Overestimating the short term

Everyone is jumping on the "AI will replace software engineers" train. Software consultancies, product companies, analysts. The whole industry agreed on the narrative before checking the ground truth.

I'm not saying the future doesn't converge to some version of this. It might. But we're oversimplifying the short term.

I keep hearing that software engineers are becoming obsolete. Meanwhile, I see no significant increase in digital literacy. None. The gap between what AI can theoretically do and what people can actually use it for is enormous. Is the expectation that everyone suddenly knows how to run software products? That they'll go from struggling with spreadsheets to orchestrating AI agents overnight?

What I actually see is AI pushing mass software adoption. People who were browser users at best are now looking into automation. Not because they're replacing software engineers, but because there's suddenly visibility into what's possible and the barrier to entry dropped. These are people who had no idea what software could do for them. I've had family members ask me what I actually do for work.

"What do you mean, 'software'?"

That's the baseline we're working with.

Expecting that population to become digitally capable overnight is optimistic at best.

I do think there's a future where models generate working products from a prompt and some back-and-forth. When that happens, even people without much digital literacy will solve their own problems. But those people were never customers of software engineering companies. They weren't buying custom software before. They're not taking revenue away from anyone.

The people who buy software consulting, who commission enterprise platforms and complex integrations, they have real problems that won't be solved by prompting a chatbot. Not yet. The gap between "I can vibe code a to-do app" and "I need a system that handles compliance across twelve jurisdictions" is not closing next quarter.

But the biggest risk isn't any of this. The biggest risk is the confusion.

When everything could be possible, nobody knows what's actually possible. The lines are completely blurry. And that uncertainty is pulling the handbrake on the entire industry. Companies are freezing software purchases. Not because they found something better, but because they don't know what to buy anymore. Every vendor promises AI will solve everything. Every analyst says the old way is dead. So the rational response is to wait. Don't commit. See how things shake out.

That's the real damage. Not AI replacing developers. The narrative around AI replacing developers causing a purchasing freeze that hurts the industry more than any technology shift could in this timeframe.

I think about this every day. The future probably does involve fewer traditional software engineers doing the same work. But the short term is not the long term, and confusing the two is how you freeze an entire market. I suppose the billions of dollars flowing into AI funding does benefit from it. Creates budget and demand for a new type of software, which emerges.