Artificial intelligence is developing so fast it can be hard for humans to keep up. The technology is becoming commonplace in everything from customer service chatbots to managing hedge funds, researchers are getting computers to reason more like people, and the European Union has taken the first steps toward regulating the use of AI. All this activity can make it difficult to separate hype from reality, and paralyze officials seeking to prepare for the changes AI will bring to the job market and public services.
Data can cut through the clutter and help policymakers design solutions. That, at least, is the aim of the Organisation for Economic Cooperation and Development. The Paris-based body on February 27 launched an online platform, the OECD.AI Policy Observatory. It’s designed to help the group’s 36 developed-nation members share data, research, and policy ideas for promoting trustworthy AI that’s beneficial for society. The website reveals how much the field has changed in recent years.
China’s rise as an economic superpower is one of the biggest stories of the 21st century. That rise applies as much to artificial intelligence research as it does to manufacturing. The country now is the leading publisher of AI research papers, ahead of the US and the European Union. That’s not surprising for a country whose e-commerce and social media leaders rival the giants of Silicon Valley in scale. India has made big strides too, befitting its status as a computing and engineering powerhouse.
Another notable trend is the growing clout of big technology firms in driving AI research, something that harks back to the days when Bell Labs was a leading innovator in the 1980s and 1990s. These companies have massive computing power that’s critical for pushing the boundaries of artificial intelligence, resources to hire some of the best talent in the field, and the incentive to turn new breakthroughs into commercial products and services. The likes of Harvard, MIT and Stanford still churn out plenty of top-notch research, but the brain drain to Silicon Valley causes some to wonder how long that will continue.
The rise of big tech in research coincides with a surge of interest in artificial neural networks. They enjoyed brief periods of popularity in the past but quickly hit roadblocks. This time may be different. The combination of today’s massive datasets and greatly enhanced computing power enables neural networks to learn through trial and error, and achieve stunning breakthroughs in pattern recognition. They power smart speakers and the facial recognition technology that unlocks your smartphone, and they hold the key to developing self-driving cars.