Submitted 6 months ago by bridgeenjoyer@sh.itjust.works to showerthoughts@lemmy.world
I can see someone in the future watching a program run and asking “wow, is that ai? A PERSON typed those cryptic letters? No way!”
I think it’s like photo editing where you’re manually using tools like “remove background” but the tools are semi-intelligent and kinda get what you want and do it for you.
Code is the highest level from which deterministic output can be derived.
It is the unambiguous spec of the application.
I’ve seen talks from people at openAI talking about how you should treat prompts as specs, and that a good spec can generate the code you need, so the important thing to keep is the spec (prompt).
But LLMs are necessarily non-deterministic, it’d be like every time you compiled the code, you got a different app that fulfilled the same purpose. That’s not really useful.
And if you make the spec unambiguous enough that it always generates undistinguishable apps, well then congratulations, your prompt is just as complex as the code is, except you don’t have any kind of static checking and compiling takes orders of magnitude longer. You might as well just be writing code.
Now will coding look very different in the future? Maybe. I hope not, but it’s looking like it will.
But I don’t think we’ll get to a place any time soon where people write prompts and never fuck with code.
LLMs are necessarily non-deterministic,
There’s nothing about LLMs that force them to be non-deterministic. Given the same prompts and context, they will always output the same probability distribution. It’s then up to you what you decide to do with that distribution. If you decide to always choose the most likely output, then the entire thing will be deterministic. We just don’t do that because it’s less useful than stochastic output.
Yes, that’s why it’s necessary for it to be non-deterministic. Without non-determinism, there is no error recovery if it chooses the wrong token somewhere in the middle of the completion. That’s much less useful.
should treat prompts as specs, and that a good spec can generate the code you need
"Tell me you've never compared what user said they needed vs what they actually needed, without actually saying it"
I fully agree with you, and just like to add my take on how developing will change in the near future:
I think AI can help developers speed up their work. Its easier to find information from a documentation when you can just straight up ask the documentation about something, rather than having to research and find the correct paragraph/term yourself. I don’t think AI will replace documentation, just help you sift through it. Also tools like copilot are pretty useful, I think. It just helps speeding up the process of getting your ideas into code.
Not only does it not speed up work, it actually slows down work: metr.org/…/2025-07-10-early-2025-ai-experienced-o…
If all you’re using it for is to find the relevant parts of the relevant documentation, then a vector search would do as well.
What an LLM can do is synthesize readable docs out of poor or missing docs. When it isn’t hallucinating, that is.
But I don’t actually see anyone using LLMs just to cut through the docs, they’re using it to code. And the results I’ve seen are pretty mixed. It does seem to help ramp up in a new area, but it also seems to become an impediment to moving past the ramp-up phase.
Beginners can now produce beginner code at 2x the speed, and senior folks can now produce beginner code at 3x the speed. But nobody is producing senior level code.
You mean that less people need to understand code? Like that’s just natural progression. There are many coding languages because hardly anybody knows how to code in machine language or punch cards and we’ve been making coding easier for humans. The progression to natural language is a natural one.
LLMs bridging the gap between coding and natural language isn’t going to be a mystery in the future. Just like we don’t go like “a person punched all those holes in the cards? no way!” Because yes they did and we all understand that that was something that needed to be done at that time. We simply appreciate the amount of work it took. It’s a lot more that typing ‘cryptic letters’, which we will all know in the future also as coding. i don’t understand why you believe everybody in the future is an idiot.
I think you much overestimate the intelligence of the average human.
I was just saying that if you show a computer program run something (let’s say a choose your own adventure text game or something) to an average 12 year old today, its likely they would say, thats ai! Because they probably wont even know that a human could write code themselves to do something like draw a circle on screen.
Also, I would have loved to learn actual assembly and talk directly to the hardware. But like you said there is no use at all for it nowadays so thats going to be a lost skill.
I don’t know why we should take 12 year olds as an example? Surely you didn’t know certain things at 12 that you know now. Assuming you are 13+
I’m just saying that things will trickle into common knowledge. And there will always be people who know jack shit. Obviously.
But when you play music from a streaming service you don’t go like. Wow people had to rewind cassette tapes! No, you understand the progression music carriers have made and just enjoy the music. You could buy a tape recorder and tapes and they are still available, if you would want. The same applies to coding, and other technologies.
You can still lessen assembly, it’s not that hard actually, just tedious. It can play TIS-100 if you want a “fun” way to learn it.
I don’t see the connection here. AI describes a type of software and programming is something you do. Wether a software counts as AI or not depends on what it does, not who wrote the code.
I’m very concerned about GIGO “garbage in garbage out” because AI needs humans to write the code it “learns.”
"After all, it was one of the first sectors to deploy A.I. programming in the 1980s, with the four ghosts who chase Pac-Man each responding differently to the player's real-time movements."
All the lines look blurry when you're squinting at things from a position of complete ignorance.
yeah, we should talk about “probabilistic” vs. “deterministic” behavior instead.
In the not too distant future…
NPC: “ChatGPT told me I got ghosts in my blood and I better inject bleach bout it”
AI Corpo: “thankfully, no human programmed it to respond with that, so we are not liable”
Fascist court: “NPC’s family must pay AI Corpo damages for negative publicity”
I laughed and cried too much about this.
littleguy@lemmy.cif.su 6 months ago
No it won’t be.
Programming will always be giving instructions to execute a task.