Share this Post
By Anthony Olabode Ayodele (Chart.PR)
It was in the late ’90s. Grandmaster and World Chess Champion, Garry Kasparov, after winning the first, suffered defeat during a second encounter in the hands of the IBM created Deep Blue. As a freelancer for The Guardian, Nigeria then (I was staffed later) I monitored the final moments of the historical duel and reported the same for the paper.
All headlines regarding the feat heralded it as the first time man would be defeated by his own creation. A major controversy surrounding the issue (aside from the possibility of the machine being human aided during play) was since Kasparov’s moves in his previous encounter with Deep Blue were fed into the machine, seeing it eventually win, was it ethically right to claim the machine was better than him?
Nevertheless, it remained a milestone achievement for Artificial Intelligence (A.I).
Around twenty years plus as a delegate, I attended the Digital Transformation Expo held at Excel, London, from October 9-10th, 2019. Presentations included those bordering on Artificial Intelligence and the digital age with Keynote speakers such as Stuart Russell, noted as the Godfather of Modern A.I, and Gary Kasparov himself, now a speaker and public commentator on Innovation, Digital Age and A.I, addressing the gathering.
Stuart’s address on the first day, in a presentation titled “How not to destroy the world with A.I’ traced its history. He referenced Allen Turing’s 1950 paper which mentioned the possibility of humans losing control to machines more powerful than themselves.
Additionally, Stuart cited an article in Baltimore Post pointing out that A.I could end the human race. An issue he raised in his book: Human Compatible: AI and the Problem of Control, stating that AI could become "an enemy of the human race".
While Stuart further claimed the upsides of 'beneficial' A.I include the 'enormous enhancement of civilisation and a probable ten times World GDP increase; he explained dangers included the need to check the "irrational", "nasty" "myopic" and "malleable" nature of humans as it relates to A.I's usage.
Further, he warned that we need to prepare for ‘Human-level A.I’. He said recent advancement saw the creation of ‘simulated babies’! Noting there’s been several attempts - as the ‘Turing Test’ which seeks to identify the extent to which a machine can pass off human addresses – to produce machines that in appearance and voice could be mistaken for humans!
Kasparov, on the second day, sharing the stage for the Keynote address spoke of his experiences playing Chess against machines, including the Deep Blue. He said in 1998 he introduced the concept of Advanced Chess, “a form of Chess where Humans use computer chess programs to explore the possible results of candidate’s moves.
“It is not about us competing against machines, it is about us collaborating with machines,” Kasparov told the gathering.
Organisers seemed to have other ideas in mind, however, as questions fielded at the end of the session only featured questions asked via a stipulated digital platform, with no one in the audience recognised/ invited directly, to ask a question.
I met up with Kasparov after the session noting that since Deep Blue only defeated him after his moves in their first encounter had been uploaded and used to aid the machine during the second, doesn’t this suggest that A.I could only portend danger to mankind with the deliberate help of man himself?
He noted his book on the subject had addressed the issue.
My takeaways from the event? My questions from a PR, ethical perspective include, is it a goal of man to create machines that could pass off as humans, programmed to defeat man (like Deep Blue) and possibly wreak havoc amongst mankind? Or, better put, what’s the rationale behind man producing machines to compete with – and probably destroy - rather than, as Kasparov put it, complement - the human race?
Further, I ponder, whether it isn’t time for the introduction of a mandatory ‘Transparency Disclosure’ in A.I policy, where the creator/users of technology are made to clearly indicate where a machine that mimics a human being isn’t actually one! As the real intent shouldn’t be to deceive man into believing otherwise
Or is it?
Maybe the untoward nature of humans that Dr. Stuart pointed out in his presentation is what we need to address, squarely, first?
Bode Ayodele is a Chartered PR Practitioner and the CEO of Ideas Concepts and Products (ICP Media). He is also a member of the CIPR Education and Skills Committee.
This Article first appeared in Influence