These are the biggest dangers of artificial intelligence

Automation is a good thing, says Starmind founder Pascal Kaufmann, but business might not be prepared for automation at mass scale.

These are the biggest dangers of artificial intelligence

Starmind founder Pascal Kaufmann tells TechRepublic's Dan Patterson that giving menial tasks to AI gives humans more room and time to focus on more important things, but that automation on a mass scale takes a lot of preparation - and potentially a big loss of human-executed jobs. The following is an edited transcript of the interview.

Dan Patterson: What are the dangers if we progress down a path where we think of the brain, metaphorically, as a computer we keep plugging more and more data? This is something I talk with, not just with you, but many people in the space, and almost everybody says, Well, we need more data. I'll tell that to my bosses, my colleagues, my friends, we just need more and bigger data. If this is the path we're heading down right now, what are some of the dangers that we might encounter along the way?

Pascal Kaufmann: First of all, exploiting big data, the deep learning algorithms automating processes is a great thing. There's so much stupid work that could be given to the machines. I've actually bought a... with processes could be automated.

On the other hand, the negative respects could be that we are not prepared for that. If we define our lives by what we work, you run into difficulties, because at some point in time, the robots or the machines can take over our work.

We really should stop defining ourselves by means of our work. There are much more interesting things to do, than just doing a job. Why not give away all these tasks to machines and robots, so that we can focus on a more interesting things?

I think society's not yet there, but ultimately it would be great, ambitious, to give the work to robots and machines, so that we can focus on the more interesting parts of life.

Dan Patterson: So, per scarcity economies, is that what you're saying? Eventually, we need to get to a point where all work is automated, and humans can spend their time with more philosophical pursuits?

SEE: Artificial intelligence: Trends, obstacles, and potential wins (Tech Pro Research)

Pascal Kaufmann: Well, if you asked that 200 years ago, what do they do, if they do not work for 18 hours a day? But, one day they just have to work for eight hours-a-day, they would have told you, What will I do with my spare 10 hours?

So, if you now ask this question, what will we do in the world if we just have to work for one hour-a-day. There are many, many things that we could do with this time.

Dan Patterson: I'm not asking what we'll do with that time. I'm asking you is this the inevitable outcome of plugging more data into automated systems?

Pascal Kaufmann: If we do it right, ultimately, there are not so many jobs on this planet that can notbe automated. As soon as something is defined, as soon there are rules behind it, you can actually automate it. This, however, doesn't have to do a lot with intelligence, it's just brute force statistics. If you do it, if you optimize it, you can get rid of many, many jobs, indeed.

Also see