Preventing AI tool use is a talent risk

Playa Viva beach

An interesting topic arose over the holiday period which is people using AI tools for the first time or at least first time in a while and realizing how far they have come since last year. Claude code in particular had new users seemingly everywhere. The strange thing about this for the AI maxi crowd is that we have been using these tools all year. Some of the reviews have been panicked as people realize just how far some of this tooling has come in the last 12 months.

The other side of the coin is those seeing the paradigm shift and rightfully concluding that avoiding these tools or not being allowed to use them at work carries the risk of being left behind. I am more conservative in my timeline for AI than others and do not believe we are 6 months away from all knowledge jobs being replaced. I do not believe this will happen even in the next 10 years since jobs will change as the tools develop.

I do however believe that the improvements to productivity cannot be ignored. I view generative AI tools as similar to the migrations from jQuery to Angular and then React with JSX. The industry built better tools that solved old problems faster and much like the migrations to those technologies no one wants to be left behind.

I distinctly remember the rush for React jobs and companies offering the technology almost as a perk. I saw this happen again with Node where the best people went there instead of staying on legacy PHP and Java projects.

The companies that waited lost their best developers to those offering React roles. The individual contributors who delayed learning it found themselves explaining jQuery experience in interviews where everyone wanted React.

React released in 2013, by 2016-2017 it was the dominant choice. That’s a 3-4 year window where career trajectories diverged significantly. Developers who adopted early had their pick of roles and salaries. Those who waited found themselves playing catch-up or stuck in legacy roles.

We’re in that same 3-4 year window right now with AI tools. The developers learning to work effectively with Claude, Codex, and Cursor today will be the ones companies fight over in 2027-2028. The managers blocking AI adoption are making the same mistake as those who banned React “until it proved itself”.

We will see these same migration patterns as high agency, forward thinking people want to be working with future technologies.

So what should companies do? The biggest complaints about generative AI remain valid: very large PRs submitted with little proof reading, copyright and data protection of both the code going in and coming out and, security or compliance risks when the code is not written and understood by a company’s engineers.

These three are the biggest complaints I see from companies hesitant to embrace AI tools. We want our code quality to remain high and our applications to be well understood and secure.

I cannot convince people that distrust Claude Code and Codex to give over the keys to their codebases but if you are a company or manager unwilling to use these tools day to day perhaps I can recommend giving your teams the green light to rely heavily on these tools for prototyping or new internal tools.

We are in an era of self-serve software becoming cheap so take the opportunity to build that dashboard or tool you have kicking around your backlog.

For the company willing to invest in AI I recommend continuing to do so. Promote it in your job ads and find new ways to test and validate your code.

You will lose some candidates doing this, AI has been politicized and the crowd against it will gladly let you know their thoughts. Instead of worrying about them I ask you to focus on the pioneer wanting to work with frontier tools.

We need new workflows and tools which will only come with experience and use. Encourage your team to use AI and see where it takes you.

The talent you want to hire is already using these tools. The question isn’t whether to adopt AI. It’s whether you want to be the company people leave to find it, or the one they join because you have already figured it out.

Photo: This week’s photo comes from a recent trip to Mexico.

0 Shares:
You May Also Like
Read More

Playing the Right Game

In jiu jitsu there’s a training method called positional sparring, sometimes called specific training. These two names refer…
Read More

No more feeds, please

The perpetual scroll of social media feeds has become boring to me. I open up Reddit, Twitter, Hacker…
BSides Wrapup
Read More

BSides Wrap-up

BSides Wrap-up Right now I’m sitting in the airport bar, reflecting on the great weekend that Megan and…