How will researchers create artificial general intelligence

Artificial intelligence

Ulrich Eberl

To person

is a doctor of physics, science and technology journalist and book author. He is managing director of the editorial office SciPress, was head of innovation communication at Siemens until 2015 and founded the future magazine "Pictures of the Future". Most recently published: "Smart machines - how artificial intelligence changes our lives" (Hanser, 2016). [email protected]: //www.zukunft2050.wordpress.com «

250 years ago, James Watt applied for a patent for his improvement to the steam engine. Since then, resourceful researchers have been constantly building new machines to relieve people of tedious work. The steam engine revolutionized not only mining, shipping and railways, but also steel rolling mills and the textile industry with spinning and weaving machines. A hundred years later the triumph of electric and combustion engines began. Cars, planes and assembly lines made mass mobility and mass production possible. And again a hundred years later, microchips, computers, cell phones and the Internet led to automation, digitization and limitless communication.

But never in this long history of industrial revolutions has the core of being human been within reach of machines: our intelligence. This is exactly what is changing. In the past five years there has been more progress in the field of Artificial Intelligence (AI) than in the previous 50 years. Machines are already outperforming us in a number of fields that previously seemed reserved for humans. They beat human world champions in the quiz game Jeopardy as well as on the go board, they can find house numbers on millions of Google Street View images in less than two hours, they make only half as many mistakes as humans when recognizing traffic signs , they find evidence of cancer cells in tissue sections that were previously unknown to doctors, and they can read basic emotions such as anger, joy, sadness and surprise from faces better than many of us are able to. [1]

Smart machines learn to speak, see, and read

Above all, the speech, text and image understanding of machines with artificial intelligence is amazing: Siri, Alexa, Cortana, Google Assistant, and whatever the virtual assistants are called, are currently learning at breakneck speed to understand questions and commands from people and make sense to answer. Translation programs such as Google Translate or DeepL can translate long sections of text into other languages ​​in a matter of seconds - not without errors, but in a quality that would have been unimaginable a year or two ago. And the Watson system from IBM can analyze natural-language texts and filter out and summarize content - whether it comes from Wikipedia or from medical specialist literature, from stock market and company news or from reports from auto repair shops.

In clinics, banks and companies, the first of these new AI systems are already processing data and giving doctors, financial advisors and managers recommendations for diagnoses, investments or the optimization of industrial processes. The smartphone was obviously just the beginning of the era of smart machines. In addition to smart health, smart finance and smart factory, the buzzwords of the future are smart grids, smart cars, smart buildings and smart cities. [2] Intelligent power grids, the smart grids, will bring energy supply and demand into harmony and thus make sustainable energy systems with all their renewable energies, storage systems and consumers possible in the first place. Very soon we will be switching our vehicles - the smart cars - on autopilot on motorways, and also on country roads and in cities by around 2030. In factories we will work hand in hand with robots, in hotels, museums and shops machines will provide us with information and serve us. At home and in retirement homes, they will bring drinks, play games with old people, clean, carry away the laundry, maybe even cook. Buildings and cities will be full of sensors that measure energy consumption and space occupancy, traffic flows and emissions, and help computers optimize.

There is no doubt that we will need the smart machines: the autonomous vehicles and the intelligent home as aids for the aging population, the smart grids and the internet of energy for sustainable energy systems, the smart factory for a flexible and competitive industry and the smart Cities for the liveable cities of tomorrow. Soon we will be living in a community of people and smart machines - as natural as we use smartphones today.