Notes from New Hampshire, Summer, 2019
On July 26th, a group of colleagues including this writer were invited to participate in a discussion at the University of New Hampshire titled the Impact of Artificial Intelligence on Business and Society.Our host, Ben Porter, a professor in the Data and Analytics Department organized the session – masterfully herding us cats – and set the stage literally by describing the stages model of technology assimilation:
The Stages model was also the basis of our common experience as consultants for a leading technology strategy firm known as Nolan, Norton & Company. We advised major companies worldwide on how to harness information technology to their business strategies and to avoid common pitfalls and blind alleys. Hopefully this understanding equips us with a useful perspective on disruptive technologies such as Artificial Intelligence.
In fact, the crystallizing idea for the session at UNH was a concern that Artificial Intelligence presented real risks to society that were not being recognized and properly managed. The panelists were not of one opinion and, in all probability, had seven or eight views among the six of us. This writer believes that AI will do many positive things but simultaneously reduce the need for workers including those in better paying jobs, leaving only marginal employment for most people. For instance, AI has consistently outperformed physicians in making diagnoses. It has already empowered robots to better displace manual labor and is still at a relatively early stage.
Another speaker, Steve Hall, is deeply concerned about the need to harness AI or more specifically Artificial Super Intelligence, which is arriving soon. Steve cited a book by James Barrat: Our Final Invention and the End of the Human Era. ASI will empower machines with mental capacities thousands of times that of mere humans, enabling them to leverage these powers to continue to advance their own knowledge and even to create biological instances that could rule or possibly eliminate humanity. Other important figures such as the late Steven Hawkings and Elon Musk have expressed major concerns. Steve feels that immediate action on a global scale is required to keep AI under human control.
Paul Clermont returned the discussion to less cataclysmic terms, but cited that AI and related technologies are no longer viewed as an unqualified good, but rather perceived by the public as having a darker side. This is due in part to cyber attacks and the loss of privacy. Paul sees AI as eliminating the most repetitive and mundane work, which will leave behind those who lack cognitive and social skills. The disruption however, like other waves of technological change, should only be temporary.
Tom Scurrah is currently CEO of MyDataOnly, a firm focused on CyberSecurity. Tom discussed the weaponization of AI and how it is used by both nation states, criminals and terrorists to attack military and civilian targets. Tom described the phases of the “intrusion kill chain”, which is described in this link: https://www.lockheedmartin.com/en-us/capabilities/cyber/cyber-kill-chain.html. Bad actors can penetrate targets to exploit vulnerable applications and then move on to create backdoor access for persistent access. To avoid this, Tom cites three gaping vulnerabilities must be countered: Ignoring the actions of hostile nation states, the knowledge gap in the IT community and wealth inequality which creates alienated cyber criminals.
Bruce Rogow who left Nolan, Norton & Co to become EVP of the Gartner Group, described an earlier version of the “Hype Cycle” developed by Ken Branch, which highlights the long time – as in several decades – required to master a technology and achieve its benefits. He outlined some of the challenges presented by rogue AI applications and the need for controls such ascontainment vessels. These are techniques which encapsulate AI apps so that they could not create unintended consequences such as embezzlement or data breaches.
Bruce also discussed how AI applications might well be managed as employees with consideration to issues such as who mentors and monitors them.
All of us who participated thoroughly enjoyed and valued the session and hoped the students who attended found it useful. We also agree, despite somewhat diverging perspectives, that the issue of Artificial Intelligence and its manifestations is far more important than the attention paid particularly at a national level and hope that much more can be done.