If you’re a decision maker at your company, you need to be on the bleeding edge of, well, everything. But before you go signing up for seminars, conferences, lunch ‘n learns, and all that jazz, just know there’s a far better (and simpler) way: Subscribing to The Deep View.

This daily newsletter condenses everything you need to know about the latest and greatest AI developments into a 5-minute read. Squeeze it into your morning coffee break and before you know it, you’ll be an expert too.

Subscribe right here. It’s totally free, wildly informative, and trusted by 600,000+ readers at Google, Meta, Microsoft, and beyond.

A few months ago, I was reviewing a contract.

I fed the document into our AI tool, watched it surface the key terms, flag the risk provisions, and spit out a clean summary. It was useful and incredibly efficient—the tool did exactly what it was supposed to do.

And then I noticed something that scared me.

My eyes had glazed over the actual contract. I hadn't really read it. Not in the focused way I know I need to as a lawyer. I thought I had, in the way you think you're listening to someone when you're actually just waiting for them to finish. The words had passed in front of me, but I hadn't engaged with them. Because I didn't technically need to. The AI had it covered. And that, it turns out, is exactly the problem.

It was a yikes moment.

I work for an AI incubator. We build AI startups. AI is woven into my team's workflows in ways I'm proud of, and our legal work is better for it. We're finding new use cases every week, recovering time, operating at a level that wouldn't have been possible a few years ago. I am not a skeptic. I am a true believer.

But that moment with the contract rattled me. Because I realized that if I'm not careful, all this efficiency could cost me skills and capabilities I've spent years building.

That's what this issue is about.

Most of the AI conversation in legal circles is about adoption: which tools to use, how to use them safely, what policies to put in place for your team. That's all important. We've covered some of it here.

But there's a conversation happening behind closed doors and among friends that I think deserves more airtime. It's the one about what happens to us as lawyers when AI handles more and more of the work.

My worry isn't that AI will get something wrong (though it will, and you should always check).

And it's not about AI taking my job.

It's about what happens to our skills when we stop exercising them.

As we work to automate more of our processes and build out agents, I think about two categories of risk. The first is technical legal skills like the ability to read a contract carefully, spot the thing that doesn't fit, catch the clause that looks standard but isn't. The second is critical reasoning; it’s the ability to engage with a hard problem, think it through, and arrive at a well-advised judgment.

Both require practice and both atrophy without it.

The scary thing about atrophy is that you usually don't notice it happening. The muscle just gets gradually weaker, until one day you need it and it isn't quite there.

I want to be clear that I'm not suggesting you use AI less. I'm suggesting you use it—and decide when not to use it—more intentionally. Here are four strategies I've been actively trying to build into my own practice.

When a complex question lands on your desk, resist the urge to start by asking AI. Take the time to process and think it through. Do the things you used to do, like go for a walk or brainstorm with a colleague. Then sketch an answer before you consult with AI.

This is about keeping the part of your brain that does hard legal thinking in good working order. Using AI to pressure-test your analysis after you've developed it is very different from outsourcing the analysis to AI and signing off on the result. The outcome may look similar. The skill-building is not. (And often, the outcome doesn't look similar for complex problems involving human judgment—that's part of why AI won't replace you any time soon).

For repetitive work, you absolutely can and should use AI. First-pass review of a routine NDA, summarizing a long document against a known checklist, or generating a first draft of a policy you've written a dozen times are exactly what these tools are for. They free you up for the higher-stakes work where your judgment is the whole point.

But when your judgment is the value-add, exercise it.