About Julie

The lab came first

I’m a scientist by training and a science writer by profession, in that order, and the order matters.

I studied chemistry and biology, did graduate research at Dartmouth, and then spent a decade at the bench: drug-discovery research at Bristol Myers Squibb, where a single protocol could carry months of work and millions of dollars, and then research at Sandia National Laboratories. The bench teaches you one thing above everything else – to notice when a claim isn’t backed up. You learn to catch yourself being wrong before someone else does.

I didn’t know it yet, but that habit was going to become the most valuable thing I own.

Then the writing

I moved to Germany at the end of 2010 for Sebastian, my husband, a German scientist. I landed on the Lower Rhine and switched from the bench to writing full time – first in-house at QIAGEN, and since 2016 as a freelancer for regulated biotech and pharma clients.

For a while I still felt like a researcher who happened to write well. That changed. I’m a writer now – one who understands the science, and who knows what the person reading actually needs because I used to be that person, standing at the instrument, drowning in documentation that wasn’t written for me.

I also run my whole life here in my non-native language: the tax authorities, the health insurance, the Handwerker who come to fix things, the geothermal heat pump. I mention this because building systems that keep a complicated life running is not new to me. AI is the newest one, that’s all.

Then AI showed up, and I got scared

When everybody started talking about ChatGPT, I thought: this is either the most amazing thing that’s ever happened, or it’s going to take my job away from me. My edge was doing the scientific work in native English. I was sure AI would hand that edge to everyone. Maybe that was ridiculous, but that’s what I thought.

So at 51, I decided not to be the person who can’t work the new thing. I started using it. Day one was salad recipes from what was in the fridge. Within ten days I’d used it for cooking, gardening, German bureaucracy, and my professional science work.

And nine days in, it lied to me. I asked for a scientific reference, pushed for a more interesting example than the boring true one, and it produced a study that didn’t exist – real author, real journal, properly formatted citation, completely fake. I went looking for the paper and couldn’t find it, because there was nothing to find. I’ve been checking AI’s work ever since. That reflex, more than any tool, is what I teach.

What three years of daily use built

My first attempts went the way most people’s do. I made a custom GPT to write my blog posts, fed it my style and my guidelines, and got back well-written garbage – fine sentences that weren’t mine and weren’t in my brain, because I hadn’t written them. Fixing them took longer than writing from scratch. AI made me slower, because I refuse to put slop out under my name.

The thing that finally worked wasn’t a better prompt. It was a system: my notes, my decisions, my ongoing projects, kept as plain files on my computer, with an AI that reads them before we work. It’s not impressive to look at, and it’s not supposed to be. But the AI knows my work now, I stopped re-explaining myself every session, and ideas that used to die in my notes app get found and used. It’s just files. If a better tool ships next Tuesday, I point it at the same folder and lose nothing.

Why I’m doing this in public

I started learning AI out of fear. I keep going because it’s opening doors – and I don’t know what those doors look like. None of us do.

What I do know is who I keep meeting: smart, busy professionals who know AI could help them but don’t know where to start, and don’t have time to dig through hype to find out. Most of the people teaching this don’t look like me and aren’t accountable the way I am. I’m mid-career. I write for clients where being wrong has consequences. And I’m exactly one step ahead of you, which I’ve come to think is the most useful place a teacher can stand.

I should be honest about the rest, too. I’m a smart person, but I procrastinate, I get distracted, and I have a 51-year-old brain that doesn’t hold things the way it did 25 years ago. The system exists because of that, not despite it. If I can do it, you can do it.