I was recently reading a Wall Street Journal article, and one sentence in it reminded me of Jack Welch, the CEO of GE from the 1980s to 2001, roughly 20 years at the top. I want to talk about that, because I think there is a real connection between what happened at GE and what is happening right now with AI coding agents and vibe coding.
The article was titled The AI Superstars Who Say a ‘Vibe Slop’ Crisis Is Coming, and it argues that vibe coding is generating code we cannot really trust. As of now, I agree.
The Problem With Blindly Trusting AI Code
You cannot just vibe code an app, push it to production, and assume it will have zero bugs, zero security problems, and follow 100 percent of your instructions. I think that is one of the biggest mistakes you can make. If you are a professional software developer, you are responsible for that code, which means you have to check what is going on and review the output.
I understand the pure vibe coder with zero software development experience being excited about all of this, and that excitement is fair. But that is out of scope here. I am assuming you are a professional developer building a product on behalf of a company, and I want to remind you: that code is your responsibility.
In the same article, a software developer says:
You have infrastructure that’s falling apart, and you have software that’s now very, very buggy compared to before. We can play this game for a couple more months, or maybe even years, but eventually it will catch up to us.
This is the part that matters. It will eventually catch up. The technical debt you are leaving behind will catch up eventually.
The GE Lesson: Trading the Future for Short-Term Results
So why did I bring up Jack Welch? He made GE one of the biggest companies in the world, but he did it by trading away the future of the company.
GE was basically a manufacturing company. Welch found an opportunity in finance and, in effect, decided he could hit his quarterly earnings by playing the finance game: buy companies, use a few accounting tricks, and make the numbers easily. That is what he did, for 20 years. But along the way the company lost its roots. It was no longer a manufacturing company. It was a company playing with the numbers.
Eventually, after the 2001 crisis (I mean 9/11) and then the 2008 crisis, it became obvious that GE could not sustain itself, and the company dissolved. GE had been alive for over a hundred years. Under Welch’s management it stayed alive for two more decades, but it took 20 or 25 years for the damage to fully show.
It may take a couple of decades. I understand that. But at least now we know it was not sustainable. Basically, GE was preparing its own death.
What This Means for Vibe Coding
I think the same thing is true for vibe coding. It will eventually catch up. Maybe not next month, maybe not even in the next two years. Maybe it will take 15 years before we understand the mistake of vibe coding without checking what is going on. I do not know the timeline. But I do know that if current AI systems stay the same, this is not sustainable.
Eventually we may end up in an era where companies are attacked by hackers easily, because they keep shipping security holes just as easily. Look, I am not saying that will definitely happen, and that is important. But I am saying that everybody should understand we are trading the future.
The only thing that can make this process sustainable is if AI tools get very, very, very much better. Right now we compare AI coding agents to junior software developers. Maybe when we can compare them to senior developers, principal engineers, or staff engineers, we can start talking about high-quality code that we can blindly trust. As of now, we cannot. And if you are a software developer blindly trusting the output of AI agents, you are making a mistake.
Now Let’s Talk About CEOs
There is also a pattern here, and I do not think it is only me who sees it. There are a lot of layoffs, and companies are essentially saying, “We have AI, so let’s fire people. Let’s give 15 AI agents to one software developer.”
I do not think that is possible, at least not if you want to produce high-quality software. The output will be huge, sure, and in the short term companies may well increase their profits. I believe that part. But in the long term it will not be sustainable.
It may take decades. It may take 20 years, it may take 25 years. I do not really know the timeline. But if the trend continues like this, I do not think it holds.
So we will see. I believe this is a fair warning: it may not be a good idea to trade the future for short-term profit. GE failed after 20 or 25 years. Maybe it will take another 25 years for us to realize we cannot sustain this process either. I do not know when that realization will come, but I know that blindly trusting AI-generated code is not sustainable.
Watch the Video
I also shared this perspective in video format. You can watch it here: