Intermediate

Algorithms can learn to collude. 

Two law professors, Ariel Ezrachi of Oxford and Maurice E. Stucke of the University of Tennessee, have a working paper on how when computers get involved in pricing for goods and services (say, at Amazon or Uber), the potential for collusion is even greater than when humans are making the prices. 

Computers can't have a back-room conversation to fix prices, but they can predict the way that other computers are going to behave. And with that information, they can effectively cooperate with each other in advancing their own profit-maximizing interests.

Sometimes, a computer is just a tool used to help humans collude, which theoretically can be prosecuted. But sometimes, the authors find, the computer learns to collude on its own. Can a machine be prosecuted?

In April 2017, economists at LSE looked at 51 studies of pay-for-performance plans, inside of companies. Here's what they said: "We find that financial incentives can result in a negative impact on overall performance."

There is a mismatch between what science knows and what business does. And what worries me, as we stand here in the rubble of the economic collapse, is that too many organizations are making their decisions, their policies about talent and people, based on assumptions that are outdated, unexamined,and rooted more in folklore than in science. And if we really want to get out of this economic mess, if we really want high performance on those definitional tasks of the 21st century, the solution is not to do more of the wrong things, to entice people with a sweeter carrot, or threaten them with a sharper stick. We need a whole new approach.