(2023-04-21) Baschez A New Kind Of Startup Is Coming

Nathan Baschez: A New Kind of Startup is Coming. Normally when engineers talk about Conway’s Law, they’re thinking about how managers should organize their employees to yield better systems design. But the law goes the other way, too: The systems we work with determine the organizations we get.

In the dot-com-bubble era

The internet was brand-new, so building even a relatively simple web service required significant investment

After the bubble burst and the MBAs fled the valley, there was a long period of time during which not many startups were built

But as time passed, startup activity began to heat up again. There were two major catalysts. First, open source web frameworks like Ruby on Rails launched. This made it much easier for smaller teams to build products. Second, cloud computing became a thing.

The archetypal startup org became much simpler: just a handful of people, mostly technical. They were able to ship product quickly and iterate with customer feedback

A whole new wave of startups with a new cultural DNA took over.

it’s likely that AI will unlock a new kind of organization in the 2020s.

thanks to AI, they will stay smaller for much longer, and the most successful of them will achieve staggering scale with only a handful of employees.

More importantly, this new type of organization will be able to try out new kinds of ideas.

When experiments get cheaper, a wider variety of people can run a wider variety of experiments.

Conway’s Law will help us understand what effect that might have on organizational structure.

what are people using AI for today? (For many of the below cases, the results are handy but mediocre on their own. Maybe that's ok if most of these are mediocre, if they find One Thing to do well? Most Successful Companies Get Just One Thing Right)

• Building products and features...with fewer engineers there’s less need for complex org structures.
• Working with data.
• Learning new technologies
• Testing your software
• Designing interfaces.
• Communicating with users.
• Interpreting user feedback.
• Automating sales and customer support

Humans will set the initial vision then act as the glue that links everything together and keeps it running smoothly.