Sunday, March 26, 2017

Does AI have a future in charge of people?

Toni Muzi Falconi Invited us to review the opinion of Luciano Floridi in Facebook professor of philosophy and ethics of information at the University of Oxford, and a Distinguished Research Fellow at the Uehiro Centre for Practical Ethics. His perspective is outlined in an essay "Should we be afraid of AI".

My response was as follows:

  • Singularity is a red herring. "Because it worked no better than kitchen paper, absorbing and being shaped by the nasty messages sent to it." is a key phrase (by Floridi) because capable and powerful AI will be aggressively used against humanity. It will be a weapon of man against man. It will absorb and be shaped by nasty people as much as nice ones. It could be a radical sect, or a nation state using it but it is certain that it will be used or corrupted as an act of war. Man's aggression to man is by far the most common threat to humanity (and much else besides). This is one of the reasons I believe that the PR industry must learn about it and be part of an axis to deny such threatening foes. Singularity in this context becomes irrelevant as an argument. We will be at war long before then and will be taming AI as a weapon of defence. Thus 'there can be no absolute AI". "We share the infosphere with digital technologies. These are ordinary artefacts that outperform us in ever more tasks, despite being no cleverer than a toaster. Their abilities are humbling and make us reevaluate human exceptionality and our special role in the Universe, which remains unique. " So, I am in Luciano Floridi's camp but for quite frightening reasons. Offer Daesh, the power of AI and it will use it for harm long before it could be sentient. Thus I suggest we prepare for the reality and not the fanciful which so many grand names suggest sentient AI might be.

I had hardly finished writing when the UK Prime Minister ordered Google, Twitter and Facebook to launch a fresh crackdown against online radicalisation in the wake of the attack on people and the Palace of Westminster in London last week. 

The PM’s spokesman said internet search and firms “must do more” to stop extremist material being posted online. Mrs May’s warning came amid a growing backlash against the world’s biggest digital firms which make billions while allegedly al­­­l­­­­owing would-be-at­tackers easy access to terror instruction man­­uals and hate videos. The reputation of such online organisation is now on the line.

There are commercial pressures too. Many companies are withdrawing advertising so not to be associated with such content.

Google, Facebook and others need sophisticated weapons to achieve this.

What then will the PR industry deploy to recover the reputational damage and the commercial disadvantage.

The answer is Artificial Intelligence. Its capability to identify the awful content is already being deployed.

AI is already being used as armament in this battle. But it can be used by unsavoury.

The soldiers in this effort will come from institutions like Bletchley Park. Many of them will be recruited into crisis management teams in PR consultancies and departments. 

In this way, the PR industry is inevitably dragged into the use and application of such Transformative Technologies.

AI will as varied and diverse as the competing factions attempting to use it. It cannot be marshalled into one amorphous capability to control humanity and the embryonic battles for the reputation of Google Twitter and Facebook show us how.