USA – SANTA CLARA –In 1965, Intel Corporation’s co-founder Gordon Moore famously predicted the rate of advancement of integrated circuits, noting that the number of transistors in an integrated circuit would double every year. A decade later, he revised his statement to doubling every two years and the concept is now known as Moore’s Law. Moore’s Law translates to a doubling of a computer’s computational capability every two years and continues to serve as a guiding principle for the technology industry. These advances enabled by Moore’s Law now produce a new mechanism to derive value from computing.
As we enter 2017, we are crossing the threshold of a new era for human innovation – one in which machines can analyze significantly more data and recognize difficult to find patterns in that information.
With the increased ability and affordability of computing, machines are able to continuously analyze and learn from data and accurately predict outcomes; this is what we call Artificial Intelligence (AI).
Researchers have been working on AI for more than 50 years. We are now starting to see the benefits of those decades of investment in capabilities such as machine learning and machine reasoning. These technologies exhibit tremendous potential to improve society in areas ranging from better healthcare and safer driving to more efficient agriculture and energy development.
For example, AI used in cars is now capable of interacting with not only the static elements of the road – signage, painted lines, etc. – but also pedestrians and other cars on the road. AI-enabled (also called automated driving) vehicles are currently being tested on roads worldwide and have the ability to assist drivers, optimize traffic flow, and reduce collisions resulting in saved lives.
Of course, AI has applications beyond vehicles that will impact the future. AI will benefit the industrial and agriculture sectors through the rise of smart factories and cities, precision agriculture and power management. The healthcare and pharmaceutical sectors are incorporating AI to assist with early diagnosis, precision medicine delivery and drug discovery. The financial sector is benefitting from fraud detection, risk assessment and asset management. Cybersecurity firms are using AI to predict threats of network intrusion, thereby, using data to protect data. Even the sports industry is incorporating AI for performance optimization, injury protection and fitness management.
To keep up with the potential of AI innovation and the benefit it can bring society, the world will need laws and regulations that embrace these innovations while protecting people from unintended consequences. Governmental regulation will need to keep up with the pace of AI advancements to protect the public and invest in long-term sustainable policy approaches. To ensure the benefits of the burgeoning AI field are made available to the greatest number of people possible, we must do three things: 1) Drive; 2) Democratize, and 3) Guide.
First, we need to “drive” AI by building the technology, optimizing its performance for AI applications, and advancing public policy to keep pace with AI development. At Intel, we are actively working with a wide range of industries to accelerate AI development.
To do that, Intel and many of our technology partners are also maintaining an open dialogue with national governments and international regulatory bodies to identify and implement long-term and sustainable policy approaches. By building and maintaining strong partnerships as these emerging technologies blossom, the computing industry and its regulators can work together to mitigate risk, drive accountability and reap incredible benefit.
Second, we need to “democratize” AI by working with developers, academia, government, industry, and civil society to promote open access to the data and tools that will fuel broad development of AI. This means engaging with the developers of tomorrow today. Isolated data stores, complex tools and a small talent pool currently limit access to AI. Government and industry both have roles to play to democratize AI.
Government can play an important role to encourage the creation of an AI workforce, including supporting graduate students and funding research. Intel has committed to democratize AI innovation, leading the charge for open data exchanges and initiatives. One example is the Intel Nervana AI Academy created to increase accessibility to data, tools, training and intelligent machines for a broad community of developers, academics and startups. Democratization will improve overall access to the technology and allow societies to reap the benefit from AI. Democratization also requires global data flows that protect privacy and the reliability, integrity, and security of the data. Restricting data within national boundaries may inhibit AI from taking advantage of greater data diversity that comes from study across different geographies and cultures.
Third, we must “guide” AI into the future to ensure it is used for the collective good and to solve critical problems previously believed unsolvable. Guiding AI will require that we grapple with complex questions of ethics, privacy, discrimination, employment and even human autonomy. Public discussion and debate must happen early and often as we advance AI to experience a safer and more productive world. While these will not be easy conversations, they will enable the successful and beneficial use of these promising technologies.
In 2020 we will look back on the capstone decade of artificial intelligence. That is to say, the time where years of research combined with the computing power made possible by Moore’s Law translated into bold applications. This is an exciting moment to be building the brains of the platforms that will fundamentally change the way people engage with the world.