START Summit 2019: Artificial intelligence: the human being must be centre stage

Artificial intelligence (AI) is potentially capable of changing every sphere of life. For this reason, in particular, it is important for people to be part of this revolution. Public discussion must ensure that everyone who is involved will be able to profit from it while being able to check the risks at the same time. This is the gist of a short address by Marc Holitscher, who participated in the START Summit as a speaker.

22 March 2019. "AI – hope isn’t a strategy" was the title of the brief address by Marc Holitscher, who as the National Technology Officer is a member of the Executive Board of Microsoft Switzerland. “Artificial intelligence is an all-purpose technology which will pervade all aspects of human life. Therefore it’s extremely important to clarify proactively how the preparation and utilisation of AI technologies should be regulated,” he emphasised.

Changes result in stress
Basic technologies had the potential to reconfigure the economy and boost productivity in all sectors and industries. This did not only require technical innovations but often also a comprehensive redesign of the infrastructure, of business models and cultural standards. Such transformations were well-known to us from the past. Marc Holitscher quoted electric power and steam power as examples.

Such novelties would result in a certain amount of stress in society because the advantages of innovations with transformative power were frequently delayed and unclear. "Up to the point where people can see and experience a very specific benefit, feelings of loss, a certain fear and aversion to risk may dominate."

Complementing human skills
Marc Holitscher regards it as extremely urgent that society should be aware of the tension between the opportunities and risks of AI. “There mustn’t be any doubt that the human being is always centre stage.” Technology must be configured in such a way that it would complement and extend human imagination and human skills.

Artificial intelligence was capable of making automated forecasts on the basis of big data volumes in conjunction with a certain probability of occurrence. What it was not capable of doing, however, was making decisions. "In terms of assessment, human beings possess natural advantages which a machine won’t be able to replace in the foreseeable future. Machines are incapable of assessing the consequences of a certain recommendation for action in relation to context and in an integrative way," for this would require empathy, creativity, emotional intelligence and humour. "This can only be done by human beings, which makes them unique."

Clearly defining values and rules
If the use of artificial intelligence should take its bearings from the human being, this would require clearly defined values and rules. "Researchers, political decision-makers and executives in government, business and civil society are being called upon to develop reliable framework conditions for the use and control of AI," emphasised Marc Holitscher. These conditions would have to be based on ethical principles that are non-negotiable. As cases in point he mentioned fairness, transparency, explainability, inclusiveness, but also security and the protection of people’s private sphere. Effective mechanisms for the concrete implementation of these principles in everyday life would also be required. "For example, this could be about companies pledging not to conclude a business deal if there are serious reservations about the use of AI."

Marc Holitscher argued in favour of being optimistic and vigilant at the same time when it came to using this technology. "The costs of machine learning are decreasing rapidly; at the same time, the accuracy of the results is constantly increasing. This boosts the economic incentive to use artificial intelligence in any conceivable scenario. In the future, many everyday products will include AI components and embed them, as it were."

Cloud democratises access to AI
In the business environment, the decreasing price of AI has a similar effect. The fact that artificial intelligence helps to rearrange the internal flow of information and to enable the creation of new networks improves efficiency and the power of innovation. This nexus was applicable to big and small firms. The sinking prices of AI solutions and the highly scalable preparation through cloud-based infrastructures also make them accessible to small and micro-enterprises. "The cloud democratises access to AI options. It’s easy to understand that everyone wants to profit from these opportunities: SMEs, doctors, stockbrokers and nation states," explained Marc Holitscher.