2029 is the year of the singularity when computer intelligence will surpass human intelligence. This is a prediction from the futurist Ray Kurzweil at Google, who has a track record of being right in 86% of his predictions.
“I’m not afraid of my calculator”
There’s a lot of fear concerning AI surpassing human intelligence. My calculator is way better than me at calculating but I don’t need to be afraid of it. My personal view, says Ray Kurzweil, is that it will take 1 000 years for AI to surpass human intelligence.
We have made a few breakthroughs with AI, but nothing near what a human can do. We shall not diminish the accomplishments so far but we are not even scratching the surface when it comes to reaching human intelligence.
Mark my words – AI is more dangerous than nukes
– Elon Musk – Founder & CEO of Tesla and SpaceX
Human Intelligence vs Artificial Intelligence
Today, we still have a lot to learn about the brain, about our neurons and how we perceive things. But we don’t need to understand Human Intelligence to understand AI. Replicating Human Intelligence is not the right way for AI.
An example is Alpha Go, that was taught by a human to play games. Then Alpha Go Zero was created, and learned how to play without human knowledge. The latter learned and played trillions of games and won 100 games to 0 against humans. It was no longer constrained by human knowledge.
Hence, understanding how humans do it is not always the best way – AI will find other ways. Human Intelligence is lots of little things: its humour, tuition, judgment, emotional intelligence or not making an “optimal” decision for a problem, because it could serve other purposes for making that particular decision. One definition of intelligence is to find different options and evaluate them.
The next step for AI is to solve problems that are easy for people and really, really, hard for a machine.
It’s very cool that we now can take images and know what’s in them. It’s not just recognizing the 10 000 things in a picture but the ability to synthesize what’s in the picture and phrasing it.
But humans have a strong competitive edge in relation to AI and that is our ability to make judgments of small data sets. You can teach a child what a dog is by pointing at it and saying “dog” a couple of times. The child can then translate this knowledge to other dogs – even if they look quite different from the first dog.
AI needs large volumes of data to understand the same thing. Learn from and draw similar conclusions from small data sets is a key piece missing in AI today.
Co-creation with AI
We are at the starting point of building data-sets to teach machines the principles we want them to have. However, the question is how we can evolve ethics in the data-sets and avoid the biases found in society today.
For example when you search for “CEO” on Google, nine out of ten images show white males. How can we teach AI without todays biases? We maybe will merge with the machines – maybe it’s not an US vs THEM. Maybe we can create things that we wish humans had – and learn how to become smarter together.
‘Machine rights’ and ethics for AI?
Should AI have machine rights, just as we have human rights? Does AI need to be protected against human violation? Maybe the most important thing is to create machine ethics and learn how to model it.
The greatest fear is not AI doing bad things but AI causing humans to do so. For example a scenario where you search for “Hispanic” on the internet and receive ads that are horribly racist in its content.
So how do we create ethics and which principles are correct? The real threat with AI is that we can create things that are not real – e.g. politicians that say things they have never said. A machine will maybe never be able to distinguish the fake version.
AI – learns from us – like our own children
Even before AI – human behaviour has not been perfect. Ray Kurzweil believes that we are moving to a more peaceful and democratic world.
AI is learning from humans, it reads millions of documents written by humans. Reading about the world, they can conclude that surgeons are male, because that’s what they read. We’re to correct this but it’s hard because at the same time, we do not want to infuse our own thoughts and values.
It’s the same thing as human children, they learn from other humans. It is important how we educate and teach values to children and the same thing will prevail to AI.
We don’t want to filter out human ideas. It’s an ongoing effort to avoid obvious problems. The world does have a pessimistic view, on the question if poverty in the world is better or worse, 90% said incorrectly that it has become worse. People believe that, even though it’s getting dramatically better.
My CEO is an AI robot
We will definitely see robot-controlled businesses. A robot will be able to perform much of the business activities and these kinds of structures will evolve. Many people will be employed by a machine.
Drugs, food and surgery will all change due to AI
10 years from now we will be able to programme ourselves away from cancer
– Ray Kurzweil – Director of Engineering at Google
- Education will be an important area that will be substantially improved by AI. We need to revamp the education system so that we are prepared for the new jobs that will arise and AI will enable a completely new way of gathering insights and translating those to students.
- The area of biology and medicine is very interesting. We are now able to use large number of data sets to understand diseases and we can develop drugs that are perfectly matched to a specific human.
We are now learning to grow food in an optimal way. We have grown food for 10 000 years but we still know only a little about the optimal way of growing food. When we use large data sets we can discover completely new things.
With image recognition, we have discovered that 50% of surgery for breast cancer is not necessary. We will be able to distinguish necessary surgery from unnecessary, in a much more specific way.
Download the report here!
Enter your email below to subscribe and get notified of new events and news from Cartina and to get access to the full SXSW18 report.
SXSW is one of the biggest digital conferences in the world, and a global meeting place for the world’s most innovative technology companies and people interested in how disruption can transform their business and everyday lives. The event takes place during during 10 days each year and this year Cartina had the chance to be part of it.
This series consists of 8 global mega trends that business leaders, experts, innovators and disruptors talked about during the days in Austin. If you want to read the full report, click the button above and we will email it to you.
Visst borde fler läsa detta? glöm inte att dela!