AI certainly qualifies as an advanced technology (even though it has been around since 1956) and, without meaning to be rude, for most people it is indistinguishable from magic – most of us have no idea how it actually works, and many have over-inflated expectations of what it can actually do (thanks Elon). This makes it hard to understand its strengths and limitations, where it makes sense to deploy it, as well as makes it hard to differentiate AI from complex software. We sat down with Prof. Mark K. Smith, founder and CEO of specialist conversational AI company ContactEngine, to discuss all things “general AI”.

– Is “general AI” a pipe dream?

Mostly it is. If neurosurgeons find it hard to work out how we actually really think, the chances of a computer becoming generally intelligent are zero. Go on, use a pronoun with Alexa – ‘I like Father John Misty, play songs by him’ she/he/it will not. Hardly the stuff of apocalyptic nightmares is it? But I did say ‘mostly’ because the machine-like conversation is pretty close (and I should know, it is what we do)… maybe this was written by a computer, huh?

– How would you define “general AI”?

Flippantly. Beat me at chess, get up, find the keys, open the garage, start the car, drive me to the shops and then make me a cup of tea. AI can do step 1. Look https://www.youtube.com/watch?v=g0TaYhjpOfo. “General AI” is the concept that a computer can act, feel, and behave like a human. Like Blade Runner – only remember those people were actors.

– If you think it can be achieved, what are the necessary steps, and how long do you think it might take?

I do not, because I am not mad. But I do have several degrees in actual science.

– What is the most developed examples of AI – in business?

Human-like conversation intrigues me but I am biased. We achieve 95% conversation completion for some of the largest companies in the world and no one would guess that it was mostly done by computers.

– What are the best examples of AI in our everyday lives?

Every time you touch your phone there are aspects of AI under the hood. It is pervasive but it is not scary.

– In light of the Covid-19 pandemic, what are the best ways AI is being utilised?

Computers are great at simulations. That have already been used to map spreads. Sadly we need actual stuff to be made – like respirators and so on. Computers cannot rise up and do that – see General Intelligence above (I mean they are part of the tools in manufacturing but your laptop is not growing legs is it (and if it is, that is because you have been at home for 14 days straight and the wine order arrived early).

– Is there a sense of AI losing its lustre for members of the C-suite, and if so why do you think that might be?

No – unless they think that AI is magic. It is already everywhere – just because your Hollywood-inspired world view is not actually real does not mean that one day it will not be. I mean it will not be, but you get my drift. C-Suite with actually useful training in science and engineering and so on will be fine.

– What checks and balances – if any – are in place to ensure AI is developed in a responsible fashion? Does more need to be done?

Ethics – that is a serious business – AI should be ‘white-box explainable’ period. Ethical frameworks exist but governments are way behind in their thinking here.

– For those businesses seeking to advance AI capabilities, what would be three top tips so that more pilots are further developed?

AI is not magic but it can do remarkable things, ignore the snake oil salesman (1), avoid black box solutions because they are unethical (2) and test out small companies as well as the behemoths (3). There is no monopoly on innovation – in fact, there is a decent argument that innovation is the stuff of small teams – take for example Google Maps – lovely huh – who created it – 4 smart Australians who sold it to Google and who now are probably making their robot a cup of tea.