AI aid Man and Machine, a Team

It is not about whether computers can think, but how can they help people thinking, writes the head of Microsoft Germany.
With a little help from my friends, including robots.

Exactly 21 years ago today, then-reigning World Chess Champion Garry Kasparov lost his first match against the super computer Deep Blue in Philadelphia.

What the IT world celebrated as an epic victory of artificial intelligence was very frustrating for many chess players.

Dutch Grand Master Jan Hein Donner responded to the question how he would prepare for a game against a computer by saying, “I would bring a hammer.”

Mr. Kasparov chose a more creative way. He organized freestyle tournaments where teams of humans and computers worked together, showing that a good team composed of humans and machines is unbeatable.

To this day, many people are fascinated with the question: Can machines think? The answer primarily depends on how we interpret the term “think.”

For my part, I agree with Alan Turing, the father of all thinking machines. What he said in 1950 is still valid today. “We should not ask‚ ‘Can machines think?,’ but rather‚ ‘What can machines do?’” I would like to supplement his basic approach and would ask, “What can we do better together with machines?”

I'm sure when we learn to work well together with machines, we will be unbeatable.

Humans and machines bring completely different talents to the table. Humans are creative and innovative, emotional and empathetic.  Computers are unsurpassable in their lightning ability to recognize patterns in huge amounts of data. Now it's time to bolster these talents in man and machine – I'm sure when we learn to work well together with machines, we will be unbeatable.

Together, we may solve some of mankind’s greatest problems at last. Artificial intelligence will help us better control environmental risks, make better use of our resources and also drastically reduce the number of people dying in traffic accidents. But it will also substantially improve the lives of individual people. For example, the blind Microsoft developer, Saqib Shaikh, has just now developed smart glasses with a built-in camera that not only read texts out loud to him but can also describe exactly what's going on around him at any particular moment. The glasses can even recognize people and gauge their emotions – for the visually handicapped, this is a huge gain in quality of life that until now was barely conceivable.

In addition, artificial intelligence (AI) will help us in diagnosing illnesses and help fight their causes. Already today, for example, amazing progress is being made in cancer research, wherever experts are able to systematically evaluate large volumes of health data with the aid of AI. Establishing the proper conditions for analyzing such particularly sensitive data is right now a concrete challenge for all regulating authorities.

Currently, many people are mostly concerned about whether robots will take away our jobs in the future. It often isn’t even a question of “if” any longer but rather simply of the numbers. Is it 40 or even 60 percent of all jobs that will be replaced by intelligent machines in the next couple of years? In addition, job market and economic activity researchers, production experts and political commentators like to discuss who the first victims of digitalization will be, factory workers or office personnel. Is it this that time the middle class will be hit or will the poorly-paid, less qualified jobs be the first to be eliminated. Will it be the highly-developed industrial nations or the newly industrialized that will benefit most from the next technological boost brought on by artificial intelligence?

Our children need at least a basic knowledge of computer programming so that they won’t experience the digital future merely as consumers of technology but rather be able to actively help shape it.

Over the course of the last 200 years, machines have for the most part relieved us of monotone, badly-paid, physically-demanding jobs that were often even detrimental to health. And, ultimately, technological progress has consistently created more new jobs than destroyed. This applies as well to digitalization. Already in 1999, two out of three employees in Germany were regularly working with computers and since then the number of jobs subject to social security contributions has clearly risen.

But, naturally, in the course of history there have also always been victims of progress. For example, the introduction of the mechanical loom robbed thousands of weavers of their livelihood, which was already precarious. To void repeating such upheavals, we must set the right course as soon as possible, for the pace of technical progress is faster than ever before.

We also need to think about how we can equally share the profits of productivity, generated thanks to intelligent machines, by massively investing in training and improving the qualifications of all workers, for example.

So that man and machine really do become an unbeatable team – and to prevent people from losing their jobs – we must continue to promote not only artificial intelligence but also human intelligence. We should focus strongly on human strengths in schools, like creativity and communication, social interaction and problem-solving skills, instead of simply on imparting knowledge. We need to make as many people as possible capable of working together in an intelligent way with intelligent machines – and for that reason, start investing massively in digital education right away and at all levels. Most European countries already have included coding in their schools’ curriculum or plan to do so in the near future. We should follow suit as soon as possible. Our children need at least a basic knowledge of computer programming so that they won’t experience the digital future merely as consumers of technology but rather be able to actively help shape it.

But we must also develop artificial intelligence from the outset so it supplements humans rather than replaces them, and here I see us, the industry, having that responsibility. At this year’s World Economic Forum in Davos, Microsoft CEO Satya Nadella said, “Augmentation not replacement must be the goal” that must be promoted as an essential design principle of every form of artificial intelligence. Microsoft is moreover committed to the goal of “democratizing” artificial intelligence by making the appropriate tools available to as many developers and users as possible to further develop artificial intelligence according to their own specifications and needs. Moreover, Joi Ito, the head of MIT Media Lab, quite rightly stressed how vital it is to integrate users as early as possible, promote interdisciplinary cooperation, and assemble diversely-mixed teams so that our future won’t be written alone by a homogeneous group of white, male software engineers.

Transparency is another essential principle to which we are committed, along with other leading IT companies in the Partnership on Artificial Intelligence. With this initiative, we seek to raise awareness, promote scientific exchange, and drive forward the discussion of ethical issues.

Only by being as transparent as possible will we be able to reduce the fears and encourage people to participate in the shaping of the society of the future. There is still a lot of room for this. For although we have made enormous progress in recent years, particularly in fields such as data analysis, and speech and image recognition, the development of “real” artificial intelligence has only just begun. So we still have every opportunity to agree on ethical standards and set the right parameters for future development. For that, we need neither a hammer nor artificial intelligence, but above all a healthy dose of common sense.

 

To contact the author: [email protected]