Ideas on Disruptive AI

Boston-based venture capitalist, Rudina Seseri, is the founder of Glasswing Ventures, and lucky for me, a close friend from my undergrad days.  Rudina has spend her entire career working in or around the tech industry, beginning with a stint with CSFB's Tech investment banking business under Frank Quattrone, (the famous, and then infamous tech investment banker,) followed by an MBA at HBS, and an managerial role at Microsoft.  From there she entered the VC world, first working for one VC and then starting her own.

Click hear to listen to an interview with Rudina at a UK tech conference this week:  ow.ly/z1mh30b19MU  Rudina and her host are discussing the future of AI and what VC firms like her own want to see.  Rudina's firm focus solely on AI and machine learning-related businesses.

One of the catch phrases the tech industry types likes to use when identifying good ideas and investment opportunities is "disruption."  AI was first invented in the 1950's but it isn't until recently that we have the opportunity to see its disruptive influence on business and human behavior.  Facebook was disruptive. The apple computer and PC were disruptive. Cars and airplanes were disruptive. Industrialization, before that, was disruptive. The gun was disruptive.  Fire was disruptive.  The first wheel must have been disruptive.  Such shifts, and perhaps AI as some do fear, could introduce cataclysmic disruption to society.  Why?  I believe it all comes down to the future we choose to imagine for ourselves.  Will it be dark, competitive, controlling, and nihilistic? Or, will we favor visions of the future that aim to ensure the continued success of life on earth, sustainability, cooperation, and an improved physical, emotional, social, and environmental experience?   

Rudina speculates that the next great innovation will likely come from a young person around the age of 20, with a focus on programming and solving problems using an innovative incorporation of AI.   I always find these speculations questionable because young people haven't been around all that long. They have yet to experience some of the greater frustrations of life, and yet, her point is fair.  If they have the time to invest in learning new programming languages at an expert level, and the curiosity to identify pain points in the world around them, plus a commitment to solve those problems with utter focus, why shouldn't they be the break-through innovators?  After all, most older people are not independently wealthy or untethered enough the explore solutions with fresh eyes.  Have you ever worked in a company that struggles with change?  If not, watch this slideshow on the 5 Monkeys Parable.  This is akin to achieving adulthood.  Its a rare old bird,  a la Bernie Sanders, who can connect with young people in honestly believing there is a vision of society that is dramatically different and improved from the status quo. Too many accept certain 'realities' as unchangeable, even though there is always great change, eventually.

In defense of the older generation, I have a few pain points I can envision addressing through AI.  You be the judge of whether they are utopian or dystopian.

1.  Adaptive Education.  I have a son who was recently diagnosed with dyslexia, after much delay.  He also happens to be gifted in function areas that are not as complimentary to the traditional model of education.  What if AI could help him?  What if earlier, more accurate diagnosis was possible? Right now a child can't be officially diagnosed until age 9 from what I'm told, which is fourth grade, and he didn't receive his testing until 6th. What if it could design education methodologies that his teacher could pull up specific for him?  Such AI tools could be rolled out nationwide, sharing the expenses nationally so that no one group of children is singled out as a financial drag on the system, as they are now with special ed funding.   Can AI help build more recommendations for more successful learning environments.

2.  Behavioral Recognition.  As a derivative of the above, if AI were trained to identify live-interaction performance during an interview for disability testing for educational reasons, could it be used to identify other behavioral indictors?  Think lie-detector testing in law enforcement, job interview(groan), or simply passively scanning a crowd for suspicious behavior with heat-sensing movement-recognition pattern identification?  Would it confuse feelings of guilt with literal guilt?  Could it catch a terrorist or would it equate a serious bad guy with a cheating spouse setting up a meet.  The one that should feel most guilty, maybe does not.  

In my Sci-Fi fiction class Sophomore year of high school we read a lot of Issac Asimov.  I enjoyed his work, as I suppose most do.  Sci-fi overall is an interesting genre because it explores the many layers of utopian and dystopian fantasies that are possible with technical advancement.   It often feels to me that we are living in a world where futuristic dreams was set long ago, and now we blindly pursue those fantasy concepts, just out of curiosity, but perhaps not out of necessity.  When we look at where capital investments are made, are they going to real-world problems that affect our ability to remain viable on earth?  Waste and recycling?  Water?  Healthcare?  Pollution?  Food? Affordable housing? Species and community preservation?   In some cases yes, in many cases no, I believe, because of a perceived lack of ROI or immediate value in wealthy societies where AI is being spearheaded.  Perhaps, its just because we are at the early stages of using AI, and as I presented in my last post, not everyone can master the tools so easily.

One thing is for sure, nothing will happen with out access to data, which may be one of the upshots to Congress's recent actions reducing customer privacy protection.  In an article entitled, "How India is Winning Its War on Human Waste," published today on Gatesnotes.com, Bill Gates observes, "What gets measured gets done."  So true.  This quote encapsulates the upshot on the exponential growth in data collection.  

Hopefully in the coming years we can dream big, dream positively, and collect data reasonably and responsibly, for the benefit of future generation, and not just the next quarterly earning or political election.