AMZN is basically organized as a conglomerate like GE.
...
The company is growing by paying aggressively for acquisitions. What could go wrong?
This is exactly the opposite of what Amazon is doing to grow. It is building internal systems at immense scale and then productizing them and selling the additional capacity to others.
AMS began because Amazon had to build enough computer capacity to run flawlessly at peak load. On Black Friday, their site has to run as well if not BETTER than it does on normal days. That is an INSANE amount of computer horsepower that is by comparison largely idle 90% of the year. Hence AMS -- an experiment in selling off surplus that turned out to be a profit monster.
FBA is the same thing. It was the third attempt to grow Amazon, after two earlier failures. One was Amazon Auctions, which failed fast. Another I think was called Amazon Stores...didn't really work. FBA worked better than they could have imagined.
Bezos is super smart, but he's willing to experiment. Fire Phone -- nutty idea that was a disaster. Alexa -- nutty idea that for some reason worked. They throw stuff against the wall, some things work.
The healthcare partnership with Buffett & JP Morgan is the same thing. It will be an internal system for all three companies. They will try to debug it, see if they can make it work better than the way things are done now. It very may well fail. But if it works, it may change healthcare in the U.S.
They may buy certain strategic companies -- like the Ring video doorbell -- but that's not a core pillar of the company like AMS; it's a quick technology leapfrog where they're already all-in (accessing people's homes for secure/trusted deliveries.)