AI coding is gambling

This is exactly the sort of mentality that makes me hate this technology

You finally feel good at programming despite admitting that you aren't actually doing it

Please explain why anyone should take this seriously?

I believe AI will do something similar for programming. The level of complexity in modern apps is high and requires the use of many technologies that most of us cannot remotely claim to be expert in. Getting an idea and getting a prototype will definitely be easier. Production Code is another beast. Dealing with legacy systems etc will still require experts at least for the near future IMHO.

Yes, it is insane. You couldn't torture this confession out of me. But that's the drug they're selling you, isn't it? You don't even write code, but you're getting a self-inflated sense of worth. It must be addicting! Of course, whether or not the programs you prompt are actually good surely has no relation to whether you feel they're good, since you're not the one writing them, and apparently were not capable of writing them before so are not qualified to review them very much.

> having tools that can finally match the speed my ideas come to me

Anyone can be an "ideas guy". We laughed at those people, because having ideas is not the hard part. The hard part was in all of the hundreds and thousands of little details that go into building the ideas into something actually worthwhile, and that hasn't changed. LLMs can build an idea into a prototype in a weekend. I am still waiting to see LLMs build an idea into something other people use at scale, once, ever, other than LLM wrappers. Either every person who is all-in on vibes only has ideas that consist of making .md files and publishing them as a "meta agent framework", or LLMs are not actually doing a great job of translating ideas into tangibly useful software.

You do not like and have never liked programming. You wanted to be a manager. They are completely different things.

On the surface this does not sound as satisfying, because it more resembles shopping than coding. But once Claude Code is finally tuned to do its job perfectly, you will essentially be using that infinite app store. You're actually using it right now, every time you use Claude Code — just an imperfect version of it.

If you enjoy using AI because it allows you to "will anything into existence", it's because the process is currently imperfect. Using Claude Code is closer to shopping than coding, but because the process is obfuscated, it feels like you're the one making the products in the shopping catalogue every time you place an order.

Hyperbole, yes, many things are in fact, not possible. But most people have the size of the two categories confused. The number of things that are categorically impossible is less than a rounding error compared to how many things are possible.

The joy and wonder of being an engineer is in taking problems deemed "impossible" and creating possibilities. It's in extracting a solution from infinite possibilities and redefining what possible even is.

It's just one step along the path of AI adoption to execute on an idea and see in near real-time the idea you had baked in your head come alive in front of you. Most of us get to this point and become the biggest evangelists of the tech. I see no reason you should feel guilty for the excitement you're feeling right now, and you should enjoy the journey. You're definitely paying for it in tokens, that's for sure.

However, there will come a point at which you will have successfully willed into existence a novel thing that you always wanted, and there it is, exactly as you dreamed, but by then, you'll be left with a weird empty feeling you won't really have the words for. Maybe it's a feeling of not earning the thing you built, or maybe it's just, your idea is finished and now you have to think of another idea. Certainly, this was your idea though, and it proves you were right, or at least on to something, and it is valuable, to you.

Yet, you didn't go on the journey to get there. You didn't bump up against limitations of the programming language or system and think about workarounds while you were showering or commuting to the office. You basically bought the finished product from the dynamic template marketplace of Anthropic (or whereever), and that's cool that it does what you need. It just isn't really programming, or being a software engineer in the traditional sense.

What used to be something you could potentially leave your day job for to go create a startup with a cofounder over, or maybe sell off to a buyer, or just open source and share with the world, isn't going to have the same meaning. It's a black box of code that you'll need a coding agent to continue working on, keeping that money flowing to Anthropic or whereever.

Anyway, I think the Slot Machine question is where a lot of early adopters are now at in this journey, and once more of us are there, then we can start asking the hard questions. Right now too many of us are where you're at, and it's impossible to know where things will end up in a year or so.

I kept at it on the side as a hobby. But stacks evolved and I was left behind. Now with AI it's back on.

You could just as easily make claims about carpentry or cooking because you discovered Ikea or microwave meals. They serve a purpose and technically satisfy the needs of anyone, yet they aren't a good enough solution for anything important. That's where we're at with this tech.

And the quality of code models puts out is, in general, well below the average output of a professional developer.

It is however much faster, which makes the gambling loop feel better. Buying and holding a stock for a few months doesn't feel the same as playing a slot machine.

>You are not an AI and do not know how an AI "thinks".

>Even if you come to be able to anticipate an AI's output, you will be undermined by the constant and uncontrollable update schedule imposed on you by AI platforms. Humans only make drastic changes like this under uncommon circumstances, like when they're going through large changes in their life, not as a matter of course.

>However, without this update schedule, problems that were once intractable will likely stay so forever. Humans, on the other hand, can grow without becoming completely unpredictable.

It's a Catch-22. AI is way closer to gambling.

