Microsoft yanked AI features from MS Paint
Microsoft was forced to remove AI features it had crammed into MS Paint and other products after customers rejected them as making the products worse.
Where this was said
Why 37signals doesn't do roadmaps
At 20:30 · chapter starts 7:37
DHH introduces the concept from Getting Real: the illusion of agreement. A bullet on a roadmap says 'upload permissions' or 'guest access' — one line that customers read and project their entire mental model onto. The feature ships. It's in the right vicinity but has the wrong handle, the wrong shape, the wrong defaults. Now the customer feels suckered: they bought the product on a promise that never existed except in their imagination. [1] — David Heinemeier Hansson "A one-line bullet on a roadmap means nothing. Customers imagine their specific version of 'calendar' or 'guest access,' while the final pro…" 09:54 Jason makes this visceral with the calendar example: announce you're adding a calendar, and every customer imagines their own perfect calendar. The one that ships won't sync with Outlook, won't have an agenda view, won't allow jumping four years ahead. Useless, they'll say — even though a fully functional calendar did ship. The word 'calendar' was never a promise about any specific capability; it was just wrapping paper. And the illusion, Jason observes, only grows larger and stranger as you move outside your own team: internal misalignment is bad enough, but external misalignment with customers who share none of your vocabulary or mental model is an order of magnitude worse.
Near-term planning horizon: ~1 month
37signals plans only about one month ahead at any given time, down from a previous six-week cycle, intentionally avoiding long-horizon roadmaps.
Why 37signals Doesn't Make Roadmaps
37signals plans roughly one month ahead and nothing more. Launching one feature opens trails to the next — rigidly pre-planning six months out kills that optionality. Basecamp's 22-year track record is the argument: if you're worried about uncertainty, look at the longevity.
22 years of Basecamp without roadmaps
Basecamp has operated successfully for over 22 years without public roadmaps, relying instead on iterative, near-term planning.
Roadmaps create illusions of agreement
A bullet point on a roadmap makes customers believe a future feature will solve their specific problem, when in reality the final shape of the feature may not match their expectations at all.
Roadmaps Are Illusions of Agreement
A one-line bullet on a roadmap means nothing. Customers imagine their specific version of 'calendar' or 'guest access,' while the final product takes a completely different shape. Buy software for what it is today — everything else is gravy.
Forward promises always end in regret
Every time 37signals has publicly promised a future feature on a timeline, it has ended in regret — not necessarily because the feature was wrong, but because the deadline was constraining.
Similar snapshots
50 free tools built total
Bhanu and his team built approximately 50 free tools to attract search traffic, each linked back to SiteGPT.
New tool built in under 5 minutes
With AI coding tools like Cursor, Bhanu can now create a new free marketing tool in less than 5 minutes by referencing existing tools.
KD filter < 10 for rankable keywords
Bhanu filters Ahrefs keyword results to show only those with a keyword difficulty below 10, making them realistic ranking targets for any decent website.
Minimum 1,000 monthly search volume
Bhanu sets a minimum search volume of 1,000 monthly searches when selecting keywords to target with free tools.
20 downloads/day at launch
PropGPT averaged 20 downloads per day right after launching on the App Store through influencer marketing.
4 months of rebuilding with no marketing
Eyal and Yali shut down all marketing and spent 4 months completely rebuilding PropGPT from scratch.
Over 40,000 downloads total
PropGPT has accumulated over 40,000 total downloads since launch.
LLM costs only $20/month
PropGPT's large language model (AI) operating costs are just $20 per month, and the cost is continually falling.
Ad monetization for games
Ad-based monetization works well for game apps where users spend extended time in-session, as seen with Grid and Wordle.
