it absolutely doesn’t. most software is shitty and buggy. I know, have seen 30 years worth of it. just now we can write the shitty and buggy a lot faster.
ask anyone that has been in the industry for awhile and they will tell you the same thing. a lot of crowd on HN write as if human-written code is any good and while there are always exception, on average, most software is shitty. if I had a dollar for every time I met someone writing software for “X” saying “dude, if you knew what I know you’d never use/do/… X” I’d be a very rich man
“Biggest innovation in tech” at times feels like 2000 era Nokia smartphones. You got phone with circle keyboard! You got foldable phone that turns into a neat camcorder! You got mp3 player phone!
All with marginal value add and having more to do with fashion than with actual innovation.
Real innovation is synonymous with problem solving. The only problem tech industry is solving nowadays is keeping its bloated valuation afloat.
In my mind, LLMs have strong parallels with self-driving vehicles. Both are impressive tech innovations with amazing future potential, and both have no shortage of evangelists that have claimed that the future is already here. Both have a swath of average joes trusting them with their lives and livelihoods, while those with enough technical knowledge realize that both can fail in unexpected and sometimes catastrophic fashion and won't use either without significant babysitting.
The primary difference I've noticed, however, is that the AI evangelists also have a flair for painting a picture where everyone's jobs are eliminated, while many CEOs are already using it as an excuse for layoffs.
I can see no reason why there'd be any negativity about AI.
I guess that depends on how old you are. The internet itself is a strong contender for being a bigger innovation. And the web. The personal computer. Smart phones, and even cell phones in general. Oh, I almost forgot GPS. And that is all just in the arena of personal computing/devices. Get outside of tech and bring in science, and we're in a completely different world.
I mean, yes, LLMs are an innovation. But outside of the tech industry, they are not having anywhere near the impact on people's lives as many other inventions.
Yet. The Internet was not impactful for the average person for many years. It took a long time for even half the population to get Internet access, and it took a long time for the Web to be developed
Uncaught TypeError: can't access property "parentElement", t is undefined
i https://io.google/2026/explore/workshop-2:29
<anonymous> https://io.google/2026/explore/workshop-2:29
<anonymous> https://io.google/2026/explore/workshop-2:29
Evidently the title is correct: we're beyond the tipping point of software quality.
I'm sick of the people, these so called builders, how they are parasitizing on the time of the actual engineers to get their PRs into the codebase.
People who have never coded, for who we now need to spend an insane amount of time validating their design, asking questions they don't even have the answers to, figuring out if the code written was needed for the prompt or whether it was just thrown in there because ... the AI had it as part of the tokens.
Remember, their design isn't built from the ground up, it's focussed on the outcomes, like the movie Bedazzled.
And then you spend time validating their code, you give them review feedback, instead of internalizing it they just shove it into the AI and let it fix it, meaning they don't learn.
Welcome to the life of enterprise developers forced to work with dirt-cheap contractors. Thanks to LLMs, this is now how all software development works.
I can't help but feel that this (not this specific article but the whole trend) is a little bit misguided. Nothing has fundamentally changed in engineering? We aren't committing raw markdown files in Github and having Jenkins re-inference entire systems at each push. The new AI tools still deal with the underlying mess and at some points the abstraction will leak.
It feels like Kubernetes? Yes, it allowed more complex setups but a decade later the leaky abstractions have become apparent and there are constant tradeoffs to be made with the problems it brings vs the problem it solves. But the awkward point is why was Kubernetes not a trillion dollar unicorn if it unlocked so much productivity gains?
"Learn to use systems thinking to understand how developer ecosystems guide the evolution of your software systems. Improve your intuition for the systemic impacts of AI-driven software development and understand how you can better prepare for the exciting changes coming to our industry."
This reads like a bunch of marketing speak that says nothing.
I watched it just to be able to comment here. The whole presentation feels AI generated. The diagrams look 100% AI generated. The text is pretty generic, I really don’t know what insights you’re expected to get from this.
I generally dislike HN cynical tone but I don’t see how not becoming a cynic myself when that’s the quality of presentation from a Principal Engineer on Google IO stage. Even at 2x speed that felt not worth the time
To continue with the janitor analogy, AI slop is like some drunk person intentionally smearing shit all over the bathroom walls and ceiling because it felt good at the time. They didn't have to do that, but they did and now someone else has to clean it up.
It is indeed at a tipping point; we are going from engineered deliverables to throwing stuff at the wall and keeping what sticks.
I wonder if the public appetite for the inevitable quality crash is there, though.
We used to ridicule "works on my machine" thinking but now we're pivoting the whole field into essentially that, shipping what "sticks" like you said.
the public is so used to good quality software that they are in for rude awakening :)
The thing is, people already have certain expectations from their software, because it is all around them.
They'd revolt en-masse if their TV had the downtime of github, of if their computer had the number of successful supply chain attacks that npm enjoys.
