It has been a while since people began to discuss about the threats that artificial intelligence might bring to them. No matter in movies like Terminator or warnings from Elon Musk, CEO of Tesla Motors Inc., AI is often regarded as a threat to mankind. However this is not an ideal result for AI. Neil Jacobstein, an AI expert who has once worked as an advisor for US Army, General Motors, Ford, etc recently talked about the threats of AI and robot. He said it’s worth noting that a lot of areas are now applying artificial intelligence technology, they acquire knowledge from it and use it to assist people’s lives. However we still need to remain vigilant on AI. Luckily, we have now figured out ways to protect mankind.
Threat 1: Replacing human jobs
Jacobstein said:” People are evaluating the concrete influence that AI will make to the labor market and now it’s unknown whether AI will subvert the labor market.” But according to studies published by the World Economic Forum, jobs of 5.1 million people will be replaced by AI or robots in 15 leading countries in the next 5 years.
Solution: Guarantee minimum income support program
As is said by Jacobstein, it’s rather critical to “positively” get ready for the inevitable unemployment, especially in developing countries. He said approaches like guaranteeing minimum income support program, increasing educational opportunities and retraining could help to reduce effects of unemployment. Some countries have been trying to launch fundamental income system in order to deal with the future unemployment.
Threat 2: Abuse of power
According to Jacobstein, the progress made in AI technology field should not only be enjoyed by technology giants such as Apple and Google, or else there might be abuse of power, the other countries and companies will be left far behind.
Solution: Increase transparency
Jacobstein believes it is important to ensure the transparency of AI research and development. He said:” AI research should apply to all the enterprises in the world and most of the advanced countries so as to make sure the others won’t be left behind. Suitable competitive mechanism has not yet been formed by now and the others are angry about this.” This explains why Jacobstein is in favor of Open AI- the nonprofit organization which hopes to make sure AI researches are open source and be applied by all.
Threat 3: AI might be too powerful
As a branch of AI technology, deep learning mainly teaches machines to learn to finish tasks by their own. The potential of deep learning is rather powerful, it has been applied into self-driving car industry and more advanced robot area. Nevertheless, once AI begins to learn on its own, many people worry that a robot like “Terminator” might appear. Jacobstein said” We would have problem controlling it while AI gets too powerful.”
Solution: Shutting down the control system of AI
As Jacobstein said, it is necessary to conduct more researches to make sure the software conforms to rules, this will avoid unnecessary influence. He said:” If there is a mistake and most of the systems make mistakes occasionally, how could we rapidly recover control of the systems? This often requires establishing control in different ways.”
Threat 4: Evolution of man-machine coordination
Jacobstein put forward that we are approaching a key point which is the coordination of brain and body. Musk has also raised similar point, he said we should begin to build “neuro lace” and allow it to add “digital wisdom layer” in our brains. As we marching towards the direction, AI will be a part of our body. Jacobstein believes it means a lot to mankind to think about the ethics influence of this technology.
Solution: Discussing about ethical issues before developing the technology
Jacobstein said:” We want to make sure we have contemplated on how would we build partnership between mankind and machines, how to maintain human-machine relationship and how to improve human dignity while building machine intelligence.”