Some AI risks more subtle than others

On a recent opportunity to chat with AI thought leader Toby Walsh, he shared that one of his concerns with the emergence of workable forms of the technology is the possibility of people losing the ability to think for themselves. While the jury is out on how many or even if the broader populace ever had this capability in the first instance, a recent high-profile example has lent some credence to Walsh's worry.

None other than Prince Harry came up short during a 'oh won't somebody please think of the children' speech highlighting the risks of AI. Since fate loves irony, the instant the perhaps AI-generated speech stopped transmitting to the teleprompter, the royal floundered and the lecture to the great and good at the Clinton Global Initiative Annual Meeting in New York ended there and then. 

Given his inability to continue off the cuff, one wonders if Harry had read and understood the contents of 'his' presentation prior to giving it.

By contrast, while piloting an early orbital spaceflight in 1963, NASA astronaut Gordon Cooper used nothing more than a grease pencil, his eyes and a wristwatch to successfully time re-entry of his Mercury 9 space capsule, along with guidance from Ground Control.

Various factors had conspired against Cooper, including problems with the stabilisation and control systems caused by a short circuit.

Cooper marked crosshairs on the capsule windows with the pencil, visually lining up the trajectory for a successful re-entry using the stars as reference points. His humble wristwatch kept track of the rocket retroburn necessary to slow the craft to an appropriate re-entry speed. With this most rudimentary technology in expert hands, Mercury 9 eventually safely landed closer to the recovery ship than any prior mission.

Had his calculations been awry, or had he relied solely on the technology for performing crucial tasks, Cooper may well have endured the same sad fate as Russian cosmonaut Vladimir Komarov.

Now, the counterpoint to Walsh's concern is that relieved from the burden of onerous tasks, people can and do apply their minds to problems society may not yet even know about. There's precedent: insulated and elevated from the hurly burly of daily toil, wealthy gentleman scientists of the past including the likes of Charles Darwin, Robert Boyle, Henry Cavendish and Antoine Lavoisier, contributed lasting advances to scientific knowledge.

Rather than an ignorant and indolent mass of humanity given to idle dissipation, perhaps the AI age will usher in a new scientific revolution as bright minds are relieved from tasks now performed faster and better by computers.

But be that as it may, the juxtaposition of these incidents highlights the very real risks of technology dependence. In Harry's case, consequence was a deeper shade of red than that to which he and his companions are accustomed. In Cooper's, it was literally life or an ignominious death.

It also surfaces some perhaps useful principles for using AI. If you must conceal your use of the technology, there's a strong likelihood that you probably shouldn't be using it.

If you aren't an expert at what the AI is doing, then depending on the output could be risky.

And finally, if you haven't done your homework, or if it makes a mistake of any sort, then the AI won't blush, but you most assuredly will.

And who would have thought, the risks about which the Prince was in the midst of cautioning against, are at once very real and now extend to making an arse of yourself on a global stage.