Welcome to Plain Inference

Plain Inference began life as Daily(Ish) GenAI Experiments, a newsletter built on a simple theory: watching someone fumble through a tool in public is more persuasive than being told the tool matters. So that's what I did. I tried things out in the open—the genuinely useful and the faintly embarrassing alike—mostly because too many smart people I knew, people whose judgment I trusted on everything else, were watching this technology from a careful distance. They were reading about it instead of putting their hands on it. I wanted to give them a reason to actually try something.

It worked. Over 42 editions and eight months, I watched people who'd been hanging back start using these tools daily. The broader shift was bigger than any one newsletter, but I saw it happen up close—in the replies, in the conversations, in people who'd written these tools off telling me they'd started experimenting again.

Now the newsletter is evolving, because the interesting question has changed. Once everyone is already in the water, the question is no longer whether a tool is worth trying. It's what that tool quietly changes for the person using it.

That's what I want to spend my attention on now. The distance between how someone who understands these systems navigates them and how everyone else does. The difference between asking a tool a question and handing it a task to carry out on your behalf. The things we agree to without quite registering that we've agreed to anything. Each of those shifts creates a need that didn't exist a year ago, and many of them are still going unmet—moments where a tool leaves an ordinary person exposed in a way they can't easily see, and no one has yet built the thing that would close the gap. That intersection of new risk and unbuilt solution is where I'm headed: less on what a tool can do for us, more on what it does to us, and on the room that leaves for anyone willing to build for these new needs.

The experiments aren't going anywhere. They're still how I think, and I don't trust a claim about a tool I haven't run myself. What changes is where each piece lands. Instead of stopping at what worked and what didn't, I'm pulling the lens up to what it means—for the person using the tool, and for anyone paying attention to where the openings are.

So if you came for the hands-on experiments, they stay the spine of this. The only change is that each one now goes a step further, under a new banner: Plain Inference—what new technology really means for the people it touches, and where that leaves room to build.

This site is home to everything—every past edition and everything new. The first piece working this way is coming soon. Look around, start wherever you like, and if it resonates, subscribe and come along.

Get curious. Stay grounded. Keep testing.

—Kimberly