john m
john m
Aug 25, 2017 · 1 min read

Some thoughts from outside the valley;

The fundamental problem with all these machine learning systems is the analysis of past data and modeling it to predict future behavior. It’s a lot more complex than computing power. Just because something has the same pattern doesn’t necessarily mean it’s going to lead to the same outcome each time. Identical twins are genetically identical yet their looks and behavior diverge over the course of a lifetime.

For example, I don’t know how many times I purchased a product and then for the next three months receive on-line ads for that same product. Maybe that generates money for Google but it’s terribly inefficient for advertisers. If that is the state of the art at the moment then Google is way overpriced.

Trump didn’t win because of computer processing. The Clinton team was much more dependent on computer processing, modeling and polling than Trump. Trump did a traditional sales pitch directly to the American people in rally after rally throughout the country. Suit and Tie included.

Humans have evolved social networks thousands of years before Facebook. We are a lot more complex then we realize.

Wealth inequality is a problem but it is also easy to fix by raising taxes and regulating monopolies. I suspect you are going to see both soon.

)

    john m

    Written by

    john m