AI in innovation: is AI the answer?
AI may have a large impact on economy by serving as an invention of a method of inventing, that can reshape the nature of innovation process, reduce costs of specific innovation activities and organisation of research and development.
Lately we’ve been experiencing investments in researching climbing sharply, while payoffs are staying constant. Economists from Stanford and MIT found out that it takes more researchers and more money than ever to find productive ideas. AI could be the answer to innovation stagnation, since most new innovations stem from complex and large amount of data.
For example: Did you know, that it takes average of 15 to 20 years to come up with new material? AI can speed the process up to 10 times. Time saved could reflect in far more potential solutions being tested and it would help optimize the materials. Deep learning has great potential in chemistry, biotechnology, drug discovery and other fields that face classification and prediction, given AI’s ability to dramatically lower costs and improve performance in research and developing projects. Breakthroughs in fields have become harder and harder, since they’ve become complex and saturated with data.
Beside innovations, AI will also have economic, social and technological consequences. Important determinant of economic growth from the development and application of deep learning is policy, that should enhance innovation in a way, that encourage competition, but also data sharing and openness. New jobs, better organisation and productivity will follow from economic growth. Developed policies should also enhance innovation in a way, that promotes competition and results in social welfare. AI’s main impact on civilisation might be totally different that we thought. Not flying or autonomous cars, emotion and speech recognition or humanoid robots, but AI’s ability to come up with new inventions to fuel innovation itself.
AI also isn’t biased like we humans are and there could be plenty paths to discoveries, that might never occur to a human researcher. Can you even imagine all the possibilities?
Important questions arise
with AI and IP.
Who is the inventor, if AI comes up with new innovation?
Who is liable if AI’s decision causes damage or harm?
As of now, the existing law of any country wouldn’t identify AI as creator or author of IP. AI produces results as outcome of an algorithm or its own intelligence. A distinction between deep-learning and algorithms must be made, since only-mechanical decisions might be lacking invention. But even if countries would grant copyrights to the works of AI, there is no answer who would get the copyright. Current legislation requires legal personhood of a right holder, which AI lacks. 
Present challenges can be solved through clear agreement between parties, where it should be noted who owns data or results generated by AI, how it will be used, will the innovation be kept safe as trade secret or registered as a patent and also who is liable or responsible for the decision-making or results that are automated by AI. Caused damage could be prevented by ensuring, that humans maintain control and have the ability to override any decision made by AI. IP laws are changed from time to time, because of IP’s dynamic nature and new creations. 
Stephen Hawking said, that whereas the short-term impact of AI depends on who controls it, the long-term impact depends on whether it can be controlled at all. The fact is, that AI also proposes a threat to companies, since it cannot be taught morality. What if AI transmits trade secrets to competitors, evolves a malicious program or otherwise compromises IP? It’s necessary that legislation changes, and answers burning questions regarding AI and IP, including possible violations committed by AI. 
Stephen Hawking also stated, the autonomy of AI can diminish the worth of human thinking and invention. Separating IP made from humans and from IP made from AI might be the right way to go, without reducing the role of the human race itself. Human element is nonetheless essential in managing the rights and obligations associated with patents, which cannot be done solely with a machine.
Since AI operates with big
data and all innovation comes from “ just” analysing inputs, one may ask
himself, is AI really creative?