The incredible artificial intelligence we were promised is finally here

For decades, we have been hearing predictions of a takeover of the world by artificial intelligence. In 1957 Herbert A. Simon predicted that within 10 years a digital computer would be the world chess champion. This didn’t happen until 1996. And despite Marvin Minsky’s prediction in 1970 that “in three to eight years we will have a machine with the general intelligence of an average human being”, we still consider this like a science fiction feat.

The pioneers of artificial intelligence surely did not have the time, but they were not mistaken; The AI ​​is coming. It will be in our televisions and in our cars; it will be our friend and personal assistant; he will take the role of our doctor. There has been more progress in AI in the past three years than in the previous three decades.

Even tech leaders like Apple have been caught off guard by the rapid development of machine learning, the technology that powers AI. At its recent Worldwide Developers Conference, Apple opened up its AI systems so independent developers could help it create technologies that rival what Google and Amazon have already built. Apple is far behind.

The AI ​​of the past used brute force computing to analyze data and present it in a way that felt human. The programmer provided intelligence in the form of decision trees and algorithms. Imagine trying to build a machine that can play tic-tac-toe. You would give him specific rules about what to do, and he would follow them. This is essentially how IBM’s Big Blue computer beat chess grandmaster Garry Kasparov in 1997, using a supercomputer to calculate every possible move faster than it could.

Today’s AI uses machine learning in which you give it examples from previous games and let it learn from those examples. The computer learns what to learn and how to learn and makes its own decisions. Additionally, new AIs model the human mind itself using techniques similar to our learning processes. Previously, millions of lines of computer code could be needed to perform tasks such as handwriting recognition. Now this can be done in hundreds of lines. What is needed is a large number of examples for the computer to learn on its own.

New programming techniques use neural networks that are modeled after the human brain, in which information is processed in layers and the connections between these layers are strengthened based on what is learned. This is called deep learning because of the increasing number of layers of information processed by increasingly faster computers. These allow computers to recognize images, voice and text and do things similar to humans.

Google searches used a technique called Page Rank to get their results. Using rigid proprietary algorithms, they analyzed web page text and links to determine what was most relevant and important. Google is replacing this technique in searches and most of its other products with algorithms based on deep learning, the same technologies it used to defeat a human player at the game Go. observers themselves were confused as to why their computer made the moves it did.

In the fields where it is trained, AI now exceeds the capabilities of humans.

AI has applications in all areas in which data is processed and the decisions required. Wired founding editor Kevin Kelly likened AI to electricity: cheap, reliable, industrial-grade digital intelligence that runs behind everything. He said it “will vivify inert objects, just as electricity did over a century ago. Whatever we once electrified, we will now ‘know’. This new utility AI will allow us to will also increase individually as people (deepen our memory, speed up our recognition) and collectively as a species IQ In fact, the business plans of the next 10,000 startups are easy to predict: take X and add the IA. It’s a big deal, and now it’s here.”

AI will soon be everywhere. Companies are infusing AI into their products and helping them analyze the vast amounts of data they collect. Google, Amazon and Apple are working on voice assistants for our homes that manage our lights, order our food and schedule our meetings. Robotic assistants such as Rosie from “The Jetsons” and R2-D2 from Star Wars are about a decade away.

Do we need to worry about “artificial general intelligence” spiraling out of control and taking over the world? Yes, but maybe not before 15 or 20 years. There are justified fears that instead of being told what to learn and complementing our abilities, AIs will start learning all there is to learn and know far more than us. Although some people, like futurist Ray Kurzweil, see us using AI to augment our abilities and grow together, others, like Elon Musk and Stephen Hawking, worry that AI is usurping us. We really don’t know where this will all go.

What is certain is that the AI ​​is there and makes incredible things possible.

© 2016 The Washington Post