Professor Tells UN, Governments Of Coming “Tsunami” Of Data And Artificial Intelligence 21/02/2018 by William New, Intellectual Property Watch 1 Comment Share this:Click to share on Twitter (Opens in new window)Click to share on LinkedIn (Opens in new window)Click to share on Facebook (Opens in new window)Click to email this to a friend (Opens in new window)Click to print (Opens in new window)NEW YORK — Technology is moving so fast it could cause harm to humans even as it brings remarkable advances unless governments act, an Israeli professor and visionary thinker told a gathering of government and United Nations representatives here last week. A hint? In the next five years we are all going to be cyborgs. In fact most of us already are. Prof. Shmuel (Mooly) Eden of the University of Haifa, Israel, presents. A joint event of the Israeli and Italian missions to the UN, was held on 14 February at the elegant Italian government cultural salon on upper Park Avenue. The Israeli and Italian ambassadors made opening remarks, setting the tone of awe for what technology can do to help us, but also caution over what it might cause. A representative of the UN secretary-general at the event said the secretary-general has said he wants to use technology like AI as an “accelerator” for the Agenda 2030. But, she said, “The secretary-general wants to make sure technology remains a force for good.” Prof. Shmuel (Mooly) Eden of the University of Haifa, Israel, called one of the most innovative minds of our time, opened his talk by saying we would engage in an “exchange of opinion.” That is, he quipped, you will come in with your opinion, and leave with my opinion. Eden (who gave legal disclaimers for his talk at the outset) said a tsunami is small compared to what is coming. He gave statistics on the extraordinary exponential rise in technology and computing power, and listed a few events that showed the effect of technology now, such as a single death in Tunisia leading to the so-called Arab Spring, , the spread and draw of ISIS, a terrorist who posted a selfie in front of what he described as a terrorist headquarters leading an American bomber to race over and destroy the site within the same day, and an airline passenger who was bloodied when being removed from an overbooked flight which led to the head of the airline company stepping down and a precipitous (if temporary) drop in the company’s stock value. Things like these may have always happened, but now they are captured quickly and spread widely with smart phones, leading to drastic consequences and over-reaction. The biggest threat to the UN is not nuclear, said Eden, it’s cyber. Time is getting shorter and shorter to react, with urgency. Governments may not feel as much urgency but will be affected, or replaced. The biggest threat to the UN is not nuclear, it’s cyber. – Prof. Shmuel (Mooly) Eden He said this fourth revolution in human history is made up of four factors. First, computing power is at levels that were unimaginable. This power is what makes artificial intelligence now possible. The smartphone in your hand has 1,000 times the components of the first rocket to the moon, he said, which led to a chorus of “wows” from the audience. Second is big data. Every time you speak on the phone or go on the internet, someone records it, he said. The amount of data is unlimited. Eden said he would be surprised if we use 2 percent of the data we generate, but in the future “we will.” Third is artificial intelligence (AI). No one could analyse all of that data, so AI came into play. Fourth is robots. He noted that they don’t always look like human forms. Most robots are just software doing some function. Eden talked about the rapid improvement in performance of technology, and highlighted the point in time at which a computer was able to beat a human at chess, and the point where IBM’s Watson soundly beat the top human champions at the tv game show Jeopardy. But more concerning, something about which inventor Elon Musk and others have raised alarm, is that robots have gotten to the point where they can do a job better than a human and we may not understand how it did it. The example he gave was facial recognition, distinguishing between men and women, where AI quickly learned to better distinguish than humans can. At first it was not so easy for the AI, but it learned something that humans could not. Even after distinctions in hair, or lips, or other features were removed, the AI could still pick out which ones were women at a high rate of accuracy, where the humans could not. What made it particularly alarming is that humans do not know how the AI is doing it. Same goes for a test involving a typical stop sign. Humans recognize the distinctive eight-sided sign even from the back. But they are less likely to be sure if they only see a portion of