This is exactly what is happening.
I think HE doesn't understand what vibe coding means.
Vibe coding is NOT "AI-assisted coding", it is letting the AI do everything with 0 checks beyond manually validating surface-level functionality and passing back the obvious issues to the AI.
That was the original intent, but now vibe coding is exactly synonymous with AI-assisted coding with the public at large. That's probably irreversible, and it's what Ng is highlighting.
I think Rick Rubin has it right that this is a kind of "punk" coding and I think vibe coding is actually a good term in that regard.
This rings especially true for me as a non-software engineer that took decades to appreciate punk music. I grew up on "virtuoso" shred guitar music and jazz. "A real musician spends hours a day in the woodshed learning their craft. Not just anyone can pick up a guitar and start making music."
I think the analogy also works because the virtuoso had a huge leg up on the untrained musician if they also decided to make simpler music.
On a longer time frame with the analogy, the next fed chairman obviously will not have a degree in clarinet from Juilliard like 99 year old Alan Greenspan.
So basically everyone?
The Deepwater Horizon movie clips show workers arguing over all the things that have failed diagnostics testing, and how important that is or isn't. They're interacting with machinery 5.6km down.
Developed-world underground miners are already using heavy machinery that has to conform a cut to a specific 3d model which is adapted to the seam and the structural engineering of the tunnel. Open pit mines have similar structural concerns, but also environmental remediation issues. The trucks making that long journey down into the pit are all going robotic now.
An infantry unit in 2025 that hasn't already won or lost on overwhelming air superiority, is rapidly adapting drones to help it fight.
If you aren't making software, you're at least using software that somebody else created, you need to understand the problems and strengths of a software aided approach (as opposed to an apprentice aided approach, or a mechanically aided approach) and you're in a (hopefully) collaborative relationship to make that software better.
(though "obviously" I only meant people interacting with computers - among which every single one could benefit from adding some level of automation)
Not everyone should hold a PhD in compiler design, but simply being able to write some BASIC commands might be hugely beneficial to society.
Reading and Writing is not something you learn just because it's beneficial to society. It's beneficial to one-self's life as it's the bases for communication. It can be placed on safety needs or belonging needs on the Maslows' Hierarchy of needs.
Basic arithmetic is requirement for continuity of one's life as you can't plan resource management (money, food) to function without it. It could be put on the safety needs.
I am having trouble positioning programming in any level below self actualization.
Do we gain more by having more people understand software? Do some risks get mitigated? I'd argue yes on both counts. And that's why I encourage people to get their kids to poke around with Scratch and maybe some Python. It's why smart people doing liberal arts degrees at Harvard still want to go and do CS50.
However "learning to program" and "understanding software" is not the same thing. Learning to program helps one to better understand software, but it's not a necessity or requirement of it.
"Everyone needs to learn" vs "People who needs to interact with software needs to better understand it" are very different points on the "understanding software" axis.
Most people don't want to learn to code because it simply doesn't interest them as a subject, a perfectly valid way of living your life.
Most professionals work with data but have low technical literacy. Imagine if most professionals could just query databases themselves. In a generation or two it could be made possible.
So, no, not everyone needs to be a car mechanic. But everyone who wants to drive a car needs to learn how to drive it safely, and would benefit from checking the oil, tire pressures, and knowing something about how their car worked to keep it in good, safe working order. If the only people who know all those things are professional mechanics, we're in a worse place.
To be fair, Karpathy coined the term specifically for a playful, just-for-fun, quarter-serious approach to software development where you just tell the machine what you want and then be open for surprises. Assisted coding is not vibe coding in that sense.
I have a Vibecoding sticker in the SUPREME style on my work laptop, which I think accurately summarises my opinion towards it.
I'm not even sure in what use cases, even toy ones, vibe coding works. It's been a few months since I tried using Claude to even do simple things like generate a React-based prototype for part of a web page (the sort of use case I would expect it to do well on), and even there it wasn't a "haha isn't this cool" hands-off "vibe coding" experience - I had to intervene and to do my own google research to find out why it was failing and then tell it explicity what to do.
I also have to wonder how many of these AI researchers fully realize the massive gap in complexity between what they are writing (typical ML model or prototype) and what real software development looks like. Let's see Karpathy "vibe code" a 100L-1M LOC system that needs to interact with a dozen undocumented legacy systems via proprietary interfaces, then come back and tell us how that went.
My suggestion:
Pro-coding
The "pro" can be thought of as either "process" or "professional" or similar.
The main point is that unlike vibe coding it's not just flying by the seat of your pants, you're actually doing intellectual work as described by Ng.
The thing we're talking about is most of the time pretty much the opposite of "professional", which your term alludes to.
UIs, e.g., are not my field of expertise, but I can use AIs to help me produce UI code that is useful to me. It's still a hassle as described by Ng and I'm certainly not flying by the seat of my pants.
The term for someone who dabbles in a field without being an expert in it is not "professional" but "dilettante". I therefore propose the term
Dilletante-coding or short Dilletanting
It is many, many times more apt than "pro-coding".
Although it brings to my mind rich people who don’t need to work for a living and just play at dabbling in things until the shine wears off?
I guess AI-assisted coding (sounds like assisted living for the elderly) is replacing Stack Overflow "cut n paste" coding, where there was similar lack of attention to what people were copying. In the case of AI what is being generated/pasted comes from the training data, so perhaps "statistical paste coding" ? "hive mind coding" ?
"Hive mind coding" almost sounds like a good thing (leaning on the collective experience of the global software community - what could be better ?!), until you realize (e.g. as Fred Brooks describes in "Mythical Man Month") that any software project large enough to need multiple developers (or AI equivalent) needs a clear vision and an architect, lead, etc. 1M uncoordinated coding monkeys isn't going to cut it, even if they are all guided by a "hive mind" cookbook of what to write.
