Lost In Time And Space

it is better to have application programmers deal with performance problems due to overuse of transactions as bottlenecks arise, rather than always coding around the lack of transactions

It is true that eventual consistency is hard to program against. So relying on transactions can give you a lot less complexity. As soon as you run into consistency problems…

If you offer people database columns, they will create a column with a type for each and every property. If you offer relations, you will get desired relations and every other relation, too. If you give them acid, chances are that we get huge transactions.

As usual you have to pick the right tool for the job. And then you have to have the right people to do the job.

Ken says: The goal of education is not command and control but climate control. Education is about humans and not mechanics. And there are three conditions under which humanity flourishes: diversity, curiosity and creativity. But what education often doing instead is the contrary: standardizing and demanding compliance (through standardized testing)

TL;DR

Innovation is a search:

in a start-up, innovation is not just about implementing a creative idea, but rather the search for a way to turn some aspect of that idea into something that customers are willing to pay for

But big companies want predictability:

discovering a new business model is inherently risky, and is far more likely to fail than to succeed. Blank explains that this is why companies need a portfolio of new business start-ups rather than putting all of their eggs into a limited number of baskets. But with little tolerance for risk, established firms want their new ventures to produce revenue in a predictable way — which only increases the possibility of failure

And add their high-potential managers, with a high potential in big companies, not in innovation:

internal entrepreneurs are more likely to be rebels who chafe at standard ways of doing things, don’t like to follow the rules, continually question authority, and have a high tolerance for failure. Yet instead of appointing these people to create new ventures, big companies often select high-potential managers who meet their standard competencies and are good at execution (and are easier to manage)