This startup has its staff working 12-15 hour days to use AI to write the code for an AI system that gets people to pay up on their debts. Collecting debts involves dealing with people who are in difficult, exceptional situations: I have no faith that they are willing to properly engage with any of the detail. I have absolutely no doubt that this company will leave a long trail of suicides behind it.
Bootcampers have brought their gold digging ethos with them into our industry and will soon outnumber those of us who prioritize safety and quality.
"The safety and welfare of society and the common good, duty to our principals, and duty to each other, require that we adhere, and be seen to adhere, to the highest ethical standards of behavior."[0]
> "Solve deep product problems like how to collect more money with a voice AI agent."
Next "killer app"...an app you can send a phone call to the moment you realize it's a spammer/collector/etc. The app uses an LLM and voice synth to have the most boring and frustrating conversation ever with the caller. It should frequently ask them to repeat themselves, pretend to misunderstand common words, and for bonus points...speak in a broken accent. I'd pay $20 for that app right now.
I mean, honestly, everything about that ad comes across as basically a parody of a terrible tech company. 12-15 hour days! The actual product! I’m having some difficulty believing it’s real.
This article makes the fundamental flaw that the only type of programming is for Work For Other People.
For me at my day job, I find success with Cursor as “fancy autocomplete”. It’s aiding me when I am writing the code. The most code it’ll ever generate is to start on unit tests.
I’ve also used Cursor on the side for little personal hobby projects where I let to go wild in generating the overwhelming majority of code. I can’t say whether it’s faster or not, but it certainly helps reduce and overhead, initial blockers, ir lowering the barrier for myself to make something.
For those who are skeptical or haven’t tried it yet, ignore this article and just go give this new tool a decent try -carefully on your existing code base, and in no-stakes hobby projects - to form your own real opinion.
Yep, I also thought that the article is less about personal projects. The list of software development aspects that the author gives includes CI/CD, good documentation, and integration tests — I'd be surprised if a lot of hobby projects out there had these (apart from what's available out of the box for free, e.g. Vercel's automatic deployment from GitHub).
In the boring professional setting though, I can totally relate. The really hard questions I have to answer at work are usually not about code.
For a one-off script or a weekend project, on the other hand, even the current gen AI is a life-changing thing.
According to the top echelons of YC, vibe coders are "1000x engineers". Add in 15-hour days and you can probably expect that team of six to surpass AWS in features by next month.
Thought that listing was parody when I first saw it a week or so again. Since then the pay range has increased substantially -- originally it was, I believe, $80 - $120K for your "12 to 15 hour days", and onboarding that literally includes full time making collection calls, which...rofl.
EDIT - Both times I've joked about this ad, suddenly my account is rate limited and I can't post anything new without a "you're posting too fast" error. Maybe coincidence, but absolutely humorous.
Ignoring all of the absolutely insane things about that job description.
Do they have a list of clients I can access, so I can avoid using any of their products? The first time I get a phone call from an AI bot starting with "I see you opened our email about...", I'm throwing my phone out the window.
since this is an AI company it's not entirely impossible that they're putting out these ads to portray a certain image, rather than to actually hire people...
> At least 50% of the code you write right now should be done by AI; Vibe coding experience is non-negotiable.
> Putting in 12 to 15-hour days
These two don't correlate. AI is sold as saving time, so why not advertise it as a part-time job? If 50% of code is AI generated, then I expect it to be a 20 hour / week contract at 100% pay. Right?
> Your onboarding will be making collection calls.
...but there's AIs that are supposed to do that. Right?
This has got to be a troll advert or shell company. Their website doesn't list any employees or products. Half the people working for them on linkedin are deleted accounts, the rest are students or serial founders [0]
They need a 'vibe coder' for that because no software engineer worth their salt is going to onboard as muscle for debt collection agencies.
Not the first job post I've seen lately that has reverted to an aggressive tech-bro tone, either. I thought we moved on from the brogrammer shit of the last decade but it's back in full swing.
There are “AI artists” claiming “prompt engineering” is a skill that should give them copyright over the works the image generator produces, so this is not surprising in the slightest.
