FCC Forum on Artificial Intelligence Busts Common Misperceptions
December 3, 2018
The Federal Communications Commission convened a Forum to examine artificial intelligence (AI) and machine learning, featuring subject matter experts from industry and academia. AI, generally speaking, is the ability of computer systems to exhibit human-like behaviors including cognitive reasoning, speech recognition, and social and emotional behaviors. We increasingly interact with AI technologies on a day-to-day basis through voice-controlled personal assistants, automated vehicles, mobile check deposits, and more. This surge in AI deployment has sparked new interest and prompted viewpoints opposing views ranging from “AI will solve all the world’s problems” to “robots will take over the world.”
Panelists at the FCC’s Forum set the record straight—explaining where we are, where we are going, and how government can facilitate AI technology development and adoption. AI technologies are good at identifying patters. They have more developed cognitive and reasoning capabilities, because the rules are more clearly defined and therefore easier to program. It is simpler to teach a computer how to play chess, for example, than to recognize and interact with objects or people. But progress is being made to improve computer perception and interaction.
Advancements in AI technology will provide tremendous benefits and opportunities. Technologies that can read lips or visualize and describe the environment will improve accessibility. AI can help make more efficient use of spectrum, as well as assist farmers in making decisions related to weather, soil, and plant conditions. It can assist in disaster recovery, enable autonomous vehicles, and even enhance our ability to detect and identify diseases such as cancer.
But even the most sophisticated computers of today cannot replicate all human behaviors. As Dr. Carolyn Nguyen, Director of Technology Policy at Microsoft, explained, AI systems are not adept at creativity, empathy, fairness, judgment, or collaboration. And unlike humans, AI is not forward looking. Although it will not solve all the world’s problems (or result in the demise of humans), AI can be a change food good.
Panelists explained how government can help maximize AI potential. Importantly, policymakers must recognize reality and reject notions based on Science Fiction. The reality is that AI is an emerging technology, and we do not understand fully how it will evolve and what use cases will look like. Regulatory humility is therefore key to enabling AI investment and innovation. Government should support multistakeholder processes, like the Partnership on AI, that develop voluntary best practices and guidelines for AI. Government can also convene experts to help shape AI development. This should include not only technology experts, but professionals from a myriad of sectors including health, automotive, agriculture, manufacturing, and public safety.
Stakeholders interested in developing or adopting AI technologies should consider grabbing a seat at the table. Developing common terminology and sharing best practices will maximize AI’s potential and mitigate risks, including risks resulting from data bias and privacy and cybersecurity challenges.