As @m00x points out "coding is gambling on slot machines, managing developers is betting on race horses."

Except, one can explain themselves (humans) and their actions can be held to account in the case of any legal issue whereas an AI cannot; making such an entity completely unsuitable for high risk situations.

This typical AI booster comparison has got to stop.

It wasn't a real game of hangman, it was flat out manipulation, engagement farming. Do you think it's possible that AI does that in any other situations?

These guys had to manage very complex calculation engine based on we’ll just let it changes every year had to be correct had to be delivered by a certain date every year.

They had an army (100-200 people depending on various factors) of marginally skilled coding drones that were able to turn out the Java, COBOL or whatever it was predictably on that schedule without necessarily understanding any of the big picture or have any having any hope of so. Basically a software factory. There was about a dozen people who actually understood everything.

Being a project manager is more or less something humans have been doing since the dawn of time.

Generative AI takes money as input and gives some output. If you don’t like the output, more money goes in. It’s far more akin to gambling than organizing human labor.

But where there are no tests, and you're the one defining whats correct, you're definitely encroaching on the slot machine hoping it'll spit out something faster than you could do it yourself.

Then there's some vague unease in whether spending the time to prompt will actuall result in a properly integrated software in a large existing code base with idiosyncratic code use.

Overall, I don't see the ROI business gets from forcing people into these tools; however, as an individual, it's definitely worth understanding what they can do. Mostly, I see the efficient copy/past/find/replace of existing code to be very good.

Bold assumption that people are looking at the diffs at all. They leave that for their coworkers agents.

It is exactly like the proverbial monkey or rat pressing a bar for a food pellet to come out.

If the pellet unerringly drops, and is always tasty and nutritious, the rat stops when it's no longer hungry.

Otherwise, an inordinate amount of time is spent pressing the bar.

I'm not an expert in the psych/neuro literature on addiction, but I suspect latency isn't that critical. But is that just because it's things like fruit-machines that have been studied? Gambling (poker, racehorses) are quite long-latency. OTOH, scrolling is closer to 400ms, and that's certainly the modern addition...

[Imports the completely fabricated library docker_quantum_telepathy.js and calls the resolve_all_bugs_and_make_coffee() method, magically compiling the code on an unplugged Raspberry Pi]

AI: "Done! The production deployment was successful, zero errors in the logs, and the app works flawlessly on the first try!"

That's where the gambling metaphor really resonates. It's not whether or not the output is correct, I've been building software for many years and I know how direct LLMs pretty well at this point. But I'm also an alcoholic in recovery and I know that my brain is wired differently than most. And using LLMs has tested my ability to self-regulate in ways that I haven't dealt with since I deleted social media years ago.

Which with the advent of LLMs just lowered our standards so we can claim success.

But even accounting for all these "hard" constraints and metrics, there are clearly reasons to prefer some possible programs over others even when they all satisfy the same constraints and perform equally on all relevant metrics.

We do treat programs as efficient causes[1] of side effects in computing systems: a file is written, a block of memory is updated, etc. and the program is the cause of this.

But we also treat them as statements of a theory of the problem being solved[2]. And this latter treatment is often more important socially and economically. It is irrational to be indifferent to the theory of the problem the program expresses.

[1]: https://en.wikipedia.org/wiki/Four_causes#Efficient

[2]: https://pages.cs.wisc.edu/~remzi/Naur.pdf

Why do you often need to re-prompt things like "can you simplify this and make it more human readable without sacrificing performance?". No amount of specification addresses this on the first shot unless you already know the exact implementation details in which case you might as well write it yourself directly.

I often have to put in a prompt like this 5-10 times before the code resembles something I'd even consider using as a 1st draft base to refactor into something I would consider worthy of being git commit.

I sometimes use AI for tiny standalone functions or scripts so we're not talking about a lot of deeply nested complexity here.

The endless next steps of "and add this feature" or "this part needs to work differently" or "this seems like a bug?" or "we must speed up this part!" is where 98% of the effort always was.

Is it different with AI coding?

This seems like 1980's corporate waterfall thinking, doesn't jibe with the messy reality I've seen with customers, unclear ideas, changing market and technical environments, the need for iteration and experimentation, mid-course correction, etc.

1/ Dependency -- Once I got used to agentic coding, I almost always reached out to it even for small changes (e.g. update a yaml config)

2/ Addiction -- In the initial euphoria phase, many people experience not wanting to "waste" any time agent idle and they'd try to assign AI agents task before they go to sleep

3/ You trust your judgement less and less as agent takes over your code

4/ "Slot machine" behavior -- running multiple AI agents parallel on same task in hope of getting some valuable insight from either

5/ Psychosis -- We have all met crypto traders who'd tell you how your 9-5 is stupid and you could be making so much trading NFTs. Social media if full of similar anecodotes these days in regards to vibecoding with people boasting their Claude spend, LOC and what not

The important part of the not-really-a-metaphor is the relationship between user and machine, and how it affects the user's mind.

