The Age of the Horse Carriage Is Over
In 1865, the British Parliament — in all seriousness — passed a law requiring a person carrying a red flag to walk 60 yards ahead of every automobile. Speed limits were set at 2 mph in towns and 4 mph in the countryside. The speed of the car was legally capped at horse-carriage levels. The reasons varied: “It’s more dangerous than a horse carriage,” “It scares the horses,” “The roads were built for carriages.” This Red Flag Act remained in force for over 30 years.
Looking back, it sounds absurd. But I believe the same thing is happening in 2026.
Does Doing More Mean Deserving More?
More people are getting faster at their jobs thanks to AI tools. Code reviews are faster. Writing documents is faster. Data analysis is faster. Naturally, an expectation forms:
“I used to do 3 things a day, now I can do 10. Shouldn’t I be recognized proportionally?”
This feels intuitive, but there’s a trap. Operating 10 horse carriages and driving 1 car are fundamentally different things. The person running 10 carriages is managing 10 coachmen. The person driving 1 car is a single driver. More importantly, what the automobile unlocked wasn’t “running more carriages” — it was “reaching places that carriages never could.”
The moment we place value on processing a higher volume of the same work with AI, we’re driving a car along horse-carriage routes on repeat.
It’s Not About Speed — It’s About Range
The real change from horse carriages to automobiles wasn’t speed. It was range. When people went from traveling 40km a day by carriage to 400km by car, they didn’t think “Let’s do the 40km trip 10 times.” They thought, “Let’s go to the new market 400km away.”
AI is the same. The point isn’t writing 10x more code of the same quality just because AI codes faster. The point is being able to tackle problems at a scale that was previously impossible.
- A prototype that would have taken a week can be built in a day and put in front of users for feedback
- A single person can understand and improve a codebase that used to be too large for one person to cover
- You can make reasonable judgments in domains outside your core expertise
This isn’t “working faster.” It’s doing things that were previously impossible.
I touched on this in a previous post about how I use Claude. After a year of working with AI tools, the biggest change wasn’t the speed of work — it was the density of work. Instead of doing more in the same time, the time I used to spend on boilerplate and repetitive tasks shifted to design and decision-making. Less time typing code, more time thinking about code. This is exactly the carriage-to-automobile transition. Not traveling the same distance faster, but fundamentally changing the quality of what you can do along the way.
Loading Car-Sized Vehicles with Carriage-Sized Cargo
Yet in practice, many organizations and individuals still apply horse-carriage standards.
Let’s look at one example closely. “Use AI to code faster and close more tickets.”
The premise here is that Jira ticket throughput equals productivity. It’s like measuring transportation efficiency by how many round trips a carriage makes per day. But anyone who’s used AI deeply for coding knows this: the moment AI speeds up your coding, it reveals that coding speed was never the real bottleneck. Ambiguous requirements, unnecessary features, wrong abstractions — those were the real bottlenecks. When AI accelerates coding, the fact that time was being wasted on everything except coding becomes painfully clear. At that point, the question shouldn’t be “How many more tickets can we close?” but “Do these tickets even need to exist?”
The same pattern is everywhere. “Use AI to write reports faster” doesn’t question the report format itself. Real-time dashboards can replace reports. AI can directly summarize the insights needed for decisions. Why is someone still building 20-slide PowerPoint decks? “Use AI to optimize existing processes” is like greasing carriage wheels. You have a car. Why are you reducing friction on carriage wheels?
Redefining Speed
What does it mean when we say “work is getting faster”? There are two perspectives.
Carriage-perspective speed: Doing the same work faster. Writing a report in 30 minutes instead of 2 hours. Finishing a code review in 1 hour instead of a full day. From this lens, AI is a “faster carriage.” A tool that accelerates existing work.
Automobile-perspective speed: Doing things that weren’t possible before. Shrinking the cycle from prototype to user feedback from a week to a day. One person running a full-stack service end to end. Detecting and responding to market shifts before competitors. From this lens, AI is a “car.” A tool that changes the range of what’s possible.
