Do We Still Need Humans?

In my office, we use a lot of what seems like "futuristic" artificial intelligence approaches to understanding financial markets. Most of my team are technical or data-science specialists that develop and drive the systems that create our systems. 

Pretty soon, we may start to question where using humans is still smart or cost-effective. 

In the meantime, I'm fascinated by what is becoming possible ... and how, even when the A.I. is little more than an elegant use of brute force, incredible results are becoming commonplace.

MIT has been looking at work by many different organizations - including Google Brain - who are working on AI that can develop machine learning software. 

In many cases, the results coming from machines coding other machines are matching or exceeding work done by humans. 

 As a point of reference, virtually all of our training techniques were developed or created by an A.I. process.

AI creating AI has many benefits.

But, at this point in time, it still requires lots of human input and insight. 

I suspect that won't be true for long.

  When that is true, what jobs will humans be doing?

 I have some thoughts, but I'd love to hear yours.