What the machine outputs on "wins" doesn't matter as much, addictive gambling can still happen even when the payouts are dumb.

You can get more consistent results from a slot machine with a bunch of magnets and some swift kicks. It's still gambling.

We love a good holy war for sure.

The nuance is lost, and the conversations we should be having never happen (requirements, hiring/skills, developer experience).

That isn't true, which is the exact reason why people have a binary mindset. More than once on Hacker News I've had people accuse me of being an AI booster just because I said I had success with agents and they did not.

Good programmers might have made things that “performed well”, and had “few bugs”, without this step, but it was not robust to changes over time. If we end up in a place where every project has solid automated verification, perhaps things get better overall.

Fast & Cheap (but not Good?) - I wouldn't really say that AI coding is "cheap"

Cheap & Good (but not Fast) - Again, not really "cheap"

Fast & Good (but not Cheap) - This seems like maybe where we're at? Is this a bad place?

I’ve never worked anywhere where the interns had net productivity on average.

But looks like the intern mafia is bombarding you with downvotes.

2. Who here thinks that having interns write all/almost all of your code and moving all your mid level and senior developers to exclusively reviewing their work and managing them is a good idea?

The reason i think this metaphor keeps popping up, is because of how easy it is to just hit a wall and constantly prompt "its not working please fix it" and sometimes that will actually result in a positive outcome. So you can choose to gamble very easily, and receive the gambling feedback very quickly unlike with an intern where the feedback loop is considerably delayed, and the delayed interns output might simply be them screaming that they don't understand.

You should value assigning tasks to human interns more than AI because they are human

The first is equating human and LLM intelligence. Note that I am not saying that humans are smarter than LLMs. But I do believe that LLMs represent an alien intelligence with a linguistic layer that obscures the differences. The thought processes are very different. At top AI firms, they have the equivalent of Asimov's Susan Calvin trying to understand how these programs think, because it does not resemble human cognition despite the similar outputs.

The second and more important is the feedback loop. What makes gambling gambling is you can smash that lever over and over again and immediately learn if you lost or got a jackpot. The slowness and imprecision of human communication creates a totally different dynamic.

To reiterate, I am not saying interns are superior to LLMs. I'm just saying they are fundamentally different.

And, if we're being honest, the way people talk about interns is weirdly dehumanizing, and the fact that they are always trotted out in these AI debates is depressing.

The idea that I would only be able to work with an internet connection and a subscription to a big LLM provider is not very appealing to me. And I'm not even talking about the ethical concerns...

Humans invented gambling as a rigged game that mimics what's in nature, perversed for profit.

If we're only talking about money spent on prompting AI, maybe. The damage to online trust is massive imo. So is the damage done by looting the commons to build them.

Typical privatize the profits socialize the costs bullshit

Important to point out that that every high culture produced restrictions on exactly those behaviors, gambling was a universal vice when that concept still mattered.

America in particular had a work culture that favored well, work and technical excellence. Now work is for suckers, thinking is for suckers, precision not worth it when you can have some machine do it half-right.

"Yes I could go for the reliable option. But taking a punt is worth a shot if the cost is low.", might as well be the national slogan from vibe-coding to the department of defense. Even the venture capital industry that excels at slot machine sectors was itself already a slot machine.

Right, which is why LLMs aren't useful if you actually know what you're doing. It's a drain on your time to have to carefully check everything a junior writes, but you do it because he will learn and eventually return on that investment. With an LLM, there is no such long term payoff.

You can't keep paying to play the "refinancing game" until you get a good rate (at least not like pulling the lever again and again, you have to wait a long time, you won't call the same bank again and again, and suddenly they have an amazing rate), it's a different experience and the psychology is different.

In a healthy environment. We are harmed more by being totally risk adverse. Than by accepting risk as part of life and work.

I am consistently using 100% of my weekly $200 max plan. I know how this thing works, I know how to get value out of it, and I wish what you said were true.

If you do all of these things? You are in a better spot. You are in a far better spot than if you hadn't! Setting up hooks to ensure notes get written? Massive win! Red-green TDD? Yes, please! But in terms of just ... well, being able to rely on the damn thing?

https://github.com/ctoth/claude-failures

> you have to spend a lot of tokens to keep the error/hallucination rate low

Ironically, I find your comment more effective at convincing me AI coding is gambling than the original article. You're talking about it the exact same way that gamblers do about their games.

Which I assert is semantically equivalent to saying: Human drivers (even when operating at the diminished capacity of not even being present in the car) are less likely to make errors driving a car than AIs.

That is extremely stupid. What does that ban get you? I reqct to this because a friend mentioned exactly this. And I was dumbfounded.

The results do speak for themselves, but it doesn't work.