On October 10th, we had the luck to have a visit from Jerry Kaplan, computer scientist and serial entrepreneur and currently Fellow at the Stanford Center for Legal Informatics. Augusto Coppola, Director of LUISS ENLABS Acceleration Program, asked him his point of view about a few trending topics regarding the development of Artificial Intelligence and we would like to share with you what came out this interesting chat. First of all, Augusto asked Prof. Kaplan whether, in the near future, we shall expect startups to leverage on AI technologies provided by large companies or rather to see them developing their own technologies and what would he recommend. “Always buy rather than build”, said Prof. Kaplan. According to him, big companies provide reliable AI technologies backed by people who have experience in the field and allow people without AI expertise to create startups that utilize AI. Buying an AI system is cheaper and faster than developing your own, therefore, for startups, it is better to buy AI technologies developed by large companies (Amazon, IBM, Google…) and concentrating on their implementation to perfect their own businesses and tools, rather than creating new AI systems on their own. From Prof. Kaplan's standpoint, buying an AI system from a major corporation should be viewed the same way as a business buying Microsoft Word: using the tools available from these companies, only accelerates a startup’s process to getting a product to market and does not make the startup dependent on that corporation, it’s simply a tool used in the development of a product. In facts, getting on the same “download curve” of major corporations who have poured millions in resources into developing AI technologies saves a tremendous amount of time and money for startups. Kaplan uses the example of Amazon Web Services, describing how it started out as internal database software for Amazon, which they eventually monetized and sold to other companies to use for their own needs. "Do not think of yourself in the technology business, rather think of yourself in the business of finding markets where a technology can best be applied", said Prof. Kaplan warning developers who fail to see that the launch of an innovative startup is not just as a matter of software development. The reason why there is so much potential in the field of AI is exactly that the technology has a multitude of uses, but not many people understand it: those who have the advantage to see this need to figure out where this technology can best be applied. "The big risk is on the business/market side rather than the technology side of AI", went on Prof. Kaplan. We conclusively got to talk about autonomous driving and when it will be available to the public, as Augusto pointed out that there are different views in regard. Prof. Kaplan's argument is that we are still 10 years away from seeing autonomous cars everywhere and different views depend on the variety of landscapes where we can imagine to see self-driving cars: and self-driving vehicles will initially be implemented through rural areas where there are not a lot of other cars on the road, rather than in cities, and the first application of autonomous driving will come from long-haul truck driving, where it is more obvious to see the money-saving that companies can derive as there is no time wasted from driver sleeping. Further, the problem, according to Prof. Kaplan is not in the technology itself, but rather how we can program these cars to interact in a socially acceptable manner among other human drivers: autonomous cars operate differently from human drivers and technology must be adapted (Prof. Kaplan says “dumbed-down”) in order to co-exist with human drivers. Prof. Kaplan described how autonomous cars can travel at high speeds with only a few feet between cars, which eliminates traffic. However, this benefit is not applicable when there are other human drivers on the road. Conclusively, Prof. Kaplan said we can expect AI developers (Amazon, Google, IBM…) to partner with car manufacturers (Ford, Chrysler, Toyota…) in order to make autonomous cars, rather than technology companies pouring resources into developing a car, or car manufacturers pouring resources into developing AI.