When we talk about “working speed,” which speed are we really talking about?
The Courage to Step Off the Carriage
To truly leverage the car, you have to step off the carriage. This is harder than it sounds.
The carriage offers familiarity. The rhythm of working at carriage speed, task sizes calibrated to carriage capacity, processes optimized for carriage routes. All of this has accumulated over years, and changing it means accepting discomfort.
What’s even harder is declaring you’ll drive a car in an environment that still evaluates by carriage standards. Saying “These tickets aren’t necessary” in a “close more tickets” culture is risky. Saying “I’ll build a dashboard instead of a report” in a “write reports faster” culture creates friction.
But the lesson of the Industrial Revolution is clear. Those who clung to carriage speed were left behind. So were those who drove cars at carriage speed. The ones who survived were those who understood the new possibilities the automobile opened, and changed the way they worked to match.
So how do you actually step off? Start small. Pick one task you’re doing this week and compare “how I’d do this without AI” versus “how I could do this fundamentally differently with AI.” Not writing a report faster, but having AI extract the insights the report was trying to convey in the first place. Not coding faster, but having AI validate the architecture first so you only write the code that truly matters. Stepping off the carriage isn’t a grand declaration — it starts with asking “why am I doing this task this way?” one more time.
The Role of Leaders
In this transition, leaders play a critical role. When leaders maintain carriage-era standards, the entire team drives cars at carriage speed.
- A leader who says “Use AI to do more” is loading carriage cargo onto a car
- A leader who says “Use AI to do things differently” is finding new roads the car can take
Specifically, leaders should be asking:
- What work does our team do that AI has made unnecessary?
- What was previously impossible that AI now makes possible for us to attempt?
- Are our performance metrics still from the carriage era?
The question isn’t “how to go faster” — it’s “how to go farther.”
That Said, Running Carriage Routes Isn’t Worthless
There’s something to acknowledge here. Not every task can be redefined.
Quantitative improvement can be the precursor to qualitative change. When someone who used to make 10 round trips by carriage now does them by car, the insight “this route itself is inefficient” emerges. Faster speed increases the density of repetition, and higher density reveals patterns. In the process of rapidly repeating existing work with AI, the realization that “this work itself is unnecessary” may come naturally. Rather than “do things differently from the start,” “go fast and learn” may be the more realistic first step.
Also, not everyone is in a position to change the route. The people who can say “these tickets aren’t necessary” are a tiny minority in any organization. Most people are in positions where they execute assigned work, and using AI to do that same work faster is genuinely valuable. If running carriage routes 10x faster by car creates real cost savings, there’s no reason to dismiss that.
The Red Flag Act had its context too. Early automobiles were genuinely dangerous, and pedestrian accidents were frequent. AI also has real risks — hallucinations, security vulnerabilities, copyright issues. Slowing down isn’t always conservative resistance. Sometimes it’s reasonable safety regulation.
Still, even acknowledging all of this, the core point stands. The Red Flag Act was eventually repealed. The horse carriage disappeared. The moment we treat quantitative improvement not as a stepping stone but as the destination, we become people renewing our coachman’s license in the age of the automobile.
Takeaways
- When we say “work is getting faster in the AI era,” it should mean we can now do things that were previously impossible — not just that we do the same things quicker
- Evaluating a car by carriage standards means its potential will never be realized
- Using AI to accelerate existing work is just the beginning. The real change is redefining the work itself
- People and organizations that ask “how to do things differently” rather than “how to do more” will be the ones that survive this transition
- We laugh at the people who created the Red Flag Act when automobiles arrived — but we might not be able to laugh at ourselves, clinging to carriage standards in the face of AI
References
- Locomotive Acts - Wikipedia — The history of the Red Flag Act enacted in England in 1865. A person carrying a red flag was required to walk 60 yards ahead of every automobile, with speed limits of 2 mph in towns and 4 mph in the country. The law persisted for over 30 years before being repealed in 1896.
- Red flag traffic laws - Wikipedia — Historical context on how regulations were applied to new technologies, not just in England but worldwide.