What’s Next In Artificial Intelligence?

With the vast amount of data available in digital form, the field of Artificial Intelligence (AI) is evolving rapidly. Innovations in digitization, analytics, artificial intelligence, and automation are creating performance and productivity opportunities for business and the economy, even as they reshape employment and the future of work.

Some companies are gaining a competitive edge with their use of data and analytics, which can enable faster and larger-scale evidence-based decision making, insight generation, and process optimization. But there is room to catch up and to excel.

Coming over the horizon is a new wave of opportunity related to the use of robotics, machine learning, and AI. Companies that deploy automation technologies can realize substantial performance gains and take the lead in their industries, even as their efforts contribute to economy-level increases in productivity.

Recent advances in robotics, machine learning, and AI are pushing the frontier of what machines are capable of doing in all facets of business and the economy. Physical robots have been around for a long time in manufacturing, but more capable, more flexible, safer, and less expensive robots are now engaging in ever expanding activities and combining both mechanization, cognitive and learning capabilities—and improving over time as they are trained by their human coworkers on the shop floor, or increasingly learn by themselves.

The idea of AI is not new, but the pace of recent breakthroughs is. Three factors are driving this acceleration:

• Machine-learning algorithms have progressed in recent years, especially through the development of deep learning and reinforcement-learning techniques based on neural networks.
• Computing capacity has become available to train larger and more complex models much faster. Graphics processing units (GPUs), originally designed to render the computer graphics in video games, have been repurposed to execute the data and algorithm crunching required for machine learning at speeds many times faster than traditional processor chips. More silicon-level advances beyond the current generation of GPUs are already emerging. This compute capacity has been aggregated in hyper-scalable data centers and is accessible to users through the cloud.
• Massive amounts of data that can be used to train machine learning models are being generated, for example through daily creation of billions of images, online click streams, voice and video, mobile locations, and sensors embedded in the Internet of Things.

The combination of these breakthroughs has led to spectacular demonstrations

Formidable technological challenges must still be overcome before machines can match human performance across the range of cognitive activities. One of the biggest technical challenges is for machines to acquire the capability to understand and generate natural language—capabilities that are indispensable for a multitude of work activities. Digital personal assistants such as Apple’s Siri, Amazon’s Alexa, and Google Assistant, are still in development—and often imperfect—even though their progress is palpable for millions of smartphone users.

For companies, successful adoption of these evolving technologies will significantly enhance performance. Some of the gains will come from labor substitution, but automation also has the potential to enhance productivity, raise throughput, improve predictions, outcomes, accuracy, and optimization, as well expand the discovery of new solutions in massively complex areas such as synthetic biology and material science.

The application of AI and the automation of activities can enable productivity growth and other benefits not just for businesses, but also for entire economies. At a macroeconomic level, based on our scenario modeling, there’s estimation that automation alone could raise productivity growth on a global basis by 0.8% to 1.4% annually.

AI and other technologies can also be broadly beneficial for society by helping tackle some “moonshot” challenges, including climate change or curing disease. AI is already being deployed in synthetic biology, cancer research, climate science, and material science.