WEB SUMMIT: Artificial Intelligence: Innovation or Risk?
Lisbon-How will Artificial Intelligence change our lives?
For Daniela Braga, founder of Defined.ai, a market market place focused on this market, deleting boring and daily tasks that “we do not feel challenged to do”. Defined.ai works with what is called conversational artificial intelligence, which is a branch of artificial intelligence focused on the creation of technologies capable of treating, understanding and interacting with human language.
According to her, conversational AI is the evolution of IVR (Interactive Voice Response) technology) – basically, this bot which guides the consumer through the buttons during a call. “In IVR, there is no interaction. It is a monologue, not a dialogue,” she said on the third day of the Summit web. “It is a much better experience to be able to say what you want right away and be understood than having to go through the buttons. The conversational AI allows that.
According to her, the next big breakthrough on this market should be the development of the speech – automatic translation (and real -time) translation of any language using artificial intelligence. For her, this type of technology is closer to becoming a reality than most people imagine. “I think that very soon zoom cats will be translated in real time in any language.
Not next year, but maybe in five years. “Today, there are already prototypes of technologies in which the translation is done with the voice of the same person in a language that she has never spoken. And they can really do this translation in a short time. C ‘is an impressive thing. “
On the same panel, Sean Kerrigan, more founder, one startup developing autonomous trucks, said that for him, the keyword to define AI is “optimization”. “This is what really defines AI today: helping to improve and automate things that require a lot of data or labor.”
For him, the market is about to switch to autonomous cars. “If you look at the classic braking cycle, we have already reached the summit where we will have autonomous cars everywhere and they will dominate the streets, and now we are at a low point. I think it’s time when things Start moving, “he said. “Companies with products a real customer value creation that will survive, prospere and develop.”
Noam Chomsky, the father of modern linguistics, and Gary Marcus, a scientist in artificial intelligence at New York University, also discussed AI in another panel – but the conclusions were diametrically opposed. “Current [AI] systems only perpetuate old data and do not understand the world and human language,” said Gary.
“Because they only use past data, they perpetuate the prejudices of the past and produce disinformation. This will harm the world. It will be horrible for democracy.” For Gary, today’s AI systems are nothing more than “these semi-automatic entry systems [which complete the letters of a word] under steroids”.
Another risk, he said, is that people spend billions of dollars in industry-for example, investing in autonomous cars. “I do not think that we will never reach the fifth level of automation that Elon Musk has promised for years. Then, all this money is lost.” I am more concerned with the risks for democracy than investors who lose their money , but that’s another risk that exists.
Chomsky also noted that research on AI is in the field of engineering, not science. “These surveys do not help our understanding of the world and human beings, they therefore make no contribution to science.”
He said that when the original idea of artificial intelligence began to develop in the 1950s, the initial objective was “to use these capacities to see if we understand how humans think, if we can learn some thing on cognitive systems “. Considered obsolete. “The journalist moved to the invitation of Brivia, a communication group specializing in the digital transformation of major brands.