“You can’t skip the learning curve”

Norderstedt, February 27, 2019 – As the digital transformation gathers pace, methods based on artificial intelligence and machine learning are increasingly finding their way into the business world. Today, AI systems are capable of handling complex tasks that, just a few years ago, only humans would have been able to perform. Nevertheless, AI is still subject to limitations. After all, an AI is only as good as the data used to train it. What opportunities and risks does AI actually harbor – and how can companies make the most of such technology?

“We already have plenty to drive us forward,” was one conclusion of the ‘Artificial Intelligence – From Promise to Practice’ round-table discussion at the Hamburg IT-Strategietage. “We have to start experimenting now. You can’t skip the learning curve.”

Those present at the event agreed that AI errors will continue to decrease while the technology’s potential uses grow. However, this will also need to be accompanied by increased levels of transparency, because humans will eventually be unable to monitor AI systems entirely. How exactly an AI makes its decisions remains an unsolved mathematical problem. “We’re going to have to learn to trust artificial intelligence,” commented one participant.

This issue of whether other countries have progressed further than Germany in this field was another lively source of discussion. In this regard, those present painted a rather bleak picture: while they agreed that Germany has excellent researchers and developers, the prevailing view was that the country falls behind in terms of investment. Canada was highlighted as one example, having invested in artificial intelligence since 2017. The federal administration is providing 125 million Canadian dollars over a five-year period, with a further 250 million dollars coming from the Canadian provinces and other public funding sources. By contrast, the German government’s 2019 budget sets aside just 142 million Euro to support artificial intelligence projects.

Yet, in light of the risks involved, private enterprise in Germany is also reticent to invest heavily in AI – in part due to the risk of wasting huge amounts of money. “The issue is that companies can’t simply set aside a few days and a handful of employees to handle artificial intelligence, a new key technology,” said the event’s host, Heiko Packwitz of Lufthansa Industry Solutions. “The important thing now is to allocate resources, capacity, and investment and to start working on tangible business cases.”