Guest post from Jim McCarten
Recently I went to a retirement party for a former colleague, who’d put in over 37 years at our previous employer. This made me reflect on my own time, from joining, to leaving, the world of ‘IT’ as was and how this has altered so drastically in that time.
It started in 80s, the 1980s, not 1880s, (to be clear, I never knew Babbage or Ada Lovelace) but it feels like I could have done when looking at what has changed during my working life. My introduction to programming was at the legendary ICL Beaumont training centre (it had a croquet lawn!). It seemed at the time very much like managing filing card systems that just happened to make use of computer technology recently evolved from steam engines and valves.
Then and Now
My first project on returning to the ‘Computer Centre’ after training was to work on some new-fangled ‘PC system’. The old hands of the early Eighties didn’t believe in them – “not real computers”! It was bad enough that their beloved mainframes from ICL and IBM now had DEC VAX and PDP-11s ‘minicomputer’ interlopers nestling alongside them in the computer room. But now, there were ‘PCs’, tiny little things that could be carried around in a car, so these would be left to the trainees. An interactive system was required to handle medical records at Paddington General. It had user screens (‘Forms’), users typed words in, it recorded them and created records for processing. Still used COBOL, (but “Micro COBOL”, not the real hard stuff), but then a step further – ‘Code generators were implemented to deal with the repetitive tasks of screen handling. The code created was serviceable but inefficient, requiring ‘fine tuning’ by human intervention. And that was the template for a few years hence. Fast forward to early ‘2000s’, databases had become ubiquitous replacing the earlier record-based filing systems, and SQL was dominant. And it became a boom time for new technologies, paradigms (they didn’t half ‘shift’ a lot apparently!) and jargon. Having survived the ‘Millenium Bug’ relatively easily, and ‘Object Oriented programming, reusable blobs of code, frameworks, shared services, Virtual Terminals, Cloud Computing, Software as Service, etc, etc all arose, becoming the norm.
But still the human overseer was there directing all. ‘Expert systems’ crept in, configuring specifications and management of our HP infrastructure, but only as an aide. And suppliers still delivered ‘application packages’, technical staff configured, tweaked, programmed “add-ons”, “custom interfaces” and Oracle DB required specialised DBAs etc. So, for example, when modules of our main application package turned out to be extremely inefficient, to the point of grinding to a halt, the expensive DBA Administrator contractor could reverse engineer components to discover where the ‘[non-human] code generator’ had been used, fix it’s deficiencies and improve performance no end. There was ‘Art’ as well as ‘Science’ as instructors on the courses often mentioned. And it took a fair bit of experience to acquire those skills. But now?
The world of AI goes mainstream
Full circle – organisations in the future may only need to acquire the right AI tool, skills in designing prompts and questions, then let it write all that code, optimised using the expertise of all who went before and there you go. No need for any more programmers, no need for packages – the cottage industry returns, on a mega scale.
When living and working in a world where intellectual property theft more generally becomes the norm, and is in fact a driving force, how will the world avoid bleeding the well dry?
If nicking everyone’s else’s data, expertise, knowledge is mainstream – who will be left to be the creators? As with considerations of Knowledge Management systems in organisations, what are the incentives to play ball? How much is it worth to me to give it all to AI or rather ‘Artificial Knowledge Parasite (or Vampire if you prefer?) (AKP/AKV) or ‘Artificial Automated Plagiarist Technology’ (AAPT) – just thought I’d invent some new acronyms there before AI does it!
And looking at the ‘user’ side of things, staff had to be given access to dedicated terminals, then it became PCs running interfaces, virtual terminals, etc and you knew where you were managing a limited range of technologies. Incidentally it also rendered obsolete all TV drama/film where it was necessary to find a ‘terminal’, brilliantly crack the password in 30s, then proceed to type very noisily on a character-based screen which made the noise of an electronic typewriter for some reason. Now, massive computer power was in your hand with tablets and smartphones. And organisations, such as the HE Institution I worked at had the challenges of ‘Bring Your Own Device’. It was bad enough trying to support systems to work with Windows and Apple devices securely. Suddenly there were many more headaches when everybody wanted to connect everything into your nice secure infrastructure. And that’s before the ‘Internet of Things’ and users insist their smart toaster wants to access the university library…….
