As AI becomes an ever more widespread term, Geoff Codd CITP FBCS FIoD warns against misuse and contemplates the implications of misleading messages about what it actually is.
I started my career working with computers in 1957 following five years engineering in the Royal Air Force. If I told someone then that I worked with computers their eyes would immediately glaze over — it was a guaranteed social conversation killer, where public ignorance and zero interest in the subject predominated. As time moved on, my IT career included various board level roles in global businesses and as an undersecretary in the public sector.
Since those early days, now over half a century ago, we have experienced a transformation of our lives by successive technology innovations and our ability to absorb them effectively. However, public perception of IT today is at the other end of the scale, with widely held enthusiasm for Artificial Intelligence (AI) and all that it seems to stand for. This is due mainly to the fact that the term AI has now been widely adopted as a generic description for the latest clever version of an IT product, even though the actual use of an AI facility was not involved. On the face of it, we should all be delighted at such a relationship reversal with our users — but am I alone in having deep misgivings about where we are and where we are headed?
The web of AI
Interpretations of what AI is all about vary widely. I believe that conceptually it can be likened to a web of relationships, a bit like a complex spider’s web where constantly changing values and relationships flow along each strand to be constantly processed at nodes and directed accordingly — a bit like the human brain, except uninhibited by our natural brain capacity limitations. And it is that limited capacity where quantum computing science becomes an important enabler of an AI capability.
For you
Be part of something bigger, join BCS, The Chartered Institute for IT.
Following papers by Alan Turing and Viscount Nuffield in the 1940s and 50s it became clear that an AI capability, which would make ‘impossible mental processes possible’, was inevitable in the long term. In the meantime, exploiting ‘established’ intelligence (EI) which flowed naturally from all those successful systems experiences, was an obvious way to improve product effectiveness. Many developments today appear to clearly fall within that category, although under the totally misleading label of AI. However, most people clearly do not have the remotest idea about what AI really means — and misuse of the term simply devalues its long term significance, and perpetuates a totally misleading message.
The consequences of misinterpretation
If my life’s experiences tell me anything in the context of the latest quantum computing developments, it is that AI capabilities are a growing reality as we approach the crest of another wave in the exploitation of information technology. There needs to be even more heightened awareness that mature weaponised AI in the wrong hands will be capable of inflicting calamitous disruption on the world order. And whilst it is said that state attention is being given to addressing that challenge, it is also clear that public perception is being totally misled by the false interpretation of AI as currently promoted. The IT industry is certainly doing clever things with EI, which derives from half a century of information gathering and processing, but that is not remotely a genuine AI offering.
We have therefore now lost the original interpretation of the term ‘Artificial Intelligence’, together with the extreme caution which needs to accompany its exploitation. Surely it is our responsibility to do our best to bring the situation back on track. Perhaps simply encouraging the use of the EI label rather than AI would gradually reflect reality, and would permit proper preparation for the true artificial intelligence challenge that is now clearly on the horizon.
It is never too late to correct a misleading message. In this case it is too important to ignore.
