Tech, Tech ethics, Shift 2018

I went to tech prison in Finland – here’s what I learned

The business festival Shift took place in the middle of May in the old prison of Kakolan Lääni. The festival revolves around tech and ethics, having a human-centered focus. Here, some of the world’s expert discuss our future. Ultimately they are optimists.

In the words of Steve Jobs, tech alone is not enough. The more digitized our societies get, the more we also have to start looking at the consequences and the long-term effects on humanity, not just the short-term benefits of tech.

Last week, I went to Shift in Turku, Finland. The oldest city in the country now home to one of the more forward-looking and quickly growing tech festivals. Here, humanity is in the center of every tech talk.

Here’s what I learned:

AI will soon be able to plan on a human level

The 3rd generation of AI is coming. Harri Valpola of the Curious AI Company has developed an AI which cannot only collaborate and communicate with people but also plan. Until now, computers have only been able to do what they learn from data. The 3rd generation will be using techniques like those in the human brain that makes us creative and able to handle situations we have never seen before.

To make sure, these new types of AI grow up in a way we humans are comfortable with; we have to understand the AI and the AI need to understand us, Harri Valpola told me. The best way to make sure that AI does not want to harm us is to let it live among humans and be part of our society. That way it will learn our values and be as diverse as the human race. Hopefully, then we can avoid scenarios like the rebelling of robots in Westworld!

Check out: Sweating it up in Finland: Why the sauna is a perfect mix of business and pleasure

Harri Valpola delivering his talk at Shift 2018.

Three significant obstacles to reach AI success

Besides the technical challenges, we need to fix three significant obstacles to make AI succeed in the way we vision. So said Nick Boström, philosopher, and professor at Oxford University where he runs the Future of Humanity Institute.

First, we need to address AI safety. If you think about what it would mean to have a system that was capable of doing anything you wished for you have to be careful about what you demand of it. Think the myth of King Midas. Everything you touch turns to gold might sound like a swell idea at first – until you die of hunger because of the lack of food.

The example illustrates a few different things; we need to be able to specify what we really want without causing unwanted side effects. Which is hard because human values are complex and context dependent. Also, thinking of instrumental convergence, we need scalable control systems; as the AI gets smarter, we need a control system that works even better as the AI system improves.

Tech, Nick Boström, AI
Nick Boström. Photo: Shift

Human level machine intelligence will be with us in a few decades

According to a study done by the Future of Humanity Institute, we’ll probably see AI with a human level of intelligence around 2040 or 2045. That was the median answer of the world’s most prominent experts in the field. But some think that it’ll never happen just like some say it will be just in a few years.

We should already begin to think hard about the consequences of algorithms. Not only for when computers will work alongside us communicating like humans. Algorithms already control big parts of our world. Like whether you can go to college or not, if you qualify for a job, if you’re the person to get kicked off an over-booked airplane, or how long your prison sentence will be should you ever go to prison.

Cathy O’Neil, algorithm activist and founder of ORCAA, want us to ask more questions about how algorithms work. When the developer says yes, does it mean it is efficient, or that it is also legal, fair, and has no bias? Probably not. We are embedding our values into the algorithms without thinking about it. Though not all algorithms are bad, there are bad algorithms, and they are not only unfair to individuals, they also undermine democracy.

Tech, Cathy O'Neil, AI, Algorithms
Cathy O’Neil. Photo: Shift

Also read: Six reasons not to go to Slush

Big is better

Urbanisation is a growing trend, so much that the number of people living in cities all over the world increases every week by an amount equivalent to the entire population of Finland. Cities grow because of costs in energy and other resources per unit decrease as size increases.

Professor at the Santa Fe Institute Geoffrey West has done the research and found a simple pattern of growth across all life forms. All life forms, big and small, conforms to the same simple laws. As the size of an organism doubles, the energy demands of supporting it increase by 75 percent. This also applies to man-made organisms like companies or cities.

Mega cities, future, tech
Geoffrey West. Photo: Shift

Human and tech will merge

The technological evolution is way faster than the biological. To keep up, humans need to merge more with tech. At least, Hannes Sjöblad, Chief Disruption Officer at Epicenter, Stockholm, thinks so. He a biohacker, and Epicenter is famous for its chip parties, where employees can get microchips implanted. It can be used to replace keycards and IDs. In the future, it’ll even be possible to use it instead of a credit card, key, transportation card, or bitcoin wallet. And it doesn’t even hurt to get one.

Read more: Are you a freelancer because you have to or because you want to?

Share this: