Page 5 - Logistics News July 2019
P. 5

Thought Leadership








































          the potential described by many pundits. One       anonymised on the blockchain and its anonymity
          very common argument is to say that AI data        safeguarded by the distributed ledger. It could
          is not suffi  ciently secure and that it could be   then be run through AI algorithms in a centralised
          hacked. However, most data is not sensitive and    fashion in order to receive necessary insights.
          thus not worthy of the expensive protection of       Although their amalgamation still lacks mass
          the blockchain.                                    roll-out, pioneering initiatives break ground.
            Take the Internet of Things (IoT) as an          DeepMind, a London-based AI start-up owned
          example. All kinds of chip-equipped devices        by Google, which has access to sensitive data
          are permanently radiating reams of data, but       thanks to a deal with The British NHS, announced
          is the energy consumption of your fridge really    it will now use a kind of centralised blockchain to
          information you want to protect at all costs? And   ensure a robust audit trail. A public, immutable
          the costs can be overwhelming, especially when     ledger called the ‘Verifi able Data Audit’ should
          you are using decentralised blockchains.           make visible what data has been used for which
                                                             purpose by the AI algorithm. The initiative isn’t
          Reason #3: Blockchain and AI have opposing         as advanced as DeepMind had assumed, but it
          tenets                                             is characteristic for the most likely way to wed
          Despite the fi rst impression, blockchain and AI    blockchain and AI.
          are no natural fi t. Blockchain’s decision-making     The problem is not that the two technologies
          mechanism is decentralised, AI’s is centralised.   cannot be combined, but rather that too many
          The former is immutable, the latter probabilistic   innovators in the blockchain-AI complex appear
          and adaptable. The blockchain is transparent and   to be the hammer that makes them perceive
          tamper-proof, the decision-making process of AI    every problem as a nail. But, if the potential is to
          is often a puzzle.                                 be exploited, use cases and problems must be
            Of course, each of these dimensions brings       crystal clear. Encrypted number crunching comes
          with it advantages and disadvantages. However,     at a cost, and with the current state of blockchain
          the idea to remedy the weaknesses of one           technology, there is a big cost diff erential to
          (AI) with combining it with an opposite tech       traditional databases.
          (blockchain) doesn’t really work out because it      Thus, an AI-blockchain merger is not the be-
          would change the very nature of the technology.    all and end-all for AI’s shortcomings. Yet once
          If, for example, AI ceased to be probabilistic, it   both technologies become more effi  cient and,
          would be AI no longer.                             above all, once their combination starts to answer
                                                             pressing business problems, the merge could be
          A possible road ahead                              done rather quickly. After all, tech behemoths such
          That is not to say that the two technologies have   as Google, Amazon and IBM are leading players
          no synergies, nor that they cannot be combined.    in both fi elds. Those giants have all necessary
          But rather than merged, they should be applied     resources under one roof – large data centres,
          sequentially. For instance, sensitive data can be   armies of coders and abundant data pools. •


          July 2019  |  Logistics News                                                                          3
   1   2   3   4   5   6   7   8   9   10