The quality may not have been good, but the market had stabilised on what the public would accept. AI changes this substantially.
it absolutely doesn’t. most software is shitty and buggy. I know, have seen 30 years worth of it. just now we can write the shitty and buggy a lot faster.
ask anyone that has been in the industry for awhile and they will tell you the same thing. a lot of crowd on HN write as if human-written code is any good and while there are always exception, on average, most software is shitty. if I had a dollar for every time I met someone writing software for “X” saying “dude, if you knew what I know you’d never use/do/… X” I’d be a very rich man
Most software may be shitty, but users absolutely are not prepared for the level of shittiness we see in vibed projects.
That's what I'm saying: users are used to a certain level of reliability. They might not be so accepting of a decline.
example? or this is theoretical?
And yet, quality can always get worse.
it could and likely will but it’ll be in the news for a day tops and the world will move on to the “next thing.”
only on hacker news, do we see the biggest innovation in tech in our lifetimes, and you get this weird cynical negativity
“Biggest innovation in tech” at times feels like 2000 era Nokia smartphones. You got phone with circle keyboard! You got foldable phone that turns into a neat camcorder! You got mp3 player phone!
All with marginal value add and having more to do with fashion than with actual innovation.
Real innovation is synonymous with problem solving. The only problem tech industry is solving nowadays is keeping its bloated valuation afloat.
In my mind, LLMs have strong parallels with self-driving vehicles. Both are impressive tech innovations with amazing future potential, and both have no shortage of evangelists that have claimed that the future is already here. Both have a swath of average joes trusting them with their lives and livelihoods, while those with enough technical knowledge realize that both can fail in unexpected and sometimes catastrophic fashion and won't use either without significant babysitting.
The primary difference I've noticed, however, is that the AI evangelists also have a flair for painting a picture where everyone's jobs are eliminated, while many CEOs are already using it as an excuse for layoffs.
I can see no reason why there'd be any negativity about AI.
I guess that depends on how old you are. The internet itself is a strong contender for being a bigger innovation. And the web. The personal computer. Smart phones, and even cell phones in general. Oh, I almost forgot GPS. And that is all just in the arena of personal computing/devices. Get outside of tech and bring in science, and we're in a completely different world.
I mean, yes, LLMs are an innovation. But outside of the tech industry, they are not having anywhere near the impact on people's lives as many other inventions.
Yet. The Internet was not impactful for the average person for many years. It took a long time for even half the population to get Internet access, and it took a long time for the Web to be developed
Is it really only on HN? I ask because I'm seeing it more or less everywhere I look, in and outside of tech.
Only on Hacker News, and everywhere else.
The page is blank.
Evidently the title is correct: we're beyond the tipping point of software quality.You might have an adblocker or something blocking some JS. Its loading fine on my iPad
I'm sick of the people, these so called builders, how they are parasitizing on the time of the actual engineers to get their PRs into the codebase.
People who have never coded, for who we now need to spend an insane amount of time validating their design, asking questions they don't even have the answers to, figuring out if the code written was needed for the prompt or whether it was just thrown in there because ... the AI had it as part of the tokens.
Remember, their design isn't built from the ground up, it's focussed on the outcomes, like the movie Bedazzled.
And then you spend time validating their code, you give them review feedback, instead of internalizing it they just shove it into the AI and let it fix it, meaning they don't learn.
Next PR will have the same issues.
Welcome to the life of enterprise developers forced to work with dirt-cheap contractors. Thanks to LLMs, this is now how all software development works.
Sounds exactly like my experience.
I can't help but feel that this (not this specific article but the whole trend) is a little bit misguided. Nothing has fundamentally changed in engineering? We aren't committing raw markdown files in Github and having Jenkins re-inference entire systems at each push. The new AI tools still deal with the underlying mess and at some points the abstraction will leak.
It feels like Kubernetes? Yes, it allowed more complex setups but a decade later the leaky abstractions have become apparent and there are constant tradeoffs to be made with the problems it brings vs the problem it solves. But the awkward point is why was Kubernetes not a trillion dollar unicorn if it unlocked so much productivity gains?
"Learn to use systems thinking to understand how developer ecosystems guide the evolution of your software systems. Improve your intuition for the systemic impacts of AI-driven software development and understand how you can better prepare for the exciting changes coming to our industry."
This reads like a bunch of marketing speak that says nothing.
The main content is the video. Not sure if it's any good
I watched it just to be able to comment here. The whole presentation feels AI generated. The diagrams look 100% AI generated. The text is pretty generic, I really don’t know what insights you’re expected to get from this.
I generally dislike HN cynical tone but I don’t see how not becoming a cynic myself when that’s the quality of presentation from a Principal Engineer on Google IO stage. Even at 2x speed that felt not worth the time
Then it’s a perfect embodiment of the AI era. You’ll read the slop and you’ll like it!
Thanks for taking one for the team. I flagged this post since AI slop is not allowed.
Someone has to clean up the slop and explain why their product got more outages than before. Reputation and responsibility cannot be vibe-coded.
Case in point: GitHub.
We're janitors at best and I don't mean to disrespect janitorial who do real and very important work. Fixing AI slop doesn't feel real.
Not even janitors. Messy toilet clogs may demand immediate attention. Software sanitation is relatively low in priority.
To continue with the janitor analogy, AI slop is like some drunk person intentionally smearing shit all over the bathroom walls and ceiling because it felt good at the time. They didn't have to do that, but they did and now someone else has to clean it up.