Page 23 - INC Magazine-November 2018
P. 23

in the work being produced. To win   “To start off, you watch what they          p
               your employees’ trust, Coplin recom-  are doing via the nanny cam, but after
               mends taking an incremental approach   a while you start to relax as you gain
               to A.I. “Apply the algorithm to a small   confidence in how they work.”       SURfacInG
               portion of the overall workload to give                                       tokenomicS
               humans time to see how the algorithm   the end of Work as We know It?
               works, and to build trust that the out-                                       In the next decade,
               come is what was expected,” he says.  The growth of machine learning with-    your customers will pay
                                                                                             for goods and services
                  One example he cites is a new   in companies ultimately raises the type
               algorithm-based table-booking system   of existential questions that executives   in ways that sidestep
                                                                                             traditional currencies:
               a large restaurant chain implemented.   don’t like to confront: How many of us   with tokens, blockchain­
               Initially, individual restaurant manag-  will actually work with machines in   enabled and distributed
               ers were skeptical that an algorithm   the future?                            units of value that
               in the cloud could do a better job at   The reality is that change is inevi-  businesses can issue to
               managing the tables than they could.   table, so companies need to work on    transact with customers.
               To alleviate their concerns, the com-  ensuring a soft landing for those cur-  A current example is Civil,
                                                                                             a decentralized network
               pany agreed to allocate just a small   rently doing the type of functional or   for news with for­profit
               portion of available tables to the algo-  repetitive tasks that automation can do   and nonprofit arms that’s
                                                                                             raising capital through
                                                                                             token sales to fund some
                           “leaRnInG to tRUSt a.I. ISn’t                                     of its work. Those who
                        that dIffeRent fRoM When yoU                                         own Civil’s tokens can use
                                                                                             them to start their own
                            eMPloy a neW BaBySItteR.”                                        newsroom on Civil’s
                                                                                             platform. And they can
                                                                                             also use those tokens
               rithm, and if the managers were happy   better. For some, that may mean       to barter with others on
               with the results, more tables would be   retraining or upskilling staff to get the   Civil’s platform to, for
               added. After starting with a pool of   most out of their institutional knowl-  instance, build out adja­
                                                                                             cent services and apps.
               just 10 percent of available tables, the   edge and experience. Others, however,
               managers quickly realized that not   will inevitably find employees auto-       The value of tokens is
                                                                                             set by their issuer, but that
               only did the algorithm do a great job,   mated out of a job, just like those at   value won’t rise unless
               but it also freed them up to do more    Foxconn, which in 2016 replaced       there’s market demand for
               useful tasks.                   60,000 workers with robots.                   them. Think of them as a
                  Stephen Ufford is the co-founder   Yet, as we look to the long             privately issued currency
               and CEO of Trulioo, an A.I.-powered   term, could it be that the fears of     that skirts traditional issu­
               global verification service to support   automation-induced mass unemploy-    ers, so what was once
                                                                                             controlled by governments
               financial services’ anti-money-   ment have been overblown? After all,        will soon be available to
               laundering monitoring. Traditionally,   the next generation of workers—the    all, thanks to blockchain
               this important area of banking security   Alexa generation, for want of a better   technology and clever
               was handled by human workers, but   term—will already be used to living       cryptocurrency entrepre­
               increased computing power and the   with and learning from machines. And      neurs. In the near future,
               sheer volume of digital data now being   these new workers are showing signs   we’ll see new token­based
                                                                                             business models, which
               produced have left them outnumbered   that they are more motivated by the     could revolutionize not
               and outgunned by criminal gangs.   experiences, freedom, and creativity
                                                                                             only how payments are
               Now, algorithms like Trulioo’s scan   they can have in their work than by     made but also pricing.
               millions of transactions at a scale no   conventional incentives.             That’s because blockchain
               human could manage and are trained   “Two of my grandparents worked           technology can facilitate
               (by humans) to spot potential fraud or   in factories, yet my grandfather also   micropayments with no
                                                                                             transaction costs: No
               block suspicious individuals.   loved to paint,” says Ufford. “Shouldn’t
                  When dealing with something as   work harness that creativity rather       smart merchant should
                                                                                             accept a two­cent credit
               sensitive as identifying fraudulent   than crush it?” It could turn out that   card charge, because that
               transactions, Trulioo’s employees    machines, in the end, are the mass       cost is higher than the
               had to be certain the algorithm they    creativity catalyst we’ve all been    price. But future busi­
               had built wasn’t showing any bias in    waiting for.                          nesses might charge tiny
               its decision making or making rogue                                           amounts for certain goods
               recommendations. “Learning to trust   MATTHEW YEOMANS is the author of        and ser vices—all facilitat­
                                                                                             ed through tokenomics.
               A.I. isn’t that different from when you   Trust Inc.: How Business Gains Respect
                                                                                             —AmY Webb
               employ a new babysitter,” Ufford says.   in a Social Media Age.

               8 2  ●  I n c .  ●  n o v e m b e r  2 0 1 8   ●  ●  ●   ● ● ●
   18   19   20   21   22   23   24   25   26   27   28