying problems in natural languages.
oh okay. Well I personally like when I can make the computer do what I want, that's why I got into it.
-
Embed this notice
@WandererUber @loonycyborg @sun i guess that's at least similar to the split karpathy mentions.
-
Embed this notice
@loonycyborg @lain @WandererUber everybody should just use ocaml
-
Embed this notice
@lain @WandererUber @sun
In any sort of craft there exist various operational standards and common sense rules. Those replace constant checking of everything. They're like type systems in programming languages. So when I use actual programming language rather than imprecise natural language instructions I'm, like, automating those standards and rules.
-
Embed this notice
@lain @loonycyborg @WandererUber @sun I'm auditing over your shoulder as we speak. You need a tactical solid aluminum milled keyboard like I've got so you know nobody knows you're copying configs
-
Embed this notice
@loonycyborg @WandererUber @sun i don't see how this relates to the post you answered to
-
Embed this notice
@lain @WandererUber @sun i'm a pwofeshional
-
Embed this notice
@shibao @WandererUber @sun yeah, it's not a fantasy. it's happening right now, actual professionals are doing it.
Wanderer über dem Nebelmeer
Blurry Moon
Living Space Studios :verified:
lainy
loonycyborg
shibao
anime graf mays ?️?
All GNU social JP content and data are available under the