the sign or if the light is too bright or other interfering factors come into play. Not so for AI, which can still identify it no matter what, and humans don’t know how it is doing it. The last example was particularly stirring, a test in which a husband and wife are on opposite sides of a wall and can speak to each other. AI learned so well what one of them was saying and how they said it, that the other could not tell if they were talking to their actual spouse or not. Eden noted that oftentimes now we are speaking to a computer when we speak to a company and we may not know. AI can be taught feelings too, like responding to something with anger, and can be even become lovers with humans (perhaps a fitting message for a presentation given on Valentine’s Day evening). All of this leads us to the point of asking: “What makes us human?” he said. All of this leads us to the point of asking: “What makes us human?” AI can even be taught to lie better than a human being, he said, the possible implications of which are very disturbing. AI can even be taught to lie better than a human being And then comes Moore’s Law, which says computing power doubles exponentially every two years. The theory was identified in the 1960s. And by now, 7,000 transistors can fit on a cross-section of a strand of human hair (the audience emitted another “wow”). The human brain has 100 billion neurons, the building blocks which act like transistors. In 2013, AI had 2 billion. Under Moore’s law, AI will surpass the human brain about twelve years from then (2013=2 billion, 4, 8, 16, 32, 64, 128). By 2026, we will be able to build a microprocessor more powerful than the human brain. By 2026, we will be able to build a microprocessor more powerful than the human brain He sees it as inevitable that “We will have computers that are much smarter than you.” We are going to be cyborgs, he said. In fact, “You are already cyborg.” He showed a picture of a crowded subway in Asia where virtually every person was looking at a phone. If your phone is taken away, “you are without oxygen,” he said. He asked the room who believed we will have chips in humans. Most were skeptical. But in fact, it already is the case, he said, for a variety of medical reasons such as an automatic insulin pump for diabetics. But then he talked about a key issue for the UN and governments: the displacement of workers. He showed a line of self-driving cars and asked what the audience sees. What he sees is 300 million drivers out of work. Eden showed a line of self-driving cars and asked what the audience sees. What he sees is 300 million drivers out of work. He also showed an Amazon company shop floor, with no human workers, just drones to get things from shelves and fill orders. And he showed two photos side-by-side of a vast bank stock trading floor that was filled with bustling workers just a few years ago, and now has only a small skeletal staff remaining, as AI does the rest. Lower social jobs will be first to go, which are often dangerous and low skill so the thought is not so abhorrent. But what makes governments sit up and notice, he said, is the realization that with the lost jobs will be lost taxes. The world is changing exponentially, he said. The internet revolution took 10 years. This one will take 5 years. Another concern about what AI will learn came from an experiment called Tay, which was an AI presence on Twitter. The company crowed that the more people tweeted at Tay, the smarter it would get. But it was very surprised and had to make apologies when in fact it started tweeting out viciously racist and misogynistic messages. This led Eden full circle to a quote from renowned scientist Stephen Hawking, who said, “without a world government, technology will destroy us.” Eden ended by quoting a hero of his, former Israeli Prime Minister Shimon Peres, who told him: “Technology without ethics is evil. Ethics without technology is poverty. That’s why we have to combine the two.” Eden challenged the governments, the UN and all others to think about how to address this rapid change and come up with ideas. He challenged the governments, the UN and all others to think about how to address this rapid change and come up with ideas. Exponentially. [Note: for a recent related news article on AI and ethics, see here.] Image Credits: William New Share this:Click to share on Twitter (Opens in new window)Click to share on LinkedIn (Opens in new window)Click to share on Facebook (Opens in new window)Click to email this to a friend (Opens in new window)Click to print (Opens in new window) Related William New may be reached at firstname.lastname@example.org."Professor Tells UN, Governments Of Coming “Tsunami” Of Data And Artificial Intelligence" by Intellectual Property Watch is licensed under a Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License.