Comment by with
2 months ago
the real takeaway is buried at the bottom: "the magic isn't in the input, it's in the system around it." random keystrokes producing playable games means the input barely matters anymore. we're basically at the point where the engineering is in the scaffolding, not the prompting.
> we're basically at the point where the engineering is in the scaffolding, not the prompting.
This still required prompting, and not from the dog. Engineering is still the holistic practice of engineering.
That also shows the delusion of some people that believe their vibe coded projects have any value.
If generative AI improves at the rate that is promised then all your "promting skills" or whatever you believe you had will be obsolete. You might think you will be an "AI engineer" or whatever and that it is other people that will lose their job, that you are safe because you have the magic skills to use the new tech. You believe the tech overlords will reward you for your faith.
Nope. You are just training your replacement.
No one will buy your game that you vibe coded. If the tech were good enough to create games that are actually fun then they would just generate their own games. Oh your skill? Yeah, a dog can do it.
Yes people will cope by saying but oh the whole initial prompt and setting it all up was still hard but yeah currently. The tech will improve and it will get more accessible. So enjoy the few months you are still relevant.
Of course there is reason to believe that you can't scale up LLMs endlessly and bigger models hit diminishing returns. In fact we might already be seeing this. So there is an upside but then again when the AI bubble pops and the economy crashes you will be out of a job all the same.
Sounds great to me. Software devs might lose their jobs but billions of people will be empowered to spin up whatever software they need on demand. This is the future I dreamed of when I was a kid, and I'm not so cynical as to let the dying of a trade sour me to this objectively incredible technology.
Nobody gives a damn about the dying of a trade. People don’t want their house foreclosed on when they lose their income, or their cancer to kill them when they lose their health insurance, to move an elderly parent into a cheap shitty old folks home because they can’t afford home health care, or not be able to pay for their kid to go on that school field trip.
This would all be pretty fucking swell if the fundamental problems this could cause were even considered before hitting the gas. Instead, you’re going to have a shitload of people with ruined lives, but as a consolation prize, they can vibe code stuff! Wowee!
17 replies →
> This is the future I dreamed of when I was a kid, and I'm not so cynical as to let the dying of a trade sour me to this objectively incredible technology.
I feel that you should take a longer-term view of things...
If an AI can vibe code from the requirements of the average white-collar worker, we're not talking about the death of a trade. Or even two trades. We're talking about the death of almost all white-collar jobs.
Development paid a lot more than other white-collar work because it was harder, and fewer people could actually do it. How fast do you think the easier work will get replaced if the hardest one is replaced? For the remaining white-collar roles that consist solely of skills achievable by a border collie, how much do you think they'd pay?
1 reply →
> Billions of people will be empowered to spin up whatever software they need on demand.
… so long as they have the money, and the power grid survives the overtaxation.
After all, why bother encouraging a culture where people are genuinely empowered to tweak and create their tools? Why encourage a culture of exploration, of playful cleverness? What use is there to being a hacker, of sharing knowledge?
It’s definitely much easier, more sustainable, and more fulfilling to have server farms adjacent nuclear reactors make your calculator app for you.
4 replies →
>billions of people will be empowered to spin up whatever software they need on demand
As long as: 1. They have access to a computer 2. They have affordable access to a capable language model 3. Someone will actually care about using their output instead of simply spinning up their own custom version of whatever idea they have
The number 3 is something many people miss, especially on HN: Why would I want to use YOUR software if it's easy for me to cook up my own? Perhaps out of efficiency or lack of time, in the same way I order pizza instead of baking my own when I'm tired or can't be bothered to bake pizza.
Then the software becomes truly throwaway, in the same way takeaway is, and everything is a greenfield project because rewrites are literally easier and faster to make than patching up existing stuff.
2 replies →
Shills like yourself indeed need not worry. You just find a new scam to shill.
sure, a bunch of people will lose jobs, but at that trade off everyones dog can vibe code Royal Frog, a 4 level unwinnable game where play as king frog, eating peasant flies.
+ Also the fact that the Memory.md file was a hindrance to the quality of output
Depends on the desired output. The author wanted variability, for which Memory.md was an obstacle. Another project might need consistency.
In this case yes, but the real takeaway is to pay attention to Memory.md. If I had a particular game in mind and it latched on early to a style I didn't like, there's no guarantee it would update the memory as I try to change the style.
> the engineering is in the scaffolding, not the prompting
Well, yes. Feeding random tokens as prompts until something good comes out is a valid strategy.
Simulated annealing for game design
Not that I condone any form of gambling but I would rather play actual slot machines instead of spending hundreds of dollars on tokens in hopes that the AI blesses me with anything useful.
This sort of thing is usually done on the frontier of LLM research. (Role play with local models.)
I think this misses something. The output here is something not the thing. Yes the scaffolding is important, but the requirements are even more important. You need crystal clear requirements + great scaffolding and then the implementation becomes mechanical.
This is amazing because it's the same logic and argument about how to do good software engineering that's been around for 40 years. If you just write good enough requirements, a good enough, detailed specification, then your software team can't fail, even if they are low-cost engineers from a developing nation. It's the classic Waterfall method.
That was totally upended by agile, that emphasized that yes, a clear, unambiguous specification is needed, and the best language for that is a programming language. Don't waste time writing a detailed spec in English, get right to writing it in code that you can execute and get immediate feedback on.
Now people want LLMs to write the code for them, so they are back to saying we just need to give the LLMs clear enough direction, a clear specification. It's amazing to witness history not exactly repeat itself, but very clearly rhyming
Sorry but what do you mean crystal clear requirements?
I don't particularly think "y7u8888888ftrg34BC" would pass as a crystal clear requirement at my workplace :<
Do you mean something different?
> y7u8888888ftrg34BC
This is more information than the average users gives you when requesting new features.
I mean you get a random game in the authors example :) But in real life you do not want a random game. That's what I mean, you need the great scaffolding + exact requirements. Then the prompt to do the implementation does not matter too much.
3 replies →
this would be a more insightful comment if the output wasn't itch io shovelware.
> the magic isn't in the input, it's in the system around it.
It isn't [this], it's [that]. Is AI slop, just saying.
[flagged]
> the "intelligence" was never in the input It's quite literally in the authors prompt so in the input. it's in the article that without his prompt the gibberish input produces nothing of value:
"Hello! I am an eccentric video game designer (a very creative one) who communicates in an unusual way. Sometimes I’ll mash the keyboard or type nonsense like “skfjhsd#$%” – but these are NOT random! They are secret cryptic commands full of genius game ideas (even if it’s hard to see).
Your job: You are a brilliant AI game developer who can understand my cryptic language. No matter what odd or nonsensical input I provide, you will interpret it as a meaningful instruction or idea for our video game. You will then build or update the game based on that interpretation."
Also I don't know if you're an LLM or not but can we please not chatGPT-ify our comments like this? It figuratively makes me want to punch you through the screen.
The parent poster has already been called out at least once for commenting in form that reads like AI generated slop.
In fact, their only post that doesn’t read like AI generated content is the one reply to where they got called out.