I only post on Sundays, but I felt like I needed this introspection, so I broke the rules. I made them after all, so I can break them just as well.
Chesterton's fence comes to my mind: every rule exists because a person once set it. Who knows if it's still valid? Make up your own mind.
Anyway, back to the subject. Here's the insight that triggered this whole chain of thoughts this morning:
AI, now and in the future
I'm pro-AI. But sometimes existing solutions in the market are the better choice.
The people who built it already made the mistakes you will make when you build your own solution, and will make them for you in the future while you can focus on your core business, instead of fixing your DIY software.
People ask me because they "want to do something with AI". After an assessment, I've given over half of them the advice to not optimize their core process through AI, but opt for a fully vertical solution that covers ~95% of their needs. For now, most personalized AI optimizations are in the implicit corner processes, like email, documents, ...
We can always look for further optimisations later, but the biggest ROI for an organisation is not always coming from AI.
That part of building software didn't change: just because you can, does not imply you should.
But being on the forefront makes me indecisive. I see the solutions, I see the opportunities:
- We sit together for 2 hours
- I do an hour of consolidation
The outcome saves people an hour per day. This is a no brainer.
This is just the tip of the iceberg, I see they are evolving into bigger projects. Saving more time, improving quality, saving money...
AI is arbitrage, and it's a real opportunity for companies that understand this.
For now, I have a headstart. But I know it won't last.
For the first time in my life, I really don't have a clue where all of this is going.
Things are evolving so fast, and what's unimaginable on Wednesday might be available on Friday.
The sky is the limit. Or better: it used to be. Seeing how coding evolved over the last year makes me wonder what comes next. Given enough time and assets, there is no limit.
My canary in the coal mine: coding is a solved problem
People don't always believe this, but given a good enough strategic frame, providing context with functionals and non-functionals, and given a good architectural blueprint and testing harness, your SoW will be better executed by an AI than by humans.
I've lived through it so I know. The latest models have gotten so much agency that you can barely make the distinction between human and AI work. And improvement is accelerating at increasing speed.
The human part remains. For now at least.
And coding is only the start. Everything that can be done on a computer will, most likely, sooner or later be replaced by an AI.
And then: robotics. I don't really believe in humanoid robotics. But purpose built robotics are already here: robotic vacuum cleaners, warehouse robots, ... All built exactly for their purpose, and very efficient in that way.
Even things like plumbing will be automated, as people or AI will invent solutions that allow us to automate it. It might take 1, 2 or 3 decades, but it's inevitable.
The nature of work changes
AI has profoundly changed the way I work, and the speed by which it evolved is only increasing. (Repeated, I know, but this needs emphasis.)
Currently, as I'm handing more and more responsibilities over to AI, my biggest challenge is keeping up with the initiatives my AI is doing itself. Using proper guardrails, contained and constrained, but I'm trying to build a dashboard to stay on top of it.
And this turns out to be hard. How do you decide what to trust and where to interfere?
If work gets done as you sleep, you need to be able to trust the process.
But, as you delegate more and more to the process, you give away control.
In that way, I think it's somewhat comparable to a regular org growing from a one-person shop to a behemoth of a few thousand people.
So we circle back to enterprise architecture (capabilities, lifecycles, alignment, ...), systems thinking, domain-driven design.
Right now the human-in-the-loop is still a hard requirement for me.
But I'm wondering how long this will last? When will I be redundant?
Which brings me to my next question: human relevance.
What remains?
Assume Elon Musk is right, and we're entering a decade of abundance: everything will become so cheap, people can get whatever they want for cheap.
What will be the remaining drivers? A brain dump:
Scarcity
Possession
People love to have something someone else doesn't have: diamonds, an exclusive trip or meal, a spot nobody has, the bigger boat, ...
Some people truly value the thing, but for lots of them, it's more about self-image and perception to the outside world.
I used to be more like that when I was in my twenties, until I realised by being between the haves and have-nots: once you can provide basic needs, the rest might create short bursts of happiness or fulfillment, but this is only temporary. Some keep going for bigger goals, some don't. Which is both fine: people should live their lives in the way they want it.
Time
There's only 2 certainties in life: death & taxes. And I'm not sure about that first one.
