Social-as-a-service, Roomba and opportunity costs

Sajith Pai's 'sort of weekly' newsletter #2020-04


As more and more products go digital, with the social aspects getting unbundled out of the offerings, we will begin to see separate social offerings to help customers add these social elements back.

We replaced movie theatres with Netfix but we want Netflixparty; we replaced offices w WFH but now need offsites and mixers to bond.

The more digital we go, the more social we will need to sprinkle back. My piece ‘Social-as-a-service’ is linked here.


I did listen to a few (Invest Like the Best with Hamilton Helmer, Venture Stories with John Danner & Michael Staton, and North Star Podcast with Matt Cooper, come to mind) but havent been able to sit and write out the highlights. That I listen to these while dishwashing is perhaps the reason:)


What I enjoyed reading last week

(1) Review of the book ‘Debt: The First 5000 Years’ authored by David Graeber. Review by Alex Danco..

Details out the central idea from the book, that money or coinage arose not so much to facilitate easier commerce as much as to enable a contribution / IOU that was not to be paid back (religious or military tribute). Economies which predated the invention of coinage worked well, Graeber says, even without money, because the economies were local and people could enforce trade / barter easily via IOUs given how well they knew each other. Coinage came (in the Axial Age of 800-200BCE) when wars and invasions led to conquerors demanding tribute to be paid back, and money or coinage was a convenient way for these non local IOUs to be transacted. I have probably simplified this a bit much, and really, you should read the essay.

The other interesting point is about how important China’s decision to back to precious metal currency in the 1400s was. It resulted in a precious metals shortage in Europe -  the Great Bullion Famine per this link though there is no mention of China - which led to the exploration and discovery of Central America with its abundant gold and silver reserves, and of course all that death and destruction that came with it.

(2) This piece, by Joe Jones, inventor of the Roomba is a great read. Lots and really lots here to unpack including 

  • frugal (hardware) engineering: remember the Roomba (original name was DustPuppy!) was conceived and developed when sensors were prohibitively expensive. They used a technique-called behaviour-based programming - based on two behaviour modes, Cruise and Escape, to build the Roomba out frugally.

  • robot business economics i.e., how to design or conceive robots that succeed - the principles that Joe Jones details include (a) perform a valuable task (b) do the task today using established tech (c) do the task for less, are all general purpose principles that guide any hardware or even software product development.

and, finally

  • why robots will not develop general AI: not enough compute cycles beyond what it is designed to do, he says. There is no redundancy.

(3) Fun illuminating piece on how the blogger and his friend who runs a restaurant game Doordash’s discount offer to generate free money for themseves(!), and segues from there to explore the poor economics of the delivery business, concluding mostly correctly that it is due to perverse incentives engendered by access to cheap capital from VCs.

(4) I thought this post by Erik Bernhardsson, CTO of, was really good. It counters Hanlon’s Razor, i.e., the principle that we should never attribute to malice what can be adequately explained by stupidity, as an explainer to why an item X or Y doesnt get done in a co.

He suggests that a better way to think about this to say: Never attribute to malice, stupidity or laziness what can be explained by opportunity cost. Essentially that stuff doesnt get done because the opportunity cost of working on X or Y is high - the alternative output Z that is getting generated because they are working on that instead of X or Y is in fact more valuable to the co. 

More simply, stuff doesn’t get done because of lack of time, or because something else is important, not because of stupidity, laziness or malice.

Corollary: prioritisation is the most value creating activity in any company. Erik says it is also a good idea to have a large backlog of to dos, so long as the prioritisation framework is well-understood.

Finally, in startups, opportunity cost is a decent explainer, but in big cos, it is usually a strategic reason. (below image from the piece).

(5) A look at the many Singapores that exist, through the eyes of a Grab deliveryman. Not the typical grab delivery man but a small business owner whose earnings disappeared after COVID and who then took up the Grab gig. I suppose it wouldn’t be too different if a similar piece on Mumbai or Bangalore dropped.