Humans and Bots — from Jeremy Josse

Technology and human collaboration — a viable business future proven by a simple chess puzzle

• A unique chess puzzle developed by the brilliant physicist Roger Penrose and his grand master brother provides insights into the difference between human cognition and computer analytics.
• It may also be a guide to a model of how computers and humans can work in the business world together collaboratively — rather than the supposed inevitable replacement of humans in the work force by sheer computer analytic might.
• Those companies that are combining computer and human collaboration are already showing they are prospering in the new tech age

Sir Roger Penrose and his brilliant grand master chess brother have come up with a fascinating chess position that can be solved relatively easily by humans, but is usually analyzed incorrectly by the most sophisticated chess computers today (those types of computers who normally can win every single game against world chess champions like Magnus Carlson). I do not want to get bogged down in the chess detail (to save those for whom chess is not a major interest), but want to look at the business and technology implication of this story. But, to give the context, let’s first just briefly describe the basics of the chess problem.

A chess game is in the position as shown below. It is a very unique position indeed, although quite legal.

It is white to move. Stockfish 8 (a powerful chess computer), that would invariably beat the world champion Magnus Carlson, analyses the above position as a strong win for Black. In fact however, it is relatively easy for a moderate level chess playing human, through some basic intuition, to find ways to make this a draw and even, under certain circumstances, a win for white. The element that confuses the computer, in brief, is the fact that you have the extraordinary situation of three black bishops on the same color square. That is probably a position very, very rarely seen, but is legal and technically possible (again we don’t need here to go into how you get to that position).

In any event this, it seems, somewhat confuses the computer, for the computer in its algorithmic fashion finds it now has to compute hundreds upon thousands of new positions relating to all these curious bishops. The human however is able to make the intuitive leap that these bishops (always being stuck on black squares) can be used effectively to force a draw.

We do not need to go into more of the chess analysis here, but is a hopeful story as it indicates humans have a real role even in an advanced computer age. Writers like MacAfee and Brynjolfsson or Carl Frey will tell you that much of humanity is heading for unemployment as computers and robots replace us in most roles in the manufacturing, production and even service processes. But if this chess example is anything to go by, it shows that what humans need to do is learn to live side-by-side, symbiotically, with computers to utilize their mutual skills.

The game of chess itself was thought to be on the edge of death once the computer program Deep Blue compressively beat Kasparov in 1997, but now serious chess players have learnt to work with computers to enhance their play while also bringing added insight into the game. And this is no doubt because, as Penrose believes, the human mind is simply not a binary processing system, but works using some other method of cognition (may be quantum type processing?)

In any event, it is becoming clearer that those tech businesses that are finding ways of combining human intuition with raw computing algorithmic power will be those who really prosper in our new tech age. As Roger Wu of Computerize.com recently said the key is to: “use the power of algorithms meaningfully — i.e. by using them as a first pass to reduce the thousands of data points to a few hundred and then use human input to add the contextual piece of the puzzle.” He points out that tech start-ups fail regularly because they miss this point — e.g Mailbox “that tried to solve email only to fall short because of the subtle personal differences from user to user behavior and context that machines are not yet geared to pick up on.”

Even in factories there is and needs to be changes. As the NY Times recently reported robots were historically caged off from humans on the assembly line to prevent the machines’ powerful steel arms suddenly hitting a human. “But now, gentler industrial robots, designed to work and play well with others, are coming out from behind their protective fences to work shoulder-to-shoulder with people.”

This may show the limitation on purely algorithmic systems such as Siri, Google voice recognition and Xbox One’s voice. Instead the investor should focus on, not just a technology-enabled platform, but a synthesized digital-driven and human-enabled model. Accenture’s annual outlook has made a similar point. Companies such as Nissan or Airbus have already developed friendly “co-bots” which work collaboratively with humans. And the chart below is an estimate in co-bot growth over the next few years:

So there it is. The Penrose chess puzzle (relatively easily solvable by humans, but very tough even for the greatest computer algorithms) is a sign not only of things to come, but solutions to how technology can enable the human work force (rather than destroy it) and enhance tech driven growth.

This article was originally published by Maxos fellow Jeremy Josse on SeekingAlpha.