Artificial intelligence has a black box problem.
Meaning, the programmers that built it, don’t know why AI makes one decision over another.
Technology can do some amazing things, we’d say. Until Donald Trump got elected as the president of the United States of America, and we entered four years of delusional thinking in our echo chambers. That’s quite criminal.
Of course, it was artificial intelligence, machine learning and weaponized propaganda which influenced the election -not the Russians. But if smart learning systems can swing democratic elections, then they might even well be capable of turning our juridical system upside down.
Can AI commit a crime? Can it decide what is right or wrong? Can it make conscious decisions? If so, who is going to accuse or defend them in front of court?
Artificial intelligence has a "black box" problem. Meaning that the programmers that built it, don’t know why AI makes one decision over another. The first deadly accidents with self-driving cars have already occurred. If a machine mistakes a person for a road sign in the streets, it could mistake a deadly medicine for a curing one in a hospital...
So, who is responsible? The human that programmed the self-learning machine, or the machine itself? Can companies working with AI classify 'AI-crimes' as equipment failures?
Why do we lie, cheat, steal?
Artificial intelligence forces us to think about what it means to be human. Why do we lie, cheat, steal, hurt, manipulate, assault and kill? Why do we commit crimes?
To receive some answers, we invited anti-futurist Theo Priestley, who is -despite what his name suggests- not anti-technology, or anti-innovation. The opposite is true. What Theo Priestley does, as more than a full-time job, is discussing the importance of a futurist mindset, in order to stay grounded as human beings. We have to challenge conventional thinking and the hypes around AI.
Take the Netflix-series Black Mirror. It is not drawing a very beautiful picture of what human driven artificial intelligence could look like. AI scientists probably never had the ambition to build killer robots, yet here they are. Already, the United States, China, and Russia are investing in autonomous weapons using AI in drones, battle vehicles, and fighting robots. Yes, some nervosity is surely at its place.
On a personal level, we are scared to death that AI will take our jobs, and any given purpose in life. As long as we’re talking about general artificial intelligence, that’s a valid concern, according to Priestley. But general AI will never surpass humans. It’s smart but unable to empathize and therefore dangerous.
Superintelligence though, will take things to the next level.
Can super AI create a better world?
If you ask Priestley, it is a very strange idea to think that human beings can create something that will be better than what we know. ‘I believe that just simply can’t happen, primarily because it will be impossible for us not to program our own unconscious bias into a machine. Our own prejudice, our own immorality.’
Priestley: ‘I believe a superintelligence will actually be evolved from AI itself. It will understand, it will evolve, and it will learn and reprogram itself. It will surpass humans in every single capability they have. They will be more creative than us. They will have better social skills than us, believe it or not. By investing in superintelligence, we challenge our humanity to actually grow, to peer into the quantum realm, to challenge our beliefs and become more conscious humans.’
‘If God used his own image to create man, what image will artificial intelligence take when it creates itself?’
Please enlighten us, Priestley! Maybe one day, we’ll give up the idea of crime altogether. How’s that for delusional thinking?
Check out the other keynote speakers we invited to bow their head over the question ‘What if AI commits a crime?’.
Access 100+ Conference Talks
Acces 100+ Conference Talks and 20+ Summit talks
At and& festival, we are building an interdisciplinary space where sharp minds merge, in the quest for answers to the big problems of our time. Because all good answers start with good questions, we’ve come up with 21 questions for the 21st century. Confirmed speakers for this question: Theo Priestley, Nell Watson and Gry Hasselbalch.