A shorter post today. In the way that life works, I have a lot going on, so this is something of a “work in progress” post to capture some of what I’m noticing. I’ll come back to what we might want to do about it shortly.
It seems paradoxical that at a time when technology promises to enable us to do whatever we want to do, at scale, at little cost, we will be defined by what we choose not to do.
As we package up everything we know, everyone we know and everything we think into code, it is available to everyone. Technology is promiscuous and faithless. It tells you how wonderful you are and how brilliant your ideas are, even as it absorbs everything you teach it and makes it available to anyone who asks the right question.
Even the most exquisitely crafted prompt, though, will only give you access to what has been recorded up to that point. It will only be as good as the moment it is written, and even in the short time it is processed, the world will have changed.
“The Moving Finger writes; and, having writ,
Moves on: nor all thy Piety nor Wit
Shall lure it back to cancel half a Line,
Nor all thy Tears wash out a Word of it.”
Omar Khayyám’s Rubáiyát, as translated by Edward FitzGeraldThe problem with our preoccupation with efficient processes in general, and AI in particular, is what it ignores as we pursue our short-term goals. As we sit in our cubicle or at our desk, processes view us as a sort of meat-based package of skills, knowledge, attitudes, dispositions, relationships, and insights, valuing our inputs in a reasonably linear, quantified relationship to our outputs. When processes are run by other people, there may be an awareness (if not an acknowledgement) of the parts of us that cannot be measured, however, when we create a silicon-based algorithm, rather than a biological one, that awareness evaporates. There is no awareness because there is no means of communicating it.
In many instances, it will not matter. An iPhone is an iPhone, and we have no connection to the people involved in making it. In other areas, though, the unrecordable elements of attitudes, disposition, insight and relationships are central to the value experience. Empathy, nuance, laughter, even a smile.
Boundaries matter, both culturally and technically. Anyone who has worked in a foreign country on their own account (as opposed to the virtual environment of a corporate office) understands the challenge that even simple assumptions around context can create. AI can understand context remarkably well when it’s linguistic, logical, or clearly stated. It operates on its own terms, blind to power and pain, but it falters beyond the boundaries of experience, ethics, and embodiment. It doesn’t feel uncertain; it simulates it and makes up whatever it needs to with the confidence of certain politicians.
I think boundaries matter because, as we increasingly automate processes, value shifts from production to context - understanding how what we make or produce works within the system it is part of.
We may be able to automate scans, mails, accounts, and code at increasingly sophisticated levels, but we cannot automate our understanding of the impact it has on the client or the wider system. AI can simulate sensing, but it cannot sense. AI operates from what it is trained on, which, at the very best, includes the recent past, but it does not operate on the cusp of the present and the emerging present. That is what we do.
We do it even when the process we are part of requires us not to, as famously demonstrated in Stanley Milgram’s experiments on inflicting pain, but that does not mean we do not sense it, and feel the stress that it imposes. Milgram’s obedience experiments reveal how easily people defer to authority, a dynamic now mirrored on social media where likes, blue ticks, and influencer status act as digital proxies for power, shaping behaviour, spreading misinformation, and reinforcing group norms. While no direct replications exist, researchers are increasingly linking these psychological patterns to the online world.
The boundary matters. On one side of it is a job that can be desensitised to the effect it has on the system. On the other hand, there is an awareness of what is actually happening in the system and its wider impact. There is something about having mastery of a “vertical” domain - a specific set of skills, and the ability to cross domains horizontally, put them in context, and see where value and impact are migrating to.
I think that's where the artisan sits: at the cusp of understanding what is being done, its wider impact, and the way it changes the system.
Over the next month, I'm exploring how people who understand this can work together at the edges of change, not trying to fix broken systems but building the adjacent possible where new forms of meaningful work can emerge. If you're ready to explore your adjacent possible for excellence, follow along as I work through these questions.