matt's blog

The law of the mechanical turk

Lately, I’ve been thinking about the future of product. In Linchpin, Seth Godin describes The Law of the Mechanical Turk:

Any project, if broken down into sufficiently small, predictable parts, can be accomplished for awfully close to free.

Now, this makes sense— the more open-ended the job, the more it pays.

Seth's example compared Encyclopedia Brittanica to Wikipedia— through project management (open-sourcing contribution), Wikipedia was built for free, or at least much closer to free than a team of 100 full-time editors.

Of course, Seth named the law after Amazon's "Mechanical Turk" service, which (in 2010) could execute tasks like audio transcription for $0.20 per minute through parallelization and assignment to remote workers.

Now, take that in the context of 2024. OpenAI's Whisper can transcribe your audio for $0.006 per minute and scales infinitely (or close enough). That's a 33x improvement.

So, who wins?

The leaders that understand how to leverage the power of automation by most efficiently breaking down their problems and, more importantly, make the smartest bets on PMF.

Now the important part:

Product isn't changing , this has always been the case.

However, order-of-magnitude changes in efficiency mean the gap between the 1% and the 99% will get much larger. Winners will be the ones that:

  1. Have the best ideas
  2. Use the new rules to iterate on their ideas the fastest.

#1 is, of course, developer experience. So “winning” product looks like creativity + project management + developer experience.

#opinion