that's a good approach. haha

google is the main source. also there are some github repositions and reddit links.

Hey jruohonen, great point. it depends on what does "survive" mean.

Monetize-able GPTs? Not many for sure. What will happen IMO is established companies will build official GPTs to better serve existing customers.

GPTs for personal use? Many GPTs will be abandoned and not be used by anyone. That's okay for one-off use cases.

There will be a loooooong tail of GPTs that people would use from time to time. There will be a lot of diversity but the top GPTs will eventually be like the top apps in the app store IMO.

> Monetize-able GPTs? Not many for sure. What will happen IMO is established companies will build official GPTs to better serve existing customers.

Care to elaborate whether the supposed plan is again a classical two-sided market setup (i.e., OpenAI et al. monetize the companies using their models and the companies then monetize their users)?

If you're a front-end dev, you'd probably worked with PMs on building features that helps the product grow. This guide covers many of those features.

If you're a backend dev, you'd probably been asked to create database tables to save marketing or business data. AirTable allows many business people to create tables and gather info , and selling info without involving backend dev.

It will be interesting to see how the self-serve business tools (nocode) will affect the type of work devs will be asked to do in the future.

It will make the work much more interesting, but potentially make it harder to break into

Super impressive work since I first saw it on IndieHackers. Congrats!

Hell yeah. Lets gooo

One of the things that bothered me for years was the lack of clarity around Product-Market Fit. So far all the explanations are a posteriori - you can only observe or measure them. E.g. "You've found PMF when you sell it faster than you can make it", or "You can measure PMF by asking your users if they'd be sad if the product is taken away from them".

Can't we systematically get to it step by step instead of observing a big black box?

Last week, one model crystallized in my mind: Product Market Fit consists of 5 "fits" under the surface, each of which can be measured separately, which I summarized in the Twitter thread.

Would like to hear your opinion on if this makes sense to you.

Reading an article this in-depth and pure on blogger brings me back to 2004, the good ol' days

how does it work? :-D

Thanks a bunch!

Priorities took 2 weeks. I decided it's time to build it on March 7th, and it's approved yesterday.

Goals took me 2 months to launch, started from an empty repo, then polished over the next 6 months.

Thank you for asking!