Famous for essays decades old, more recently famous for his “Let’s Build” essay and generally famous for one of the most successful VC companies in the world, Andreessen gives an interview with Sriram Krishnan.
Out of many thoughtful points I’d highlight two.
-Personal productivity and the unstructured model vs the structured model.
-Process (inputs) vs Outcome (OKRs/metrics) models
On productivity
Andreessen moves from an unstructured model - that he blogged on previously (see below) - where he argues for trying to never put anything in the calendar…. To a very structured model that he articulates in the interview and highlights the pro/cons to both. It’s a 180 change in model - worthy of a hedge fund manager reversing his trade.
A highlight of the structured model (or his one but others are similar) is that everything goes on the calendar - and you have to calendar “free” and “open/thinking time” and protect it.
Also worth looking at is Paul Graham on these areas. His 2006 blog still relevant on how insiders/outsiders and productivity can work.
As an aside, it’s why I remain suspicious of any claims to one “correct” way of thinking about personal productivity. Marc is extremely successful and has used model that are 180 degrees different.
On outcomes vs process
There are broadly two schools of thought.
One which emphasises metrics... Only items that can be measured can be managed… and using KPIs (key performance indicators) and OKRs (Objectives and Key Results).
The other emphasis processes and inputs as important and outcomes come out of that.
Poker (and by extension investing or enterprise that has a decent mix of chance/random/luck as well as skill, and further it’s hard to know what exact mix causes an outcome) is suited to input and process.
Sure you can measure your win/loss but it only improves via the process.
VC investing has a decent dose of luck and long horizons, say 10 years is common, which makes it suitable to the process method and Marc mentions this and his emphasis on process.
Endeavours which have very low random chance elements may be more suitable to OKR management - and say the lean type manufacturing processes (although lean itself has a lot of focus on process, it also has a big part on measurement).
As a final code... On let’s build... he would build in the areas of healthcare, education and housing.
Interview here: https://www.theobservereffect.org/marc.html
On Let’s Build: https://a16z.com/2020/04/18/its-time-to-build/
Previously on productivity: https://pmarchive.com/guide_to_personal_productivity.html
Paul Graham on being marginal, inappropriate and outsider: http://www.paulgraham.com/marginal.html
Annie Duke on poker / process (also see Michael Maboussin). https://www.annieduke.com/