How has the world become so much more amazingly complicated in such a short space of time? Amazing leaps in technology offer so much possibility and benefits but as ever once Pandora’s Box has been opened……
Who will manage, support, lead in offering the services that users want, expect, ‘didn’t know they needed’? This has been a stunning change in data systems, moving from relatively simple highly organised and structured record collection and display to huge amount of information exchange, gathering, collation, synthesis as the norm.
And what’s it all for – just what is all this data collected for? You knew where you were when you just needed names, addresses, those old records of their purchases/courses/insurance details etc.
And Now……? Now it seems like IT (the AI led version) is about managing virtual life forms who want to own your soul, your essence, and all parameters that define you. Or that can replace you? Previously we have advanced our careers by managing teams, now you may end up as the shepherd of AI entities that will do the job of everyone – how do you manage them? You really can’t take them to the pub, buy cakes on birthdays, organise team bonding exercises. It really won’t work – you may need to acquire an office puppy to give you that missing personal contact and feedback while you watch the AI consume most of the electricity and water of what used to be called the UK, prior to becoming ‘DataCentre One’ (with due acknowledgement to Orwell J ).
And the AI entity will learn. Each time you correct and moderate, it learns, ‘it’ has no qualms about replacing you. Think of the film ‘Alien’ and the evil company android played by Ian Holm describing the ruthless predator they have brought on board unwittingly, that will destroy the crew. “I admire it’s purity, it’s logic. You have my sympathy”. Of course, mentoring humans may also result in intentionally, or otherwise, they replace you ( Keir Starmer anyone, Once and Future, or not, PM?). But at least it takes a while, you get feedback, satisfaction, reward, a sense of achievement? Does that dynamic work when you pour your lifetime expertise into an unthinking ‘Black hole’ of knowledge absorption? And it learns far too quickly – the lifetime of experience is absorbed in months, weeks…days? A friend who works as a consultant in project management recently told me how his last task on a recent contract was to implement an AI driven project planning facility. He then watched it assemble a PRINCE Project Initiation Document (PID) in minutes, using 200 organisation documents, many of which he had created for said organisation, over a course of months. Retraining schemes in plumbing may have a growing appeal.
How will employers, whether commercial or service oriented, private or public manage the staff they will need then? What are the skills that will be required, what will be the reward and motivation mechanisms to achieve this?
“My job is to support the dopey AI system – I am here to serve”
And there are other major concerns at a different level of course. Who owns, manages, controls and directs the growth of this capability, this data, this ‘collective world reality’? Which is now a major debate in society and politics and we await the answers.
Will your new AI friend (or Master?) challenge you, tell you that you are ‘wrong, misguided, off track’, give you the benefit of its wisdom acquired through years of learning, and thereby make you a more rounded and enlightened person? Or just echo you for a while, then push you over the metaphorical cliff? We can encounter the problem with ‘Yes wo/men’ in the ‘Real’ world as it is. Do we need more of that? Be your own judge, think carefully and critically – “Is this right?” Now more than ever these critical thinking skills should be developed hand in hand with the technology drivers. And that’s for another essay.
Again, thinking back to the dawn of this world, and the golden prospects it offered in this New World. I still remember optimistic articles promising unlimited leisure time and idyllic lives as the technology explosion made our collective lot easier. So far, it doesn’t quite seem to have gone that way. Those horizons have moved further ahead , but how many will be on that ship sailing forward, and who’s thrown overboard? And is it a good time to check those plumbing courses out?
Or maybe the bubble will burst…