What does the Genie from Disney’s Aladdin has to do with Artificial Intelligence? If you haven’t watched or read about the Aladdin, Genie is a an all-powerful spirit (a “jinni”) residing in a magical oil lamp and the tritagonist in Disney’s 1992 animated feature film, Aladdin. Genie can basically grant any wishes and do anything as he likes. Genie is full of positive energy and full of life; also amongst the most powerful beings in the universe, Genie is benevolent, easygoing, and friendly.
It sure is nice to have Genie on your side or be your friends, you can do anything even rescuing your girlfriend or stopping the most evil person on earth. However, what if Genie is not all positive energy and friendly? Imagine you have to fight Genie and would you stand a chance? I would not even try getting noticed by an evil Genie.
A highly intelligence and autonomous AI is more dangerous than the most harmful substances in the world, more dangerous than the heaviest machinery we’ve developed, and could possibly cause more damage than some of our deadliest weapons. Indeed, part of its danger lies in the fact that the AI can orchestrate, manipulate, and make use of all of these harmful things in strategic ways.
AI governance and regulation is a solution to these potential risks. It’s too soon to quantify the risk, and I won’t stake my entire reputation on a prediction, but I think it’s safe to assume that unrestricted AI development poses a medium to high threat.
Like regulations that govern our use of drugs, alcohol, weapons, and heavy machinery, AI regulations can provide clear guidelines to the general public, clearly outlining what is considered acceptable and unacceptable when it comes to building and using AI.
Advantages of AI Governance & Regulation
There are a few advantages in regulating AI – some obvious and direct, others less so. I won’t run the entire gamut of benefits here, but it’s worthwhile to pause briefly and consider what we stand to gain from taking these important steps.
Regulation, for one thing, will mobilize campaigns that seek to communicate with the public and educate them about the dangers of misusing AI or developing dangerous technologies. As AI becomes an ordinary commodity, these educational campaigns will be essential. Similar educational campaigns about the risks associated with alcohol have been quite successful. Thanks to successful advocacy from groups such as M.A.D.D., young people all grow up knowing the dangers of drunk driving, and it’s an activity that has become highly stigmatized. Running comparable campaigns about the dangers of AI could lead to widespread awareness of AI safety and best practices.
The main advantage of AI regulation, however, will be the ability to put a stop to business operations or close it down entirely if it is found in violation in its use or development of AI. Companies that try to rush an autonomous vehicle to market before it is fully tested for safety or who work on an AI that will violate traffic laws so that it can deliver goods more quickly can be stopped before their products put people at risk.
Regulation does admittedly have its drawbacks. The first that comes to mind is cost. There’s a far lower sticker price to letting corporations develop whatever AI they please and not spending resources policing private uses of AI. Like all forms of compliance, it will take a cost not only on the regulatory bodies (the government agencies who will have to find ways to enforce regulation within its budgetary constraints) but the companies that build the AI as well. These corporations will require the services of specialized lawyers and software engineers to ensure that the AI algorithm conforms to all existing laws.
Governing and Regulating Intelligence Is Nothing New
If this seems like a heavy-handed approach, it’s important to realize that this is not the first time we’ve taken steps to regulate intelligence. In fact, even human intelligence is regulated. No, nobody is going to shut down our brains if we misuse them, but there are restrictions on what we legally allowed to do with our intellectual capabilities. For instance, we are not permitted to lie under oath, or scam people out of their life savings by exploiting their trustworthiness.