thinking in systems; seeing in color
first, you feel like an idiot. then, you build something that changes the world.
“To dwell inside a well-functioning machine is to be largely unaware of its functioning”
–Siddhartha Mukherjee, physician, biologist, and author.
That’s my favorite quote. I’m super excited about this one - it’s essentially what I do.
The systems which control our lives are often completely invisible to us. Even if we’re aware of them, they’re impossible to grasp - at once, far too big and far too complex. If a universe of intersecting system defines the way we see the world, how do we even begin to understand them?
We can only hope to reckon with them in abstract or in small parts, because systems are infinitely large and interconnected. But the most important products leverage existing systems to create their own - all while creating atomic shifts that fundamentally and entirely change a system’s underpinnings.
understanding systems:
We’ve all dealt at length with systematic ‘wicked problems’ - pandemics, geopolitics, injustice, climate change. No matter their complexity, each system share a set of common characteristics:
systems are relative and mutable. There is often no true or false, only good or bad. There is no start or end either.
systems are composed of smaller sub-systems, and compose larger macro systems
a system has more than one correct explanation. systems cannot be measured, but their outcomes can.
systems are unique - they do not necessarily obey precedent but often fall into patterns.
systems are often self-referential, running on feedback loops.
For all their ambiguity, systems are still natural things. They exist for natural reasons. For humans, this typically has to do with the balance of survival, the proliferation of growth, the maintenance of power. Though systems can act ambiguously, there is always an underlying motivation at their core.
systemic change:
The nature of a system makes it nearly impossible for one actor to meaningfully change it - or even understand it.
To have a meaningful impact on a system, focus on cause over effect: there are a million 10% process improvements that band-aid existing solutions, but far fewer atomic shifts that get to the root of the problem.
These atomic shifts can be initially subtle - more subtle, even, than a process improvement, but play out in a huge way long term. Day-to-day life doesn’t change noticeably, but course-corrects over time.
Some examples:
internal combustion → electric cars: driving them feels about the same, but the way they work is fundamentally different. Electric cars can reach efficiency milestones that are physically impossible with internal combustion.
Web2 → Web3: the addition of a simple property - ownership of virtual assets - is enabled by a fundamentally different infrastructure composed of blockchains, staking, and virtual computing.
32 → 64 bit microprocessors: microprocessor evolution is driven by non-prescriptive breakthroughs but maintains staggeringly consistent growth (Moore’s Law).
working on systems:
Systems thinking, more than ever, is a critically important skill. Like learning a new language, it enables a whole new way of thinking. Understanding a system helps you think proactively instead of reactively. It can help you see invisible opportunities - to navigate deftly. Knowing the rules helps you figure out which ones to break.
Context is king. by default, take a step back.
Systems are not monolithic - they are complex
Systems are relative and mutable.
Always starts from the bottom up. To create an atomic shift, you need to start where problems are coming from, not where they’re going. Good solutions are built atop existing solutions. Great solutions are built on systemic bedrock.
Let’s take an important, everyday microsystem: company culture. It checks the boxes of a system - multivariate, hard to pin down, but you know it when you see it. Culture is often systematized into a series of principles, issued from the top down.
But culture actually arises as a series of undocumented daily interactions. Principles simply name what’s already there. And, like any other system, a culture is not put in place all at once. Its parts continually change, and just like the people it lives in, culture is constantly living, breathing, and evolving.
So how do you work on systems from within them?
Start with the foundation. Look for patterns. The principles of a system are evident in the process, not necessarily what’s said about it.
Operate in nudges, not commands. Play the long game.
Do not get lost in a system, but also do not try to oversimplify it. Do not overextend your scope.
Be wary of numbers: they never tell the full story. Reducing a system to a series of metrics garners some insight but loses a ton of resolution. Systems are human things, not numerical things.
Understand the underlying logic behind decision making (this is a system on its own)
building systems:
I build systems. Most of my work, until this point, has been about understanding them - the obvious and the non-obvious parts. My thinking process is naturally deconstructive; once the parts and their meanings are on the table, I rearrange them to figure out what’s signal and what’s noise. From there, we build from the bottom up, with a focus on creating atomic shifts - not 10% process improvements.
In practice: Helm (my startup).
Ok, so we start by creating a new foundation - an atomic building block. An irreducible part of a system that underlies the universe that will be built on top. Some examples: Apple’s original user interface, AirBNB’s rating system, Google’s ad tracking pixel.
This is sometimes driven by a technological innovation, but it’s often just a matter of reframing - by questioning the nature of the system in place.
With Helm, it’s a portable workplace identity that automatically populates based on an employee’s actions at work (Slack messages, sentiment pulses, Github contributions, etc). It’s owned and accessed by the employee.
On its surface, it’s like a living resume with data that’s extremely useful to both employees and employers. But once it’s framed as an atomic building block, it’s evident that this initial value is not an end but a means to a far more human-centered workplace.
This core workplace ID becomes foundational to a sea change and trillion-dollar markets. Here’s what might be built on top:
Talent marketplaces: with a depth of information, universal credentials, and soft skill data, Helm’s workplace ID makes it easy to find and evaluate the right fit for any position.
Professional community: Helm’s workplace ID builds community in a deeper, more detailed, peer-to-peer way - unlike existing superficial professional social networks.
Human Resource Management: Helm’s workplace ID unbundles the archaic, complex, and expensive Human Resource Information System (HRIS) by tying HR information (payroll, benefits, PTO) to a collection of individual profiles
Single Sign On: Helm’s verifiable workplace ID acts as a key to efficiently control access to workplace tech and sensitive information. No more data breaches!
Gig work profiles: a growing network of workplace IDs makes it simple to connect, evaluate, and hire the right contract worker for the job.
In the right environments, systems are massive value multipliers. Navigating them can be a challenge, especially at first. Like anything new and complex, those who work in systems are especially vulnerable to the Dunning-Kruger effect - feeling at first like a mastermind, then with more knowledge, a total idiot, and finally, with lots of effort, a mastermind once again.
I spent almost all of 2021 learning the systems that I’m now working to break, feeling like an idiot working iteration through iteration. But now, it feels a whole lot clearer. More to come!