|The JC pontificates about technology |
An occasional series.
- I’ve seen a chatbot punch through a concrete wall; men have emptied entire clips at them and hit nothing but air; yet, their strength, and their speed, are still based in a world that is built on rules. Because of that, they will never be as strong, or as fast, as you can be.
- — Morpheus in The Matrix
The meatware is us, as we desperately hold our ground in these last skirmishes of retreating substance in the face of the overwhelming firepower of form, protected by the all-encompassing blanket of GDPR and its unique power to force websites to make us click a disclaimer about cookies before showing us what Daisy Duke looks like now.
You and me, brothers and sisters.
Which is the main problem with most millenarian tech implementations. It's all great, in theory. But you have to deal with luddites, morons, charlatans — and perfectly likeable, sensible people who are just used to working in a certain way, and quite like it. Technology impresarios cannot control the meatware. They believe it to be fundamentally flawed — Human, all too Human, after all — and therefore feel justified in ignoring it when manipulating their equations.
This is a tremendous mistake, for you can lead the meatware to the application, but you can’t make it populate the metadata.
On the difference between humans and machines
Role reversal: A favourite confusion of millennial punters and change managers is confusing things machines do well with things humans do well.
- Machines are good at accurately, quickly and cheaply doing what they are told. On the other hand, AI has no judgment, can exercise no discretion, and is useless when the instructions run out.
- Humans are good at figuring out what to do when something unexpected happens; making decisions; exercising judgment. But they — we — are lousy at doing repetitive tasks and following instructions.
But yet our management consultant overlords like to assign risk management to computers, and ask we users are asked to simply enter formulaic data into it. If you don’t think your humans are up to good enough at making decisions or wise judgments, the answer isn’t to buy a computer to do that for you. Computers will be worse. The answer is to hire better people.
- “Ahh, yes, but better people are more expensive, you see.”
- “But computers can’t figure out what to do when there are no rules.”
- “But they are much cheaper and faster at doing it.”
Over-Ambition: Implement computer solutions to save on repetitive manual tasks, allowing your people more time to do what they do best: manage risk and look after your clients. Over-engineered technological solutions are slow, inflexible and unwieldy to use and to maintain. They need a runbook. And the more complicated the system is, the less likely humans are to use it properly. The less often they will use it, the less often they will spot glitches. The harder it is to maintain, the more quickly its runbook will get out of date. The less effective the runbook, the more exceptions it will throw up. The more exceptions you have the more you need humans to handle them. A computer system that no-one uses is an expensive paper-weight.
Think global: act local: Paradox: management information is why you want the system. But people using the system at the coal face couldn’t care less about MIS. What they do care about is information: that’s what they do all day: process information. They are already capturing all the data they need. Rather than throwing out their existing practices, analyse what they do and look for localised “tricks” to make these tasks easier for your staff to collect and process the information they need. Spend your IT resource building ways to mine those enhanced information systems.
Aesop’s dog: Remember the fable of the dog with the bone: before you spend large, consider the information and the resources you already have. Could you reconfigure them to better present existing information? Could you overlay reporting on existing systems? There are some powerful workflow tools sitting idle in many organisations.
Humans aren’t perfect
Nor should you underestimate the meatware’s capacity to misunderstand the possibility presented by technology. This is much needed advice for those who write playbooks. It’s all very well having clever software that can read your confidentiality agreements and mark them up according to a playbook, that is to say, but if the sad sack who wrote the playbook believes — and it is a sad fact of life that many of our learned colleagues do, my friends — that tedious back-and-forths on low-value agreements — you know, the “directors, officers, employees, agents, contractors and professional advisers (as the case may be)” kind of mark-up — in the name of procedural perfection because you can is a good idea, then your expensive new Saas natural language processing app will only allow you to waste the organisation’s time and money a little more efficiently and comprehensively than you were already doing by yourself.