In the US, copyright is assigned to whoever takes a photograph. Pushing a button on your phone isn’t much work at all and yet the courts think that’s enough effort to allow someone to claim copyright over the work. So I don’t think the claim that the prompting, inpainting etc. required to generate an AI imagine isn’t “enough” work for someone to claim copyright doesn’t really hold water.
The argument for "vibe coding" (or generally heavy-use of AI coding) is always along the lines of
1. Vibe coding is much more efficient than regular coding
2. If you don't learn vibe coding you will fall behind
I see lots of arguing over point 1. but I think we can reason about 2. such that it makes the veracity of 1. irrelevant.
There is literally no skill you have to learn NOW (meaning today, this week, this year) that will ruin your career if you don't learn it. There are still very productive and well compensated people writing using editors and other tooling created in the 20th century.
Equally, your boss isn't going to come in to work tomorrow and say "you aren't already a vibe coder?! We expect you to be even though this is the first time I am mentioning it. You're FIRED!"
So if you want to learn "vibe coding", go ahead, but don't feel some great existential anxiety over it. People saying you will "fall behind" are just creating clickbait nonsense.
I kinda agree but the article does not do a great job at defending the position. Who cares about docs?
Vibe coding, or just letting AI take the wheel will work in some situations. It allows non coders to do things they couldn't before and that's great. Just like spreadsheets, no code tools, and integrations tools like Zapier, this will fill a bunch of gaps and push the threshold where you need to get software devs involved.
But as with all these solutions there is that threshold were the complexity, error margin gets, and scale go beyond workable and then you need to unfuck that situation and enforcing correctness. And I think this will result in plenty "oops my data is gone" types of problems.
If you know upfront that your project will get complex and/or needs to scale you might be best off skipping the vibe coding and just getting it right, but for prototypes, small internal tools, process "glue", why not.
It's not a replacement of software engineering as a whole (yet), it's just another tool in the toolbox and imo that's great. Can I use it, no.. I have tried and it just doesn't work at all for bigger more complex projects.
But there are actually uses for sloppy coding (whether "vibe," or some other [low|no]-code variant).
In particular, prototyping. This has always been a great application for that kind of thing. I'm old enough to remember when Flash and Director were used to prototype UI (Heh. I remember Microsoft demoing their "Longhorn" UI from Director, and trying to hide it).
Also, if your actual product is the company, then sloppy code is fine (I guess). You are basically just showing a prototype, anyway.
Even writing articles on this stuff feels like insecurity. Just let the AI fad come and go; the real engineers will take what works and everyone else will fail. The tide goes in - the tide goes out.
I laugh at the vibe coding nonsense now -- yeah, a million people making tiny variations of the same crappy web games is not revelatory -- but it truly does feel like there are extremes that are just as harmful.
On the one side are the super enthusiasts who grossly oversell to try to seem innovative and "with it", hoping they can claim some land in the great new AI development world.
On the other side are the head-in-sand types who keep railing about how useless AI is, it's a stochastic parrot, only super juniors find it useful and it holds no value for the Super Novel Work that they engage in, etc. You see this sort of commentary on here all the time.
Right now it's somewhere in the middle. I find the tools extremely helpful in my day to day, and they've completely changed how I work, and the tools are growing more valuable with each passing week.
I've not run into anyone who says AI is useless. I'm highly critical of AI as an "everything machine", won't let it code for me and believe AGI is a pipe dream... and I still have incorporated it into my workflow as a rubber duck, idea soundboard and documentation search engine.
On the other hand I've run into way too many people (dozens in person, hundreds online) who are overselling AI, most with a direct financial interest as their motivation.
It's not true that the sides are at all balanced unless you create an extreme "anti-AI" side that doesn't exist.
"Vibe coding" is fine for prototypes, weekend projects, hackathons etc, it's another variant of other quick coding / website making / visual programming IMO. But don't expect it to be production grade, and within the next year or so we'll probably see the fallout of people actually doing so.
I can easily "vibe code" production ready code with Windsurf. The code is SOLID, unit tested, documented, commented and resembles the code I'd write by hand. It would be impossible for you to distinguish my vibe code from the code I wrote by hand a year ago.
