When we launched Murror's delete feature, we thought of it as housekeeping. Users write messy drafts, clean up later. Standard stuff.
Then we looked at the data. About 15% of entries were being deleted within minutes of writing them. Not drafts. Not typos. Full, completed entries that users wrote, read back, and chose to erase.
Three states passed laws this year about AI that acts like it cares. If your product does that on purpose - not as a joke, as the actual feature - you're not adjacent to this. You're the definition.
I build an app that helps people process their emotions with an AI that remembers what they told it last time. When I went looking for whether that puts Murror inside any of the new AI-companion laws, the honest answer was: probably, and I wouldn't have guessed that a year ago.
In 2026 a growing share of your customers form their first impression of your product inside a chatbot you can't see and you don't get to write the recommendation.
Distribution used to have a front door you could find. Someone had a problem, they typed it into Google, they scrolled a page of links, and if you'd done your SEO homework or timed a launch right, you were on it. You could see the door. You could optimize the door.
From September 2026 you can't submit an update without saying whether your app has social media capabilities. Say yes and Apple applies a minimum 13+ rating and drops you into a Time Allowance parents can cap. The definition is broader than having a feed, and your secondary category is load-bearing too.
If you've shipped an update in the last few days you've already met this question. If you haven't, you will on your next one, and I'd rather you met it here than at 11pm with a release branch open.
In 2026 the bottleneck isn't code anymore it's the person typing the prompts. Here's what I do to keep that person functional.
A year ago, "I built a working app this weekend" was a flex. Now it's a Tuesday. Vibe coding went mainstream, agents write most of the first draft, and the gap between an idea and a shipped product has basically collapsed. That's the good news, and it's genuinely great.
Every week we get the same request: "Can I just talk to the AI?"
They want a chat interface. A back-and-forth. Something that feels like texting a therapist at 2am. And honestly, it would be the easiest feature we could build. The AI is already there. The conversation engine exists. We could ship it in a week.
When someone journals on Murror, our AI doesn't just process text. It reads emotional weight. It picks up on patterns the user might not see yet -- the way their language shifts when they talk about work versus family, the recurring themes they circle back to every few weeks, the gradual change in tone that might signal something deeper.
We built this because it makes the product better. The AI can ask more relevant questions, create more meaningful reflections, and know when to give space versus when to gently prompt.
When we first launched Murror, the number one feature request was always the same: "Can I talk to the AI?"
Users wanted a chatbot. Something they could vent to, ask for advice, get instant feedback from. Every competitor had one. Every investor asked why we didn't. So we built it.
Building got cheap this year. Judging whether the thing is actually good did not. In 2026 the eval is your spec, and it's the step nearly everyone leaves out.
There's a moment now that didn't exist three years ago. You describe a feature, an agent writes it, and forty minutes later it's running. It works, sort of. It answers, it responds, it does the thing. And you sit there with a genuinely hard question you have no tool for: is it any good?
That question used to answer itself. Old software was deterministic same input, same output, and you wrote a test that either passed or failed. AI features don't work like that. Ask the same thing twice and you get two different answers, both plausible, one subtly wrong. "It ran without erroring" tells you almost nothing about whether it's right, helpful, or safe. The build got easy. The judging got hard. And the gap between those two is where most maker products quietly fall apart around week three.
The discipline that closes the gap has a name this year: eval-driven development. The idea is boring and it's correct. Before you tune a prompt or swap a model, you write down what "good" means as a set of concrete examples, and you check every change against them. Not vibes. Not "seems better to me." A fixed set of cases with expected behavior that you can rerun. The whole shift in how serious teams build with AI in 2026 is the move away from vibes-based iteration toward something you can actually rerun and compare. Eval platforms are multiplying for a reason the demand is real.
