Did you read the article? The author means software in general, not just user-facing apps.

Yes, I did, just to make sure it was as silly as I thought at first glance.

You do realize that "The author means software in general" is already a concession that they don't actually address the question in the title, right?

The way I see it is - AI still makes mistakes, and I have to know how things work at some point anyway. So I'd rather spend my time actually understanding fundamentals (in my case, CSS at the moment), than trying to keep up with the frequently changing AI tools and models.

Once the tools and models stabilize more (as well as the pricing model), there's less risk in me learning something that is no longer relevant.

Except when I choose to wait on learning how to use AI tools effectively, I get told I am going to be "left behind".

You make it seem like AI coding has already "totally changed" our jobs. This is exactly the FOMO the article talks about ("until its too late"). It hasn't. I'm still using the same workflows without AI tools, and so are most of my teammates.

To be fair, I would have talked like you in January 2026. But things have changed since Feb.

The Nintendo Wii was like this too.

Yes, but also I often argue that the Wii is better "VR" than VR, because you can play with your friends, which is probably the real "killer app" of gaming?

I agree. This is one area I'm hoping that AI tools can help with. Given a complex codebase that no one understands, the ability to have an agent review the code change is at least better than nothing at all.

I would normally agree, but I think the "code is a liability" quote assumes that humans are reading and modifying the code. If AI tools are also reading and modifying their own code, is that still true?

You have to be able to express the change you want in natural language. This is not always possible due to ambiguity.

Next to that, eventually you run into the same issue that we humans run into: no more context windows.

But we as software engineers have learned to abstract away components, to reduce the cognitive load when writing code. E.g., when you write file you don't deal with syscalls anymore.

This is different with AI. It doesn't abstract away things, which means you requesting a change might make the AI make a LOT of changes to the same pattern, but this can cause behavior to change in ways you haven't anticipated, haven't tested, or haven't seen yet.

And because it's so much code to review, it doesn't get the same scrutiny.

What happens when there’s a service outage and you cannot debug code without an agent?

Switch to a rival service that doesn't have an outage. There are at least half a dozen competent hosted LLM vendors for coding now (Anthropic, OpenAI, Gemini, Mistral, Kimi, MiniMax, Qwen, ...)

Like any service outage out of their control, people will find other things to do until it’s over.

Afternoon latte and useless meetings won’t do themselves!

There was nothing stopping everyone from using continuous delivery today, yet many companies still rely on long cycles, manual testing and handovers. The problem isn't the tooling, it's the people.

I don't understand this part either. At some point we're writing software for people to use, and there has to be someone who comes up with the requirements based on what people want. AI doesn't change this fact.

Is there an equivalent for the JS ecosystem? If not, having Dependabot update dependencies automatically after a cooldown still seems like a better alernative, since you are likely to never update dependencies at all if it's not automatic.

RenovateBot supports a ton of languages, and ime works much better for the npm ecosystem than Dependabot. Especially true if you use an alternative package manager like yarn/pnpm.

Too bad dependabot cooldowns are brain-dead. If you set a cooldown for one week, and your dependency can't get their act together and makes a release daily, it'll start making PRs for the first (oldest) release in the series after a week even though there's nothing cool about the release cadence.

The cooldown is to allow vulnerabilities to be discovered. So auto update on passing tests, which should include an npm audit check.

I use Gemini for personal stuff such as travel planning and research on how to fix something, which product to buy, etc. My company has as Pro subscription so I use that instead of ChatGPT.