Does AI matter?
The head of the Word Economic Forum, Klaus Schwab, and some intellectuals somehow linked to his organisation, such as Yuval Harari or Ray Kurzweil, believe that we are indeed about to witness the coming of the Singularity, a state of affairs in which man-built machines will first become intelligent and conscious and then surpass human intelligence. Many horrific narratives are tied into this idea, such as the concept of the AI-overlords propagated by Elon Musk or warnings that AI will rule and enslave mankind given by Stephen Hawking. We, my co-author Barry Smith and I believe that none of this is true, but based on a lack of understanding of the mathematical foundations of AI. Hawking was a good mathematician; unlike his teacher Penrose he probably did not look at AI closely enough to get a realistic picture of it. Anyhow, in our book “Why machines will never rule the world. AI without fear” we show that there will be no singularity, a summary of our arguments is given in this interview with Digital Trends. The WEF has now announced our book in their AI news section, which is a good start to move away from their unscientific fear narratives.
Our view can be summarised like this: AI is based on mathematical models. But consciousness and intelligence are properties of the human mind, a complex system, and we cannot fully model complex systems mathematically, at least not to the extent required to emulate consciousness or intelligence. In the book, we provide a lot of arguments for this, and show that there is no reason to be afraid of AI.
Now does AI matter? Yes – but not in the way Schwab sees it.
AI is part of the ongoing usage of mathematics and science to cover our world with a broader and denser network of technology, the technosphere. But it is just a normal part of applied mathematics that can be used to plan and build algorithms and machines that are excellent to fulfill certain well defined contextual purposes. Outside these narrow contexts, AI systems fail. Nevertheless, like all sophisticated technologies, AI will contribute to our technosphere via automation and improved pattern identification methods. But it does not fundamentally change the industrial world we already have. At Cognotekt, we have deep experience with the limits and possibilities of AI in many industries and a broad range of use cases.
Contact us to explore these possibilities and to avoid mis-investments into pseudo-AI.
Register here for our biopharma and AI series.
Vor welcher Herausforderung
stehen Sie? Schreiben Sie uns.
Dr. Raija Kramer