Maybe this is the real disconnect, or a large part of it, between benchmarks indicating super-human leetcode and competitive coding AI capability, and Google reporting 30% of software AI developed (= test cases?), and not hearing of much utility outside of these one-man projects or narrow "write me a test function" type things.
Next people will be claiming that "influencer" or "twitter shit poster" is a real and exhausting job.
Second, our job in technology is to make ALL jobs easier, that’s what technology is for, not for bullshit manipulative, addictive and extractive consumer crap. The reason any of it even exists is to improve the productivity of humans.
There will always be demanding jobs, they may be demanding physically, or mentally or both, your god damn job is to figure out how to make every one of those jobs easier and LESS physically and mentally challenging.
Pointing out the obvious fact that using different metrics other jobs are harder is neither helpful, valuable nor unique.
I will however agree with you last statement, technologies abuse of people in the consumer app space is anti-social and destructive to the world, those are “jobs” we created with technology. In a sense you might say we are responsible for creating the worst jobs in the world, because as easy and valueless as being an influencer is, it destroys people mentally and turns people into shells of human beings.
So instead of trying to imply that all your fellow engineers are a bunch of whiny soft and weak complainers, you should be both simultaneously grateful that there are jobs that are physically easy and obligated to help those whose jobs still aren’t easy make as much of their jobs as easy as possible.
We live in a society we are ALL dependent on each other, specialization is what allows us to have large complex societies, without it we would all be trying to find food and build shelter. We can ONLY have our jobs because others do theirs, never forget that, that fact creates an OBLIGATION not a comparison.
Manager who sits upvoting and saying "great post" on LinkedIn all day, seems to be most middle management right now.
There's a different question to be made as to whether or not influencer is a good direction for society to strive for as a career, but that's a whole other matter.
It there _any_ empirical evidence for this, at all? Like, beyond the _perception_ of the users.
Someone managed to write a standards-compliant HTTP 2.0 server in just two weeks, 40k+ lines of code (including tests) with almost all of the coding done by Gemini Pro: https://outervationai.substack.com/p/building-a-100-llm-writ... . No unassisted programmer could do that so fast (except maybe Fabrice Bellard).
It is encouraging though that Ng admits that dealing with AI "coding" is like wading through a swamp and leaves you exhausted.
Regarding "vibe coding": I'm waiting for the first National Security Letter that demands that a secret backdoor is inserted in the generated code. No one will notice, because we "can forget about the code".
And you continue doing so with every other point you are making.
Just because some random people think they can vibecode real products, doesn't mean that this didn't happen before just slower.
A company i worked for, had their MySql server unprotected on the internet for no reason at all. They still used MD5 too.
Another company saved credit card information in their DB with a simple generic key in code accessable by everyone.
So your argument is that since we had incompetence before, let's have more of it?
It seems to me that everybody has a slightly different expectation of what vibe coding is, and what to expect from it. Same with AI generally. On the one end you get people insisting its going to change the World, and on the other you get "vibe debt" references and, well, this: https://www.youtube.com/shorts/qWJUWinWWnQ
My experience is that all AI today is good at helping experts be more productive by giving them first draft artefacts that might need editing, and sometimes that editing is extensive, sometimes maybe nothing.
That's it.
They are not "doing the job", they're assisting the job. People who expect AI to "do the job", are setting themselves up for failure, whether its filing paperwork in courts [0], writing book reviews [1], or writing software [2].
Sure you can use midjourney to produce some artwork for you, but if an artist used it, they'd be able to make it even better and fix it up in all sorts of ways, so maybe you will still need to pay the artist, they will just take less time and maybe lower their unit price of work.
If you are an expert at software engineering, and you are using AI, your job is changing, but it's not going away. I think every mature conversation about current agentic coding practices acknowledge this. And I think people who are getting good at this will be more productive, and they will get more done, but they're not going to be getting their job done in 4 hours and taking the rest of the week off, because they're just going to try and add more value to get paid more, or even just to keep their jobs. That's the treadmill of free market capitalism.
The "vibe coding" stance tries to counter this, but doesn't acknowledge the limitations of the technology. I think it can be taken in one of two ways: ignore the code and be surprised at what comes out and have fun with the results, which I think is fine for a hobby, you do you; or, ignore the code because you don't need to understand it, and let's get rid of all the expensive developers, and hey, what could possibly go wrong in this production stack, which I think is going to end very badly for all involved.
While today the expectation of vibe coding with current technology is as ridiculous as me saying I'm going to replace my lawyer with a bot and YOLO it [3], I am not sure that will hold forever, or for the rest of my career, or for even the next year. The only way to find out is for experts to try it, and qualitatively and quantitatively assess the results. I'd love to see a vibe coding index which periodically tries to get apps of increasing complexity built and then assess the results, and we track this over time. At some point maybe we'll get through a tipping point, and we can have conversations about what that means.
I think we might need a better term for it though, that isn't AI-assisted development (as the intention is that its not assisting you, it's doing it). Natural language programming? Entirely Generative programming?
[0] https://www.independent.co.uk/news/lawyers-london-england-hi...
[1] https://chicago.suntimes.com/opinion/2025/05/29/lessons-apol...
[2] https://cyberplace.social/@GossiTheDog/114546174330023840
[3] https://petapixel.com/2025/04/14/man-tries-to-use-ai-generat...
Our job remains the same: take ambiguous customer statements, transform them into well-defined instructions that lead the computer to do what the customer was trying to ask for
Not sure what would be faster. Manually reviewing and fixing the AI migrated docs or just give up and use sed or something to do it from scratch.
It's much better to ask the specific queries that you yourself would do with sed and treat the models as smarter sed (or ask them to generate the sed statements themselves)
cumo•18h ago