Despite low returns on billions in investments, technological intelligence shows potential for a massive increase in productivity and prosperity over the next ten years, according to Georg Gesek in an interview. It is increasingly successfully emulating the evolutionarily developed human brain. However, there is no threat of humans being replaced by machines.

For Georg Gesek, the term “artificial intelligence” is potentially misleading, as if artificially created intelligence were something completely different from that which has evolved. Instead, the Viennese physicist, electrical engineer, and CEO of HPC specialist Novarion Systems emphasizes: “Intelligence is the ability to solve problems.”

This means that it makes no fundamental difference “on which hardware it runs,” i.e., whether humans or machines tackle problems. Gesek therefore prefers to speak of technological rather than artificial intelligence. Furthermore, he believes it is pointless to discuss whether this already deserves the attribute ‘intelligent’: wherever neural networks “find solutions that humans have not found before, that is intelligence.”

Technology has taken the human brain as its model for good reason. Even though initial approaches are still emerging and marketable large language models still lack core capabilities such as continuous learning with flexible use of short- and long-term memory, machines are increasingly “becoming more and more like us in their thinking.” Google is also demonstrating this with its current research into such “nested learning” for AI systems.

In episode #60 of DataCenter Diaries, “Georg Gesek: There is only one intelligence, and AI is the next stage of evolution,” Georg Gesek talks to Ulrike Ostler, editor-in-chief of DataCenter-Insider, about the relationship between natural and technological intelligence and why the latter is “a logical continuation of natural evolution using technological means.”

Podcast episode #60 of DataCenter Diaries is available on Spotify, Apple Podcasts, Deezer, and Amazon Music.

It also addresses why information has no inherent truth value and how AI models will remain dependent on human-defined frames of reference, how Europe can keep pace with the US and China, and why the European Union’s tendering rules need to be fundamentally changed to achieve this.

Gesek does not see any threat to the relevance of human thinking. Certainly, any chess program can now beat a human being. But the machine achieves even greater advantages when it discusses its moves with a human being. “Although humans play worse than computers, humans and machines together are still stronger than a machine without humans.” The Novarion CEO and QMware co-founder adds: “The technological intelligence we are developing for various areas of application will work with us, and we will of course teach it to enjoy working with us.”

Gesek also does not want to overlook the fact that technological intelligence is “very, very immature.” Despite billions in investments, including in the expansion of data center capacities, “the use of AI in enterprises or large organizations is very limited,” and there is “still a significant economic gap.”

Nevertheless, Gesek remains confident. “This is not an [AI] bubble,” and there are good reasons for the investment volumes: “We know what will be possible in the next five to seven years.” Even if the exact technical implementation has not yet been determined, “we know what we need to do to build increasingly powerful and intelligent technological intelligence. Ultimately, it is much better planned than it looks from the outside.”

This means that the AI boom is not comparable to the dot-com bubble at the turn of the millennium. Unlike back then, Gesek says, AI technologies deliver measurable results. Even an exemplary return on investment of just 10 percent still conveys an idea of what is possible. And “we know what we need to implement technically so that we not only achieve the other 90 percent, but go well beyond that.” Linear increases in investment would thus be offset by the potential for exponential increases in AI capabilities.

Gesek is therefore convinced that AI will generate “massive returns on investment in three to five years” and bring about “a massive increase in prosperity worldwide over the next ten years.” He believes that the urgent task for society and politicians is to prepare for this and to ensure that the increase in prosperity is distributed fairly.