Recycled Bullsh-IT

The Day the World Turned Day-Glo

Back in the age of shoulder pads and power ballads (mid-80s) one of my more adventurous consulting colleagues went off to join HP (Hardly Perceptive?) to work on an exciting Expert Systems project that would (allegedly) transform the way we process data. Expert Systems consisted of two main components:

• A Knowledge Base that stored facts and information provided by human experts
• An Inference Engine that applied predefined heuristic rules to process a problem statement and then output a solution

 Genetic Engineering

Expert Systems were as big in the Eighties as Artificial Intelligence is now, with over half of the world’s largest companies running Expert Systems to solve complex scientific and business challenges or provide insight and recommendations like a human specialist.  Given we rarely come across Expert Systems these days, what happened?

I Am a Poseur

Like all potentially useful technologies its bandwagon was hijacked and it became the panacea for all problems, and over-pitched to impetuous punters, leading to its meteoric rise and a disappointing damp squib of a fall. See previous rants: Hype, Hype Hooray! and Crystal Balls and Analysts.

Some of the underlying technologies resurfaced in other tools – Inference Engines as Business Rules Engines in BPM and Machine Learning software; Knowledge Bases in AI, IBM’s Deep Blue, etc.

 Art-I-Ficial

Fast forward to the 2010s and a whole bunch of new data and algorithmic tools and techniques coalesced into Expert Systems 2.0, now called Artificial Intelligence (AI).  So, is AI going to succeed better than its (in)expert ancestors?  Possibly. In some areas.  At some point in the future.

 Identity

The goals for AI are breathtakingly ambitious, encompassing Search Optimisation (corruptly controlled by the company formerly known for “Don’t Be Evil”), behaviour mapping & manipulation (from all the F.A.N.G.s), Natural Language Processing (“I’m afraid I can’t let you do that Dave”, Clueless Chatbots), Perception (Big Brother Facial Recognition, Self-Crashing Cars), solving the big questions (Life, the Universe and Everything), as well as replacing humanity itself (Cobots, Skynet, The Culture).

 I Am a Cliché

Now don’t misconstrue my cautious comments for caustic condemnation. AI has significant potential to unlock new areas of fundamental research and create enormous value for us all, well the puppet robot masters anyway.  Like any bleeding edge technology, treat with care, oblige the business with some low risk POCs and use it for CV polishing, rather than problem solving.

We are still in the AI summer, with the AI winter to come, before we can rebuild Westeros or the Westworld when AI goes mainstream. In the meantime, having AI as part of our funding pitch can increase the valuation of our tech start-up by 15-50%, so let’s take advantage of investor ignorance now!

John “Germ Free Adolescent” Moe