In the article, “Thanks to AI, Computers Can Now See Your Health Problems”, (Megan Molenti, Wired.com 1/9/2017), advances in Artificial intelligence and Deep Learning are being credited with taking facial recognition to a whole new level. This is wonderful technology that will likely enhance the lives of millions of people. The benefits of life saving diagnoses and improved treatment options for a myriad of afflictions affecting millions of people are exactly the kind of thing that AI promises to provide to humanity. See how this technology is in use today .
What Me Worry?
It only takes a moment to perform a scan that could be done with, or without, the knowledge of the person being scanned. Security forces already use facial recognition in public places to identify known suspects. It would be a logical extension to include the identification of personality disorders or mental states to identify threats. There are also many possibilities for applying this technology in new and different ways. Other applications might include: lie detection, health insurance eligibility, employment screening, college admissions, marriage suitability testing, in utero abortion counseling, criminal prosecution, immigration suitability, enemy combatant identification, triage survival likelihood, and on and on. If any of these possibilities make you uncomfortable, you are not alone.
While the potential benefits of this technology are great, it opens up the future possibility for the government, and any private industry with the means, to use this technology to gather extensive personal information on individuals and use it for their own purposes. In the darkest extrapolation, this technology in the hands of malevolent regimes could be the tool of choice for genocide. Fast forwarding to a time when autonomous machines are making life and death decisions for humans, survival choices could be made based on artificial intelligence, operating beyond our understanding, scanning for an individuals economic and social value markers to decide the preferred survival outcome. Let’s hope the preferred survival outcome is based on our preference and not what the machine may determine is in it’s own best interest. In a sense, by developing this technology we may have given future AI machines the keys to sort out defective humans. This alone may be cause to advance the AI Doomsday Clock one minute forward.
What are we afraid of?
Stepping back from science fiction fears, there are real social concerns that need to be discussed and addressed. The accelerating pace of technology is not only threatening privacy and personal liberties. New technologies being developed to utilize artificial intelligence and Deep Learning are creating new challenges for our social institutions. Each new transformative technology needs to be called out on it’s potential for a net negative disruption to our social fabric. This will require a proportional amount of investment in social sciences as in computer science. Government and legal institutions will need to develop an understanding and infrastructure to work with evolving technologies. Taxing authorities will need a basis for which to deal with the economic shifts that will come from fundamentally different income streams and wealth creation mechanisms while traditional tax bases become marginalized. The fear should be focused on the consequences of technology innovation far outpacing our ability to absorb it without creating short term social and economic breakdowns.