I want to spend some time on what TenStreamsLab is not.

This might seem like an odd way to introduce a project — leading with the negatives, clearing the deck before laying down what's actually there. But I spent fifteen years in law, and a good portion of that time was spent reading contracts that had been written by people who were far more interested in sounding impressive than being clear. The damage those documents caused was usually proportional to the amount of vague aspiration embedded in them.

So let's be clear.

This is not a course. I am not selling you a system. I am not going to offer you a masterclass on passive income for six hundred and ninety-seven dollars or access to my private community for forty-seven dollars a month with a special founding member rate that expires in twenty-two minutes.

I have nothing against people who build those things. Some of them are genuinely valuable. But that's not what I'm doing, and I want to say it plainly because the space I'm operating in — AI tools, income diversification, online business — is absolutely saturated with people who started a project, learned one thing, and immediately packaged that one thing into a paid product before they had any idea whether it actually worked.

The newsletter you're reading right now is one of the income streams I'm testing. The honest documentation of this experiment has value — that's the bet. But the documentation has to be honest for the bet to work. The moment I start shaping the story to support a product pitch, I've burned the asset. So the constraint is structural, not just ethical.

Here's what we're actually building.

TenStreamsLab is a research-and-testing operation. The output is documented experiments: what we tried, what the data showed, what we're doing next. We publish that documentation publicly, in this newsletter, because we believe there is an audience of people who want to follow along with a process that's real rather than curated.

Alongside the newsletter, we're building a library of practical resources — guides, toolkits, frameworks — that capture what we learn along the way. The free toolkit at start.tenstreamslab.com is the beginning of that library. As the experiments produce findings, the library grows.

There is also a paid community called The Upstream, which I'll talk about more in a future issue. The short version: it's for people who want to go deeper than the newsletter, ask questions, and get access to what we're testing before it's published. It's not a course. It's a conversation.

Those three things — the newsletter, the free resources, and The Upstream — are the infrastructure. The streams we're testing are separate from that infrastructure. Some of them live outside digital publishing entirely.

I also want to be honest about the scale we're talking about.

TenStreamsLab is not a startup with a team and a funding round. It is a small operation run by a small number of people, with AI filling in the gaps. I'm not going to pretend otherwise.

What that means in practice: things will sometimes be slower than a well-resourced operation. There will be weeks where one experiment gets more attention than another because that's where the data is pointing. Some streams that look promising on paper will turn out to be dead ends, and we'll document that when it happens. This is not a glossy case study. It's a field notebook.

The reason I think that's actually valuable — rather than just an honest limitation — is that most documentation of this kind of experiment is written in retrospect, after the person already knows it worked. It's narrative shaped by outcome. What I'm offering is documentation in real time, which means you get the uncertainty along with the optimism, the wrong turns along with the right ones.

If you have a journalism degree, a law degree, and twenty years of practicing careful observation and argumentation, real-time honest documentation is something you're actually qualified to produce. That's my edge in this space and I intend to use it.

One more thing I want to name directly: I am not an AI expert. I am not a software engineer. I am not a serial entrepreneur with five exits. I'm a writer who started using AI seriously, found it genuinely transformative, and decided the most honest thing to do was to document the experience of building with it rather than positioning myself as someone who has already figured it out.

The people who've already figured it out are not the ones you want writing this newsletter. The ones you want writing it are somewhere in the middle — far enough along to be credible, not so far along that they've forgotten what the early stages actually feel like.

That's where I am. And I suspect that's where a lot of you are, too.

This Week in AI: Several major AI labs have released model capability benchmarks recently that are impressive on paper and require a fair amount of skepticism in practice. Benchmarks measure what benchmarks measure. The question that matters for working practitioners is almost always narrower: can this tool do the specific thing I need it to do, reliably, without me babysitting it? That's a different test.

Want to see the full scope of what we're building? Join the conversation over at The Upstream — a community for people running their own experiments. Details at start.tenstreamslab.com.