The AI wave has arrived, and it's no longer only hype. Social media is flooded with discussions about it, and everyone involved in software development feels the pressure. Where a few months ago we were cautiously dipping our toes in, we've now been thrown into the deep end. What's becoming clear is that the industry is splitting into two speeds, and the gap between them is widening fast.

Taking the plunge

The first group is fully embracing AI. They're leveraging agents, talking to them through Whispr Flow or Superwhisper, running Codex and Claude Code in parallel across multiple terminals. Their workflows might still be held together with duct tape, but they can already see the productivity gains on the horizon. These people are figuring it out as they go, and in doing so, they're building a lead that will become increasingly difficult to catch up to.

This moment feels unique. It reminds me of when Google first emerged. Some people embraced search engines early, not because they knew exactly what it would lead to, but because they were curious. They learned the quirks. How to phrase a query and when to dig into page two, three, four, ... for the hidden gems. Up until recently, you could still spot the difference between people who had been Googling for decades and those who came late. Googling was a skill, one that wasn't taught but learned over years of use. AI is heading the same way.

The naysayers

On the other end of the spectrum are the sceptics. The naysayers. The ones who dismiss it all as ridiculous hype because some AI-generated pull request didn't follow their coding standards. Outrageous. They resist change. AI still can't produce code exactly the way they want it, and they insist their standards are much higher than that.

This group poses a real risk to your company. If the naysayers become the loudest voices in the room, they push the enthusiasts into a corner. And when that happens, the sideliners who might have been nudged into the pool now stay on the deck entirely. They hear the dominant narrative, assume AI isn't worth exploring, and never build the firsthand knowledge that comes from actually using the technology. They never learn how to Google.

The resistance might be institutional, as is common in highly regulated industries. Your company might not support AI adoption at all. Getting a Claude Code subscription requires jumping through hoops, or it's blocked entirely because leadership is worried about IP leaking to an LLM. That concern isn't unreasonable. But if it stops the conversation entirely, you're leaving your team without the tools they need to stay competitive. Even if full-fledged AI adoption isn't on the table, there are steps you can take. Can you provide access to local models? Can you create a sandboxed environment for experimentation? The question shouldn't be whether to adopt AI, but how you can give your people the leverage to use it responsibly.

Attract the divers

Have you looked at your software engineering job descriptions lately? I see vacancies with zero mention of AI. Some are a copy-paste from a few years ago. If that sounds familiar, it's time to revisit them.

You want to attract the enthusiasts. At a minimum, you need people who are open to trying the technology. Hiring someone who is resistant to change, or worse, actively hostile toward it, will do more harm than not hiring anyone at all. This goes beyond your engineering team and affects your company's culture as a whole.

Curiosity has become a core hiring signal. Developers who are genuinely curious about AI tooling, experiment with it, and integrate it into their workflow are more productive and adaptable. If your job vacancy doesn't reflect this shift, you're either attracting the wrong candidates or signalling that your company hasn't caught up. Neither of those is a position you want to be in.