The future of AI: 'Computers shouldn't think like people'

Subject: Contemplating the future role artificial intelligence and its relationship humans

Image Credit: FlashMovie via Shutterstock

Image Credit: FlashMovie via Shutterstock

In the search for intelligent machines, approaching or even transcending human intelligence has been a prominent dot on the horizon since the 1950s. In addition to the various technological challenges, according to Rudy van Belkom, project leader of the future exploration Artificial Intelligence (AI) at the Future Vision of Technology Foundation, this search is extremely difficult for three reasons:

  • We don't know exactly how intelligence works with people.

  • We do not have a generally accepted operational definition for many relevant concepts (such as awareness), which makes it difficult to prove its existence.

  • We are constantly shifting in what we define as intelligence

Van Belkom drew these and many more conclusions as a result of a literature study, various expert sessions and extensive research on Artificial Intelligence

Human-level Artificial Intelligence

"Intelligence is difficult to" clone ". Experts therefore disagree when we are going to achieve human-level Artificial Intelligence. The dot on the horizon shifts with the times and seems to be just as far away. Yet it is not inconceivable that "the intelligence code" will be cracked. "

AI: A mirror for humanity

What if tasks that now appear to be very complex can also be solved with a relatively simple algorithm? Van Belkom states that AI is already capable of creating artworks and composing musical pieces. Many people struggle to accept this as real creativity. On this subject there is also a philosophical discussion that plays a role as to whether we ourselves are not simply programmed and therefore do not act as autonomously as we think.

Image Credit: Ryzhi via Shutterstock

Image Credit: Ryzhi via Shutterstock

Artificial Intelligence can thus become the next "insult" for humanity. Machines previously surpassed us in physical labor and computing power and now our intellectual capacity is also at stake. Our intellectual capacity has always distinguished us from all other beings on earth and has given us, at least emotionally, control over our environment. AI a mirror for humanity. It teaches us a lot about ourselves and forces us to asks ourselves fundamental questions about what humanity is.

Taking steps forward

Do we really need to pursue human intelligence? Van Belkom postulates that submarines do not swim like fish and airplanes do not fly like birds and then asks the question: why should computers think like humans? He then states that If you give a spider human intelligence, it will not behave like a human, but as a "super spider" that can spin even better webs to catch prey.

Van Belkom thinks we will only take real steps if we let go of the idea that we are superior beings. To further clarify his point he explains that humans are not superior to insects, we both evolved for other purposes. Although humans have more advanced cognitive skills, insects are more likely survive a nuclear disaster. "Success" is therefore context-dependent and therefore relative.

A better world…

Van Benkom declares that we have to ask ourselves for what purpose we want to use intelligent machines, instead of seeing the creation of intelligent machines as a goal in itself. How can we create a better world with the help of intelligent machines? And what exactly is a better world? Man and machine should work together from their own specializations. Leave complex statistics to computers and socially sensitive issues to people. Van Benkom asks Why should we build emotions into machines? He would rather strive to make computers as objective as possible. This could be more useful as people, with all their evolutionary programmed prejudices and emotions, are not good at being objective at all.

Source ANP press release - Stichting Toekomstbeeld der techniek (STT)


If you enjoy our selection of content please consider following Universal-Sci on social media: