Comment by thevibesman
3 hours ago
"The Lean Startup" was recommended to me about 10 years ago in the short phase when I was trying to build a startup (cool prototypes, no business model or product market fit, and starting the project with half the seed money I thought I needed to get it off the ground made it a short project).
It was a big influence on me and something I recommend and quote often.
I'm curious if your perspectives on the topics of "The Lean Startup" have changed I the era of AI tools. Particularly curious what you think about the role of MVPs to test a market.
This has been on my mind the past few weeks because of a recent experience during a company hackathon:
A few years ago I gave a talk about prototyping and MVPs at the Audio Developers Conference and in this talk to explain the concept of an MVP I proposed a silly idea for an audio plug-in (that replaces a singers voice with the sound of flamingos) as a demonstration of how we might test that there is a market for this plug-in before building it. I gave some examples of how we could test this like a landing page MVP, concierge MVP, etc.
Recently during the two days of this company hackathon I was too busy to do a project of my own because I was helping on-board colleagues with Claude and getting sucked into some leadership meetings. During the demos meeting I decided to try to build my voice replaced with flamingos plug-in and built a working plug-in in under two hours and this got me thinking:
If I can build a real functioning plug-in that a user can try in their host application in less than two hours why would I use non-software MVPs to test a market when I can build working software just as fast or faster than I could setup a non-software MVP a few years ago.
Of course there is more to learn from "lean" than just MVP (I'm also a big fan of the andon cord and the 5 why's)
(to anyone commenting on vibe coding I looked at the code and while not all of it was ideal I wouldn't consider this "vibe coded" and for serving the purpose of an MVP a couple things in the code that were a little funny are not a problem)
Definitely, AI makes certain kinds of MVPs a lot easier to produce, so that's great. Unfortunately, some people are being encouraged to outsource their own thinking to the AI. For them, it doesn't seem to make the learning any faster. In fact, I know some people that it has made the learning quite obviously slower.