It's Time to Think About How You Replace Yourself
I've spent my career inside startups. One thing has never changed: in a startup, your job will be different tomorrow than it is today.
As an early employee of a startup you wake up Monday doing data entry and by Friday you're building the KYC strategy. The driver 99% of the time is the founder. You didn't ease into new skills. You got thrown into them. You pivoted or you failed. Reinvention is the job. That's how you scaled. Good founders know this, support it and find people who thrive in these early stage environments.
What I'm watching happen now is different. Not because the change is faster. But because the thing driving the change has a new engine. Sometimes it's AI, but just as often, it's your co-worker who just automated half their workflow. It's an executive who saw a demo and wants results by Friday. It's the PE firm mandating change.
The pressure is coming from every direction, machine and human, and it's compounding.
Whether AI systems become truly self-evolving in three months or three years doesn't matter. The systems we build today will be smarter than we are at building them tomorrow.
People ask me about AI constantly. At dinner, at events, at Boxcar on Pearl St. It always goes the same way. They start casual, then it starts to veer toward telling me about the moment the machine did something better than they could. Not faster. Better. Or how it hallucinated on something obvious the last time they used it.
Here's what I've learned and why it matters for everyone now, not just founders:
In times of rapid change, institutions respond with structure. Education, government, corporate HR. They keep reaching for their historical playbooks. New curriculum. New job structures. New policies. And by the time they ship, the ground has already moved.
Startups respond with iteration. This was considered the startup advantage. Covid showed us what building in motion looks like at global scale. We iterated in public. That's the muscle.
Not predicting the future. Building systems that adapt when you're wrong.
In any fast-moving environment, execution and the space to think further out are tied together. When you're buried in the daily, you can only see days ahead. Build systems that handle the daily and you earn the space to think in quarters. Build the right culture and you start thinking in years. Each one feeds the other.
That progression used to take 7-10 years. AI is compressing it.
Recommended by LinkedIn
This is the skill everyone needs now: the ability to operate when nobody has the answers. To build in motion. To iterate instead of freeze. That used to be a startup survival skill. Now it's happening to every company, every industry, every role. It's a life skill.
The one big change:
The bottleneck is no longer capability. AI handles capability. The bottleneck is something different and evolving, and because it's not capability, it's moving faster. Historically you needed to master the skill and acquire the license. It's now the scene in the Matrix where Neo learns kung fu in seconds.
Today we can see coordination is a problem, but we are not sure what comes next. That's where isolation becomes the real enemy. Not AI. Isolation. Most of us are experimenting alone. Some share on Reddit or LinkedIn, but feeds are ephemeral. Coordination requires memory and nobody can out-think what's coming. But we can out-adapt it, if we do it together.
When I moved from New York to Colorado, that idea hit different. New York is a "me" culture, you hustle, you compete, you protect what you've built. Colorado is a "we" culture. The mantra for years here has been "Give First". That instinct to share, to build together might prove to be exactly what this moment demands.
Isolation doesn't scale. Feeds don't compound. Collaboration does.
I've started to explore how we can systemize this idea at AI Colorado, it's infrastructure for shared learning. A place where practitioners share what's working and what's not. Then software connects them to people, events, experiments, vendors that might be able to help them get to the next step on their journey.
Practitioners helping practitioners and I think it can give a state like Colorado a competitive advantage. Join us, we are slowly letting people off the waitlist: aicolorado.org
What prompted this:
"Startups respond with iteration. This was considered the startup advantage… Not predicting the future. Building systems that adapt when you're wrong.” 💕
"We" culture is the future. Great post Robert Reich !
Solid piece RR; I’d complement this with our need to decipher signal from noise in our collaborations which in and of itself is a challenge. To your point about being asked about AI relentlessly, the volume of content about and around AI is, in a word, overwhelming. Collaboration is paramount, indeed. Thanks for sharing your insight! P.S. Enjoying the beta. Finding some good signal in all the noise.
Love the visual. I spend a lot of time on the visuals that accompany my posts. It's the purely fun part, for me. No one ever says a word about them!
Yep, the Matt Shumer piece was illuminating. Like your approach here.