This opinion article was published in daily newspaper Trouw on 6 May 2022 (link).
Judging by the worldwide reactions, it has not escaped anyone’s notice that tech billionaire Elon Musk recently bought Twitter. As a result, Musk now has unlikely influence over social debate and the rules surrounding it. He himself says he will use the newly acquired power to defend freedom of speech, but for one man to have that power at all is highly undesirable.
But Musk’s power over our future is not limited to Twitter. In the next century, we are going to drive electric, in a Tesla owned by Elon, for example. We will go into space on a grander scale, in Musk’s rockets. And finally, this could well be the century of AGI: artificial general intelligence or general artificial intelligence. And there, too, Musk has his finger firmly in the pie.
Playing chess and driving a car
Ordinary artificial intelligence (AI) is often better than us at one task, such as playing chess, playing Go, or driving a car, but not at multiple tasks. General artificial intelligence (AGI) would be able to handle a broad spectrum of tasks and could even perform every imaginable task better than ourselves. The technology does not exist yet, but most AI scientists believe it will happen this century.
Because AGI is inherently better at cognitive tasks than we are, it could replace many more jobs. It could also start controlling physical things, such as drones, robots, weapons, or factories. The first steps in that direction have already been taken with Deepmind’s MuZero, who could play 57 different computer games without the rules having to be explained to him.
What are “our” values?
This presents potentially huge opportunities, but also risks. Will we soon be able to control technology that is smarter than ourselves? Currently, no reliable method is known for this, despite some 20 years of research into this.
Tech billionaires like Elon Musk therefore sponsor institutes that conduct research into how we can soon make AGI act according to our values. Fine, but who do we mean by ‘our’ when it comes to ‘our values’? If AGI can become more powerful, because smarter, than us, who will soon be in control? Perhaps Elon Musk and his fellow tech billionaires. Their influence on shaping the twenty-first century is disproportionately large and therefore risky.
That developing technologies – apart from AGI, biotech and nanotech – are going to make a big mark on the future is certain. That these influential developments do not belong in the hands of a small club of billionaires, too. Precisely in this politically hardly explored area, governments should therefore work like hell to regain influence. With limited resources, the Netherlands can become a leader in research into the control and democratisation of future technology. May sound vague, but it is crucial.
Appoint top existential risk official
By launching our own investigation, we are building a democratic counterweight to the power of the tech billionaires. If our government does so, the Netherlands can claim a leading role globally, as we are not the only country in the world that is now lagging behind. Therefore, we recommend to install a director-general for existential risk at the ministry of home affairs. Add tech-driven existential risks to the Integrated National Security Risk Assessment. And establish a Planning Office for Existential Risks, which can start conducting its own research.
Outrage in response to the considerable power of men like Musk, solves nothing. If we do not want ‘tech cowboys’ to have so much influence over us – over our communications and over our future – it is time to invest in a democratic counter-power ourselves.