I think it is a mistake for most individual news/sports sites to not participate in a subscription program offering limited use for occasional readers. It prevents the site from collecting at least some revenue from those readers and possibly converting them to full subscription readers later.

8 years post-acquisition is a phenomenal run. Looking forward to what Kevin and Mike pursue next.

At Machinima, we were one of the earliest content partners on Amazon Channels.  On the plus side, they were a relatively easy partner to work with and getting up and running on the service wasn't difficult.  The down side is that we got no data, so the value of those subscribers to us overall was only the value of the revenue stream.  That being said, we would never have done 50/50 deal.  There isn't any marketing support to justify that kind of rev share on a subscription product.

I don’t believe individual local “hosts” can provide the infrastructure necessary to run complex, high demand products at scale while maintaining quality, and for that reason I believe we’ll see Airbnb embracing reputable suppliers in the experiences space in a more traditional manner some time in 2019. 

Given the unpredictable quality control of Uber's driver pool, as a food business owner I would struggle to sign up for a service where I would receive an unvetted set of people with no connection to my values, culture, or service SOPs and have this be the ongoing value proposition as they scale out a large constantly churning pool of gig workers.

In the service business, service is the product, not a sideline (as it may be in their core transportation business).

I would suggest that there are areas closer to their core competency where they may want to focus.

As regrettable and avoidable as this incident was, until Waymo's Level 4 self-driving system is perfected, it's still a Level 3 vehicle where the driver is the weakest link. All this real-world testing has inherent risks. Safety measures and more drivers in each car can only reduce but not eliminate the probability of accidents. 

Honda spends $6B annually in R&D. This agreement w/ Cruise would suggest they are doing $167M/year into this project for the next 12 years, or 2.7% per year of their annual R&D spend. 

Going it alone with $167M/year to work on [autonomous vehicles] isn't smart capital allocation. Putting that money to work as a part of a syndicate...makes at least more sense than doing it alone.