- In reskilling employees for the AI age, we can choose to try to be better than robots or to complement them.
- We should be aiming for a middle-ground, getting the best from both parties’ potential.
- The future of machine learning should be about how humans and machines can form the best teams.
As the US goes through the biggest loss of jobs in decades, President Donald Trump is proposing to solve matters by decoupling the US’ manufacturing relationship with China and bringing those jobs back to America.
However, there is a significant challenge to that strategy. By and large, manufacturing jobs as we know them are not going to return. Instead, they are set to be replaced by automation and machine learning.
This is not a uniquely American problem. Global population forecasts say we will reach almost 8.5 billion by 2030. Add the exponential speed at which areas such as AI, computational processing power and robotics are developing, and it is safe to predict that our global workforce and the demands we put on it will change markedly in the near future.
Job functions will change rapidly to mirror the pace of technology, so creating a workforce that is educated and ready to adjust to changing demands must be among our priorities.
The form future jobs will take is likely to be shaped by how man and machine end up working together. It remains unclear to what extent the analytical power of machines will replace that of humans. Will the human presence in some job functions become entirely obsolete? Such questions could be better framed.
There’s good news for those in creative roles: so far, machines cannot really replicate the imagination. The bad news is for those who have routine, non-creative jobs, as these are indeed being eaten up by automation.
Man can meet machine in two ways
There are two opposing approaches to how we can help the workforce keep up with technological development. One is to boost employees’ analytical skills to compete directly with the machines, the other is to strive to complement machines and artificial intelligence with synthetic skills.
But it is not about polar extremes, nor is it a question of picking one or the other; it is about finding that sweet-spot of how machines and humans work best together. This will not come about by designing the fastest central processing unit (CPU) nor the strongest robot. Instead it will be the fruit of designing the best teams, best processes, and best user experiences.
We should not be sizing up the potential of humans nor machines in isolation, but taking both combined. Designers must look into solutions whereby humans and machines complement each other, maximizing the potential of both.
Why computational power should not be reserved for only the specialists
Garry Kasparov, a Russian chess-master who still holds the record for consecutive professional tournament victories said, “A weak human player plus a machine plus a better process is superior to a very powerful machine alone, but more remarkably, is superior to a strong human player plus machine and an inferior process.”
In the 1990s, Kasparov represented humans versus machines in a historic chess game against IBM’s Deep Blue computer. He went on to observe and participate in various chess contests where teams of man and machines competed against each other.
His conclusion? It is not the team with the most computational power or the highest-ranking grand masters that will win, but the team with the best interplay – the best teamwork.
In 2000, grandmaster Vladimir Kramnik defeated Garry Kasparov and became the Classical World Chess Champion. After retirement, he sought to rekindle human virtuosity in chess. Paradoxically he did so with the help of DeepMind – the makers of the best chess computer so far, AlphaZero, a far more advanced chess computer than Deep Blue. It is self-taught, and since AlphaZero can teach itself to play, it is also able to learn how to play any game by new rules.
It can explore new variants of games and reveal its bugs and beauty more quickly than generations of human play could ever do. It can test all the outcomes of a game and decide if the game is worth playing. Consequently, the human-machine team Kramnik-AlphaZero are exploring new forms of chess that bring about human mastery and aesthetics, and they have come up with all sorts of new and alluring types of chess as a result.
Reimagining the business process
We should not expect the future of machine learning and robotic design to be about humans versus machines but rather how humans and machines can form the best teams. A survey of a thousand companies working with AI published in Harvard Business Review stated in 2018 that, “Most activities at the human-machine interface require people to do new and different things (such as train a chatbot) and to do things differently (use that chatbot to provide better customer service). So far, however, only a small number of the companies we surveyed have begun to reimagine their business processes to optimize collaborative intelligence.”
Today, at least 90,000 of IBM’s 388,000 employees are applying design-thinking methods to develop the company’s business domains - such as AI and CPU. As such, IBM is iterating and experimenting with how they can improve the user’s experience of working with computational power.
The future will not be about creating the fastest CPU or cultivating prototypical employee skills, but it will be about designing the most compatible combinations of humans and machines, and optimising and simplifying the interaction between the two. And the most pioneering companies already know it.