I'm sure there's a time and a place for "vibe coding", but a related point is with larger software projects, most work is maintenance (shameless plug: see my blog post on the subject[1]). It should be obvious that if there's a problem in a serious project, "just work around it or ask for random changes until it goes away" is probably not going to get you very far. I suspect perhaps "vibe coding" was just pointed out with the attitude of "hey this is kinda cool", but hype is blowing it out of proportion.
Great point, this maps to my experience. Almost every "fix this existing code" problem posed to LLMs is a failure. The approach of simply spitting out the next probable token doesn't work if any complexity or design aesthetics is involved. We wouldn't trust a human that approached problems by stream of consciousness either.
But vibe coding does work reasonably well for solo greenfield/demo projects. When there is no external complexity, no inferred requirements, no consequences to getting it wrong, nothing to integrate with, nobody to collaborate with ... just you and a blank editor - vibe coding works.
While it can be nice living in a little self-contained fantasy world, on existing projects you'll find absolutely none of the same conditions. And a whole slew of other engineering challenges that aren't solved by throwing generated code at it. As the value of code trends to zero, the value of those other engineering skills increases.
Readit News
> At least 50% of the code you write right now should be done by AI; Vibe coding experience is non-negotiable.
That is absolutely ridiculous.. As the kids say these days, I think that company is cooked.
0 - https://www.ycombinator.com/companies/domu-technology-inc/jo...
"The safety and welfare of society and the common good, duty to our principals, and duty to each other, require that we adhere, and be seen to adhere, to the highest ethical standards of behavior."[0]
[0] https://www.isc2.org/ethics
It's kind of like using autocomplete in Gmail: nobody says, “Well, Google wrote 20% of that message.”
We just quietly thank the autocomplete gods for saving us from typing “per my last email” one more time.
"Putting in 12 to 15-hour days, the engineering team has traveled to +10 cities in the past half-year for product launches."
"Solve deep product problems like how to collect more money with a voice AI agent."
Next "killer app"...an app you can send a phone call to the moment you realize it's a spammer/collector/etc. The app uses an LLM and voice synth to have the most boring and frustrating conversation ever with the caller. It should frequently ask them to repeat themselves, pretend to misunderstand common words, and for bonus points...speak in a broken accent. I'd pay $20 for that app right now.
Which is kind of funny, because if it's an automated call you don't need to even take it. Who picks up the phone from unknown numbers these days?
Shocking that no one wanted to work there.
Deleted Comment
For me at my day job, I find success with Cursor as “fancy autocomplete”. It’s aiding me when I am writing the code. The most code it’ll ever generate is to start on unit tests.
I’ve also used Cursor on the side for little personal hobby projects where I let to go wild in generating the overwhelming majority of code. I can’t say whether it’s faster or not, but it certainly helps reduce and overhead, initial blockers, ir lowering the barrier for myself to make something.
For those who are skeptical or haven’t tried it yet, ignore this article and just go give this new tool a decent try -carefully on your existing code base, and in no-stakes hobby projects - to form your own real opinion.
In the boring professional setting though, I can totally relate. The really hard questions I have to answer at work are usually not about code.
For a one-off script or a weekend project, on the other hand, even the current gen AI is a life-changing thing.
I thought vibe coding was just a meme. There are people who put vibe coding on _their resumes_??
> Solve deep product problems like how to collect more money with a voice AI agent.
> Ready to grind long hours, including weekends, to hit our ambitious goals.
Sounds like hell for an engineer.
EDIT - Both times I've joked about this ad, suddenly my account is rate limited and I can't post anything new without a "you're posting too fast" error. Maybe coincidence, but absolutely humorous.
Do they have a list of clients I can access, so I can avoid using any of their products? The first time I get a phone call from an AI bot starting with "I see you opened our email about...", I'm throwing my phone out the window.
-- edit "impossible" not "possible"
Deleted Comment
https://www.ycombinator.com/companies/domu-technology-inc/jo...
>Your onboarding will be making collection calls.
>At least 50% of the code you write right now should be done by AI
You're doing collection calls for a company that does "Automated human-like voice calls for the insurance industry".
> Putting in 12 to 15-hour days
These two don't correlate. AI is sold as saving time, so why not advertise it as a part-time job? If 50% of code is AI generated, then I expect it to be a 20 hour / week contract at 100% pay. Right?
