Thinking out loud.

Yesterday I flew to London for the Children's BAFTAs. I wasn't there to pick up an award. I was there because SuperBam works with creators who were nominated, and because the conversation backstage at events like this is always more interesting than the stage itself.

Somewhere between the pre-show reception and the third time someone asked me what I did, I ended up in a debate about YouTube kids content with a producer who has been in children's television for thirty years. Her position: the algorithm is destroying children's ability to focus. My position: every generation says this about the new medium, and every generation is wrong about the mechanism while being vaguely right about the anxiety.

The printing press was going to make people lazy readers. Radio was going to make people forget how to think for themselves. Television was going to rot children's brains. The internet was going to destroy attention spans. Social media was going to make everyone narcissistic and depressed. And now the algorithm is doing all of the above, apparently simultaneously.

The pattern isn't that the new medium is dangerous. The pattern is that the new medium changes the distribution of power, and that the people who held power under the old distribution are very good at framing that change as a moral crisis.

None of this means the algorithm is benign. It isn't. But the thing that's actually happening — the real shift that will matter in twenty years — isn't attention spans. It's who gets to be a creator, and who gets paid for it, and how the economics of that flow through to the actual work being made.

That's the conversation worth having. The moral panic is a distraction from it.

Radical idea: dump your Discover / For You / Explore feed and replace it with a deliberately curated list of people whose judgment you trust. Not influencers. Not brands. People.

The algorithm is not your friend. It is optimized for engagement, which is a proxy for the thing you actually want — which is to learn, think better, and find things that are useful or beautiful. The proxy and the thing are not the same. The algorithm doesn't know that.

What does "follow people instead" look like in practice? It looks like: making a list of 40–60 people whose recommendations you have historically found valuable. Not people you agree with on everything — that's an echo chamber. People whose judgment on a specific domain you trust. Someone who curates good long reads. Someone who knows the creator economy better than the press covering it. Someone who is actually building something and writes about what they're learning.

Then muting everything else. Not unfollowing — that's a lot of work and creates social friction. Just muting. The Discover tab goes dark. What's left is the people.

The algorithm gives you what you clicked on last week. People give you what they thought was worth sharing this morning.

This isn't a new idea. It's how smart people have always navigated information overload — by building trusted networks rather than outsourcing curation to a machine with different incentives than theirs. The current moment just makes it feel urgent because the machine got very good very fast.

If you missed the Made on YouTube event, you missed a subtle but important signal: YouTube is still trying to figure out what live means for their creator economy, and the answer is still "maybe shopping?"

I've been in this conversation in various forms since 2019, when I was building Facebook Live Shopping inside Meta's NPE team. The fundamental tension hasn't changed: live video is a great product for building connection and urgency, but it's a terrible product for completing commerce transactions. The two things want to happen at the same time, and they keep interrupting each other.

The shopping experience breaks the viewing experience. The viewing experience breaks the shopping experience. The creator has to choose which one they're doing every thirty seconds. The viewer has to do the same. And the platform is in the middle trying to capture both moments with one interface.

Why it keeps breaking

The platforms keep reaching for the TikTok Shop comparison, but TikTok Shop works (to the degree it works) because the content is short, the purchase intent is high, and the friction from "video" to "buy" is minimal. Live video is different. The attention and the transaction are fighting over the same window.

The category that actually works is entertainment-first live with commerce as an add-on — not commerce-first live with entertainment as the hook. The QVC model works because it's always been transparent about what it is. When you're watching a YouTube creator you follow for their personality and they suddenly pivot to selling you something, there's a small but real trust tax.

Live commerce will eventually work. But it'll work as its own genre, not as a feature bolted onto existing creator formats.

If you have ever worked across multiple time zones, you have heard some version of the same complaint: "It's so hard to find time to get everyone together." This is the wrong problem to solve.

The myth of synchrony — the belief that real work requires people to be present at the same time — is one of the most expensive cultural defaults in knowledge work. It produces meeting-heavy calendars, decisions that wait for quorum, and a baseline assumption that the workday is a shared object that everyone holds at the same time.

Time zones break this assumption by force. And that, it turns out, is a feature.

What distributed time actually gives you

When you can't be in the same room (or the same Zoom), you have to write things down. Decisions get documented because they have to be communicated, not just announced. Context gets written because it can't be absorbed by osmosis. The work becomes legible in a way that co-located work often isn't.

The teams I've worked with that were most distributed were also — not coincidentally — the teams with the best documentation, the clearest decision logs, and the most reliable handoffs. Not because they were better people. Because the time zone gap made legibility a survival requirement.

Synchrony is overrated. What you actually need is alignment — and alignment doesn't require a shared calendar block.

I knew this moment was coming, and yet I wasn't ready for it. Hulu — as a standalone brand, as a specific kind of company, as the place that made me — is effectively gone now, absorbed into Disney's content machinery in a way that makes it unrecognizable from what it was when I joined in 2007.

I joined Hulu when it was genuinely weird. Not weird in a startup-cool way — weird in a "we're not sure if this is legal" way. The television industry had not yet decided whether streaming was an opportunity or an existential threat, and Hulu existed in the gap between those two positions. It was funded by the networks that were most afraid of it and built by people who thought those networks were wrong.

What I learned there — and this is the thing I've carried into every job since — is that there's a difference between a good idea and a product people will actually use. The early streaming wars were full of good ideas that nobody used. Hulu survived because it built something with a real wedge: the back catalog, the next-day episodes, the experience that was just good enough to make you stop torrenting.

What being early actually means

Being early to a platform shift is not the same as being smart. A lot of people were early to streaming. What being early at Hulu gave me was a front-row seat to what happens when an industry has to reinvent itself faster than it wants to — and a set of instincts about where the real leverage is in those moments that I've used ever since.

The thing about formative experiences is that you don't know they're formative until much later. I knew Hulu was interesting. I didn't know it was the education.

We moved to Europe. Not for a vacation. Not for a sabbatical. Not for an adventure, though it is certainly that. We moved because we looked at the life we were building in Los Angeles and decided that the architecture was wrong — that the version of success we were optimizing for wasn't the version we actually wanted when we said it out loud.

The decision took about two years to fully form and about forty-eight hours to execute once it was made. That's usually how the big decisions work. The thinking is slow. The action is fast.

We chose France for reasons that sound romantic and are also practical: the quality of life is genuinely high in ways that matter to us — food, pace, outdoor space, proximity to places we want to take our kids. The tax situation for Americans abroad is complicated and not particularly favorable, and we did that math. We did it anyway.

What we gave up

A network that took twenty years to build. The ability to grab coffee with people who are building interesting things. The casual density of information that comes from being in the right rooms. These are real costs. I'm not pretending they aren't.

What we gained: mornings that belong to us. A pace that makes space for the kids to be bored, which is the precondition for being creative. A default assumption that the day has room in it, rather than that the room has to be fought for. A version of presence that I had been trying to manufacture in the margins of a schedule that didn't actually have margins.

The question wasn't "can we afford to move?" The question was "can we afford not to?" Those are different calculations.

We've been here just over a year. The kids are learning French. We are learning French, more slowly. The light in the south is different than I expected — more varied, more directional. I've started noticing things I didn't have the bandwidth to notice before. I think that's what we came for.