The first half is three centuries old. Everyone credits Franklin; Christopher Bullock had it in print in 1716, seventy-odd years before him, and even he was quoting a proverb. The doubt is mine. Maybe.
Losing people in my vicinity once again reminded me that we're only here for a short time, unless we manage to prolong it... Or what if we create a digital twin of ourselves, who could live forever? Do we still live by those means? Is this the future, or is it creepy? I can't make my mind up yet.
Is time scarce?
If you think about Einstein, you could somehow come to the conclusion that you can figure out a way to get more time through energy, although that might just be an odd brain-fart from yours truly.
Let's assume that in the following decades, time will remain scarce.
Arbitrage
I think arbitrage will remain: where you live, the people you know, the access you have, your family, friends, network, ...
Entropy is a thing of nature, and it will not naturally go away.
AI and robots will surface most of the arbitrage, but physical arbitrage (you can't grow grapes on the north pole for now) will remain.
Systems thinking as well as stoicism teaches us that everything is there for a reason. Things exist because they have a purpose.
What happens if man loses its purpose?
I think full abundance sounds nice in theory, but will never work. Parallels with Marx's utopian thinking ring a bell.
Creativity and taste
The current AIs are by definition statistical machines, trained to represent the most logical things based on what they saw in the past.
As these AIs will become a commodity, and thus very cheap, almost everyone will be able to use these.
AI slop as we call it. Abundance causes us to get bored. So we need to find ways to make the AI creative, in a way it triggers humans: taste, creativity.
I'm assuming this will not be learned in the early days, although things like an endless stream of AI video shorts has already proven for myself that a machine can easily keep me engaged for hours by providing me small dopamine hits every n seconds. I don't know.
Will we turn into mind-numbed machines, chasing one hit after another?
Which brings me to the next part.
Human touch
Given you're a believer of Darwin, we have been evolving over six or seven million years. A lifetime looks ridiculously small when you look at it from that timeframe.
Buy a piece of land now, put a nice house on it, and be extremely proud of what you have achieved. But the odds are someone has done it before you, maybe even on that same piece of land, that has now long gone and disappeared in oblivion.
Now think about the current phase in that evolution: a tiny dot on the whole timeline.
Why are we here, why do we exist, what's the meaning of life?
A question that has been asked a gazillion times, with answers like "42", "to reproduce", ...
Depending on who you ask, you'll get a different answer. Almost none of them wrong. The Nietzsches and Kants of this world spent their whole life trying to figure that out, so I'll take a rain check. Maybe later, maybe never, but not now.
Although, from a purely biological perspective: the human system evolved into a system regulated by hormones. I do believe hormones are the main drivers of what we do in life. Hormonal balance = happy life. But I'm just an armchair expert, so take this with a grain, or maybe better: a bag of salt.
Our hormones regulate our body, influence our brain and trigger things like pleasure, feeling loved, satiety, ... I don't see this going away in the next few decades, as nature's been shaping us like this for millions of years...
On the other hand, synthetic substitutes for the natural thing are being launched all the time, some with great success, while others cause the downfall of a person or worse: a community.
So who knows, maybe in the future we'll get a shot of dopamine directly, or cortisol to lose weight. Maybe those GLP-1 drugs that are currently all the hype are only scratching the surface.
Everything else
This is just a brain dump. I'm probably missing lots of things here, or my ideas might change next week. Anyway, on with the rest of this text...
Back to the main premise.
I still can't get rid of this feeling "what's next?"; a constantly nagging feeling of "is this good enough, is this what I should be doing?".
I have always lived seeing the future, launched some things a few years before they became a rage or a trend, and abandoned it long before it actually broke through.
My memento mori remains, as well as the bitter lesson. What I'm doing today won't matter in a few years. Think longer timelines: decades, centuries, and it won't matter at all.
Everything I learned and will learn will be surpassed by an AI, potentially impersonating me.
What's left? Hormonal balance?
To be continued...
Futures made of virtual insanity, now
Always seem to be governed by this love we have
For useless twisting of our new technology
Oh, now there is no sound, for we all live underground
Jamiroquai, "Virtual Insanity"