> Your onboarding will be making collection calls.
...but there's AIs that are supposed to do that. Right?
This has got to be a troll advert or shell company. Their website doesn't list any employees or products. Half the people working for them on linkedin are deleted accounts, the rest are students or serial founders [0]
[0] https://www.linkedin.com/company/domu-ai/people/
Not the first job post I've seen lately that has reverted to an aggressive tech-bro tone, either. I thought we moved on from the brogrammer shit of the last decade but it's back in full swing.
- Social Media Engagement Strategist
- Brand Evangelist
- Vision Alignment Strategist
- Chief Happiness Officer
...are also real job titles nowadays.
I see lots of arguing over point 1. but I think we can reason about 2. such that it makes the veracity of 1. irrelevant.
There is literally no skill you have to learn NOW (meaning today, this week, this year) that will ruin your career if you don't learn it. There are still very productive and well compensated people writing using editors and other tooling created in the 20th century.
Equally, your boss isn't going to come in to work tomorrow and say "you aren't already a vibe coder?! We expect you to be even though this is the first time I am mentioning it. You're FIRED!"
So if you want to learn "vibe coding", go ahead, but don't feel some great existential anxiety over it. People saying you will "fall behind" are just creating clickbait nonsense.
Vibe coding, or just letting AI take the wheel will work in some situations. It allows non coders to do things they couldn't before and that's great. Just like spreadsheets, no code tools, and integrations tools like Zapier, this will fill a bunch of gaps and push the threshold where you need to get software devs involved.
But as with all these solutions there is that threshold were the complexity, error margin gets, and scale go beyond workable and then you need to unfuck that situation and enforcing correctness. And I think this will result in plenty "oops my data is gone" types of problems.
If you know upfront that your project will get complex and/or needs to scale you might be best off skipping the vibe coding and just getting it right, but for prototypes, small internal tools, process "glue", why not.
It's not a replacement of software engineering as a whole (yet), it's just another tool in the toolbox and imo that's great. Can I use it, no.. I have tried and it just doesn't work at all for bigger more complex projects.
I do (just sayin')...
But there are actually uses for sloppy coding (whether "vibe," or some other [low|no]-code variant).
In particular, prototyping. This has always been a great application for that kind of thing. I'm old enough to remember when Flash and Director were used to prototype UI (Heh. I remember Microsoft demoing their "Longhorn" UI from Director, and trying to hide it).
Also, if your actual product is the company, then sloppy code is fine (I guess). You are basically just showing a prototype, anyway.
On the one side are the super enthusiasts who grossly oversell to try to seem innovative and "with it", hoping they can claim some land in the great new AI development world.
On the other side are the head-in-sand types who keep railing about how useless AI is, it's a stochastic parrot, only super juniors find it useful and it holds no value for the Super Novel Work that they engage in, etc. You see this sort of commentary on here all the time.
Right now it's somewhere in the middle. I find the tools extremely helpful in my day to day, and they've completely changed how I work, and the tools are growing more valuable with each passing week.
On the other hand I've run into way too many people (dozens in person, hundreds online) who are overselling AI, most with a direct financial interest as their motivation.
It's not true that the sides are at all balanced unless you create an extreme "anti-AI" side that doesn't exist.
There are even very skilled and accomplished engineers that don’t even use language servers.
Not programming, but even some legendary Disney animators still draw out their key frames by hand… on paper… in pen.
Build with what you build best with.
Personally, they help with little refactors and occasionally a quick, difficult to google, question, usually about syntax.
We are weavers looking at a loom for the first time.
[1] https://www.construct.net/en/blogs/ashleys-blog-2/reality-lo...
But vibe coding does work reasonably well for solo greenfield/demo projects. When there is no external complexity, no inferred requirements, no consequences to getting it wrong, nothing to integrate with, nobody to collaborate with ... just you and a blank editor - vibe coding works.
While it can be nice living in a little self-contained fantasy world, on existing projects you'll find absolutely none of the same conditions. And a whole slew of other engineering challenges that aren't solved by throwing generated code at it. As the value of code trends to zero, the value of those other engineering skills increases.