"Can machines think?" - this was the question posed as early as 1950 by Alan Turing, a leading mathematician of the 20th century and the driving force behind the team that had cracked the Enigma code ten years earlier and thus made an important contribution to ending World War 2. Turing's work is considered the birth of the computer age and of "artificial intelligence". But what does "artificial intelligence" actually mean? Alan Turing proposed a test for this: if a human and a machine "talk" to each other and the human cannot tell at the end whether he was talking to a human or a machine, then the machine is "intelligent". In recognition of his work, the Loebner Prize has annually rewarded software that withstands a so-called strong Turing test for at least 25 minutes since 1991. Chatboots like "Mitsuku," however, have so far only been able to reach the bronze category.
Machine Learning: Learning Playfully
Machine learning took an important step when people started letting computers play against themselves. In 1997, world chess champion Garry Kasparov lost to the Deep Blue computer, in 2011 IBM Watson won the U.S. television show Jeopardy, and in the world's oldest brain game, "Go," the machine has now also outperformed humans. However, the algorithms of such games could not be transferred to things that are otherwise very easy for humans: like walking, reading handwriting, interpreting pictures or understanding languages and dialects. With the development of artificial neural networks and sufficient computing power, this changed.
Neural networks are not programmed, but learn themselves. Using a variety of data and the associated solutions, neural networks train themselves and can achieve better performance than humans through continuous improvement. Such systems are behind Siri and Alexa, translation programs such as DeepL, face and person recognition programs, or medical applications, for example for analyzing X-rays and MRIs.
Today, commercial applications that use artificial intelligence have reached all sectors of business and research. Artificial intelligence is indispensable for autonomous driving , innovative energy transition solutions , the materials development , the medical engineering and nursing , and even for innovation management methods.
Innovation management methods based on AI
According to a forecast by consulting firm PwC, Germany's gross domestic product could increase by 430 billion euros by 2030 thanks to AI. The German government therefore wants to invest three billion euros in the implementation of its AI strategy by 2025 and create 100 new professorships. Naturally, Bayern Innovativ is also closely following the trendy topic and regularly presents new results on its platforms and in publications. And with services such as the Experten Netzwerk Bayern , Bayern Innovativ has long had its own innovation management methods based on AI in its portfolio.
