Hamilton Helmer's 7 Powers Rewritten
AI is challenging his rules of business
Conviction
One sharp idea each week about what and how we consume, why we crave it, and how it shapes culture.
My Working Thesis
We're witnessing a fundamental disruption of strategic power dynamics that's happening faster than most business leaders recognize. While AI startups may be temporary, the new rules they're establishing will permanently reshape how we think about competitive advantage.
But if we're truly entering this new era, what does it mean for how companies should build defensibility? Which of the traditional moats still matter and which have become obsolete? Will this velocity-first approach create sustainable businesses, or just a wave of profitable but short-lived ventures? And what happens when the foundation models themselves absorb the innovations happening on top of them?
This newsletter is me writing my way to the answers.
Temporary Companies, Permanent Change
I've been obsessing over Hamilton Helmer's 7 Powers framework for years in large part due to the Acquired Podacst. It's been my research lens through which I evaluate defensibility and moats or just new businesses in general. But something's been bothering me lately.
The AI companies I'm watching don't play by these rules. They break a lot of them. Repeatedly.
This week I’ve read a couple stories of founders who have built $1M ARR AI products in under 12 months with just a handful of people. No venture funding and no years long path to profitability. A million in a year before the year was done. And done so through rapid iteration, direct distribution via social media, and high ticket pricing from day one.
“They’ll probably be obsolete within a year” is what im always thinkings with new consumer AI software that im exposed to weekly. But what I also keep hearing and thinking is that they’ve “probably built enough to build whatever comes next.”
As someone who has been working in the startup ecosystem for 6 years, for as fast as I’ve been moving, this right here is next level. It contradicts almost everything I have come to known about building startups. Im seeing dozens of these companies emerging weekly, building significant revenue before most traditional startups would have finalized their pitch decks.
So I've been reexamining Helmer's framework through the lens of this AI revolution, and I've realized: three of the powers are being completely dismantled, three are being supercharged, and one has morphed into something entirely new.
The Dismantled Powers
1. Scale Economies: The Playing Field Flattens
I used to believe scale was everything. More users meant lower costs per user, creating an accelerating advantage for market leaders.
But I'm watching AI completely rewrite this equation. The founders Im seeing are leveraging foundation models to deliver capabilities that would have required massive proprietary infrastructure just months ago.
Some are reaching $100K MRR before hiring any employees. Their cost structure doesn't meaningfully improve with scale. They pay roughly the same API costs whether they have 100 users or 10,000. This fundamentally breaks the traditional scale economy advantage.
I think we're entering an era where being small and nimble might actually be the advantage we always used to think it was. The minimum efficient scale has collapsed to near-zero.
2. Cornered Resource: From Scarce to Abundant
There was a time where proprietary data was the ultimate moat. And now we are watching AI companies build high level voice, image, and text capabilities without those datasets. Simply leveraging foundational models trained by the LLM providers to match or even exceed what proprietary systems built over decades can produce.
I’ve used adobe’s AI capabilities and MidJourneys image generations as well. The difference is marginal if there is any. You can argue that MidJourney is significantly better.
The very concept of "cornered resources" is being challenged when the most valuable resource, intelligence, is becoming rapidly commoditized and accessible via API.
3. Process Power: From Systems to Speed
I've long believed in process as competitive advantage. The company with superior methods wins in the long run.
But I'm watching AI companies succeed with almost no established processes. They operate in constant unpredictability, changing direction weekly based on user feedback and model improvements.
Its almost impossible to establish process power. By the time a process would pay off, companies will be building something different. What matters is having superior adaptation speed. When the landscape changes monthly, documented processes become obsolete before they deliver returns.
This isn't sustainable for traditional businesses, but these companies aren't trying to be traditional. They're trying to capture value during a specific window of opportunity, then pivot or sell before that window closes.
The Supercharged Powers
1. Network Effects: From Connections to Collections
Network effects are evolving. The AI companies I see gaining real traction are those creating new types of network effects through data collection and user contributions.
ElevenLabs has built a library of thousands of voice models contributed by users. Each new voice makes their platform more valuable to everyone. This isn't the social network effect we're used to. It's more of a contribution network effect and it compounds rapidly.
I'm seeing the same pattern with AI image generation, character creation, and even productivity tools. The platforms that enable users to contribute to and benefit from collective resources are pulling ahead of technically superior but closed systems.
2. Switching Costs: From Lock-in to Personalization
Switching costs have become more powerful, not less. As AI systems learn from user interactions, they become increasingly personalized and therefore increasingly difficult to replace. Our AI text models know us better than google or any social media platform.
I recently tried switching from ChatGPT to Claude and there’s a large learning curve for the new tool and im teaching it very manually. After a few days, I switched back. Despite superior technology, or at least better equipped for my needs these days, the new tool doesn’t know me as well, my preferences, my voice, my common requests.
What I'm realizing is that AI creates a new form of data lock-in that's more powerful than traditional switching costs. Losing a system that has molded itself to your specific needs is more painful then data lock in of the past.
The companies that understand this are designing for progressive personalization from day one, creating an ever-deepening moat around each individual user.
3. Branding: From Status to Identity
Branding has always been powerful. But in the AI era, I'm seeing it transform from status signaling to identity alignment.
The AI tools people choose are becoming extensions of how users see themselves. I watch people fiercely defend their choice of AI image generator or writing assistant in ways that reveal deep identity connections.
"I'm a Claude person," they'll say, or "I'm a MidJourney artist"—statements that sound more like creative identities than product preferences.
Smart AI companies will be leaning into this, designing not just functional tools but identity ecosystems. They'll create communities, specialized languages, and cultural touchpoints that make users feel part of something bigger than a tool.
This is brand loyalty amplified to tribal affiliation. And it's incredibly sticky.
The Transformed Power
Counter-Positioning to Velocity Positioning
Counter-positioning was about adopting a model incumbents couldn't replicate without damaging their existing business. This might be a stretch but in this AI era im seeing something a bit different which is velocity positioning.
It's not that incumbents can't replicate what AI startups are doing it's that they can't do it fast enough. By the time they respond to one innovation, these companies have moved on to the next.
This isn't traditional counter-positioning. The advantage isn't structural, but rather it's temporal. It's about operating at a speed that makes your current position impossible to counter because you've already abandoned it for a new one.
What This Means For The Future
I don't believe most of today's AI startups will become enduring companies. Many will be acquired, copied, or rendered obsolete by advancing foundation models.
But I do believe they're collectively establishing new rules of competition that will outlast them.
The companies that thrive in this new landscape, whether startups or incumbents, will be those that understand:
Velocity trumps perfection. Shipping fast and improving faster matters more than getting it right the first time.
Revenue retention matters more than user retention. The ability to extract more value from the same users over time is the new growth metric that matters.
Identity alignment creates deeper moats than functional superiority. People choose tools that reflect who they are or aspire to be.
Distribution advantages compound faster than technical advantages. Getting attention quickly is more valuable than having marginally better technology.
I'm still processing what all this means for long-term business building. Can companies sustain the sprint pace indefinitely? Will consolidation eventually restore traditional power dynamics? I don't have all the answers.
But I'm convinced of one thing: Hamilton Helmer's framework is evolving. And understanding how these powers are being dismantled, supercharged, and transformed is essential for anyone building or investing in the next generation of technology companies.
The rules are being rewritten in real-time. The companies that recognize this first will be the ones writing the next chapter.