AI rant

The Knowledge Horizon.
Mistral doesn't use Artificial Intelligence, all Mistral intelligence is real!

From the time of Archimedes, Aristotle, Euclid, Plato, Pythagoras, through Isaac Newton, Leonardo da Vinci, Stephenson, Trevithick, Curie, Edison, Brunel, Nobel, Watt, Einstein, Logie Baird and countless others, all mankind’s knowledge base could, in theory at least be understood by one, single, intelligent human being. Given enough time to study.

Meaning that in the event of cataclysmic devastation such as global plague or war, following cessation of the event civilisation would have the means to recover and rebuild. I believe that a threshold was reached around the end of the nineteenth century which eclipsed that possibility. I call it the Knowledge Horizon.

Mankind is now dependent upon machines and electronics in addition to basic knowledge in order to survive. From crucial food production and distribution, to transport, manufacturing, construction, heat and power, defence, health care, education etc.

The driving force, for all this creativity and invention, apart from kudos and curiosity was also personal reward of course. Now we are faced with a new and seemingly impossible to resist dilemma. Artificial Intelligence.

Knowledge consigned to books, for education purposes is good of course. Handing over the product of lifetimes of genius, free on a plate to anyone who demands it, without any inkling of how to deploy that knowledge sensibly, is the irreversible opening of Pandora's box. It means of course that man will have no incentive to dedicate a large chunk of their time and energy, perhaps even a lifetime’s worth, into researching and creating anything whatsoever, if in the certain knowledge they will receive neither credit nor reward for doing so. Creating music, poetry, novels, innovative new tools and machinery, yes and even complex computer software code, will all cease. Incentive and initiative destroyed, forever!

My own personal story comes in here. I witnessed the dawn of the personal computer. During the late '70s I was encouraged and taught by my then American employers to use one. To save time and therefore money for the corporation I worked for. I did so. It worked and everyone was happy. Including me. I pushed the technology further and saw countless opportunities for saving time and thus enhancing profits. accountancy, marketing, engineering, staff training, manufacturing processes and tool controls for example.

My career had taken me into the world of energy engineering. Everything from a 20 metre long feed water heater for a nuclear power station down to a tiny heat exchanger for a kidney blood dialysis machine. Refrigeration condensers and evaporators. Even a car radiator or a diesel engine turbo charger intercooler. These are all energy movement tools. Along with hundreds of other examples. All indispensable tools for sustaining life in a modern world. Consuming over 20% of all mankind’s generated power on planet Earth in the process.

I decided to computerise the long, painfully skilled processes for calculating and designing heat exchangers. For all and any application. From kidney dialysis machines up to nuclear power stations.

Without exception, colleagues, friends and relatives told me I was barking mad. I was told no simple, relatively low power by today’s standards computer would cope with the sheer volume and complexity of the coding required. I ignored my cynics. I saw a solution. At times, researching and working, all too frequent 24 hour days, (yes really!) over 5 years of totally unpaid work. I admit that on occasions I began to doubt myself. I parted company with friends and even family alike, even permanently until this very day.

I succeeded and I am proud of having earned millions for my work as a result. I am also proud of having saved perhaps as much as 2% of the 20% of the world’s generated power. 0.4% isn’t a hell of a lot but it is infinitely more than what any duplicitous ancient TV presenter or hysterical, mentally sub normal, truant schoolgirl has ever achieved or ever will achieve!

Leeds De Montfort University is a leading college for IT (Information Technology). Two of De Montfort's lead professors saw my work and said it was impossible. Must be some sort of hoax. They interviewed me and I proved to them it wasn’t. They said "Who taught you how to think like that and to do things like that?" My reply "I have an advantage over you. You were formally trained and followed the rules. I wasn't and didn’t, so I had to teach myself. Quite simple really."

I was offered an honorary Bachelor’s degree for science. Conditional upon me touring the country with them on occasions over a period of up to twelve months. I rejected it. Upon the basis that it would not aid my then fledgling company by one jot, and therefore me or any of my new recruits. Besides which I couldn't afford the time. Now clients from many of the world’s largest ‘Blue Chip’ corporations and government bodies, through to self employed ‘One Man Bands’ depend daily upon the Expert System software I instigated. Over 25,000 licensed users across 140 countries world-wide. A tiny number in the world of IT but the market the company I founded in 1984 can best be described as serving is a relatively small 'niche market'. One embracing a collection of professional specialists.

And now AI wants to give my work away. So far and I believe indefinitely AI has failed dismally. In any event, no one wishes to spend days developing a set of unique design parameter questions for each new project to pose to ChatGPT or the like, simply for it to take two hours before presenting a hopelessly muddled, jargon infested, non editable, inevitably inaccurate result. Having 'scraped' input from any number of annonymous, dubiously qualified, untraceable and therefore totally unnaccountable sources. When MY intellectual property can achieve a certified, peer group assessed, provable, traceable, editable input and guaranteed result, along with an audit trail and a presentation quality printout, all in from under ten seconds to sometimes thirty seconds from start to finish. No main frame computer required. Just a simple PC.

We cannot uninvent AI. As I stated, it’s an irreversably opened Pandora's box. Every day though I see AI errors getting increasingly worse. Misspelled words and atrocious grammar today. Inaccurate mathematics today. Aircraft driving 346 passengers and crew into the ground at twice the speed of sound a few years back. AI could though be reined in and should be. By intelligent political leaders. Unfortunately few politicians of that calibre exist in the world today. You embrace AI if you want to. I never will.

Mistral's commitment:
Bringing benefits of computerisation to our RAC industry - without the commonly associated problems.

Close and return to previous page