… And for any new readers, welcome to Signal // Noise — the newsletter read by CEOs, founders, & scrappy builders every Thursday. Each week, what I’m listening to, one deep dive, notes from the field, three links worth your time. No buzz, no bullshit.

🎧 Want the whole vibe? The 2026 playlist is right here.

A post and chart went viral this past week (2M views and counting).

If you haven’t seen it, shame on you.

John Burn-Murdoch at the FT plotted what's happened to app releases since agentic AI showed up, and the line goes damn near vertical.

Releases are up around 80% over the 2024 baseline.

Then there's a second line. The one tracking whether any of those apps actually got used.

It's flat. If anything, it's drifting down.

So everyone's building, and almost nobody's getting traction.

Caveat: app reviews might not be the best proxy for engagement, but, I still think the lesson here is straightforward, especially when you consider that the chart comes from a study out of MIT and NBER, bluntly titled "Writing Code vs. Shipping Code."

The researchers in that study found AI agents boost the raw amount of code you write by 180%, but those gains shrink the further down the chain you go, falling to 50% for projects and just 30% for what actually ships.

They call it the weak-link hypothesis: AI makes the easy part easier, and the hard part, the human part, stays exactly as hard.

Building isn’t the real bottleneck now. It just isn’t.

So here's my take.

When everyone has the same agents, building stops being a moat.

The things that don't get easier with AI are the things that really matter.

And the biggest one, my friends, is distribution.

Remember the old Wu-Tang song?

Cash Rules Everything Around Me.

Welp, right now for founders, it's D.R.E.A.M.

Distribution Rules Everything Around Me.

I wrote a while back about how the moats have changed.

Old moats were capital and scale.

New moats are data, distribution, and depth.

I still believe in all three.

But if you're ranking them for where we are right now, distribution is the one carrying the most weight. AI just made the supply of products basically infinite, and attention didn't get any bigger.

Think about what actually wins now. Owning a channel beats the crap out of renting one, every time.

HubSpot didn't out-feature anybody. They built a blog and a pile of free tools, and they owned the front door to their funnel for a decade.

Figma didn't buy its growth either. The product spread because every shared link pulled in three more people who needed access. That's distribution baked right into the thing.

Now picture the founder shipping their fourth feature this month, sure the next one's the one that really helps it “take off”. But it isn't.

The product was probably fine two features ago.

The problem was that nobody knew it existed.

The other thing that doesn't commoditize is taste.

When anyone can generate a halfway-decent app in a weekend, the edge is knowing which one's worth making, who it's for, and what to cut.

That is a level of judgement that AI can’t really just ‘hand’ to you.

Distribution is the thing that gets you in the room, and taste, hopefully, is what keeps you there.

So here's the somewhat sad but realistic read on that chart.

Most of those new apps aren't dying because they're bad.

They're dying because building is the part founders enjoy, and selling is the part they avoid.

AI just made it that much easier to hide in the fun part.

Build less. Sell more.

Managing Google Analytics and Search Console across a team is still a dumpster fire: shared logins break as you scale, and chasing people for access gets old fast. The workaround making the rounds is to route everything through a shared 1Password vault so nobody sees the real credentials, then pipe verification codes into a dedicated Slack channel and a shared Google Voice or Vonage number so nobody's pinging you for 2FA.

Goldman Sachs CEO David Solomon makes the case that the AI jobs panic is overblown, arguing the economy will adapt the way it has to every disruption from electricity to Excel. He grants AI could automate a quarter of current work hours, but bets that time gets reallocated to higher-value work instead of vanishing. Worth a read whether or not you buy it.

After an epic Spurs collapse and the Knicks taking a 3-1 series lead, there’s certainly some big market NYC hype right now, sure. But it also makes it the right time to check out The Ringer's take on why the league feels broken the rest of the year and why the regular season's such a slog. The part that'll stick with founders: how a company quietly takes on the personality of its owner, for better or worse. If you’re an NBA fan, or just a student of business CEOs, this is a must-listen.

Before you ship another feature, consider if you really have distribution locked.

And ask yourself - what is my main source of distribution?

Hit reply and tell me.

And until next time, thanks for reading.

Jordan