For investors and founders
Investing in the AI era
The market is shifting from rewarding AI stories to demanding AI durability. This is our writing on what that changes, for investors sizing up an AI company and for founders reframing to navigate it.
The short version
How to evaluate an AI company
Five questions separate a durable AI company from a wrapper the next model release erases. This is the Wrapper Test, in brief.
- 1
What survives the next model?
Does the company get stronger as base models improve, or does the next release make it redundant? Durable companies ride the wave; wrappers get swallowed by it.
- 2
Where is the proprietary advantage?
Proprietary data, distribution, or workflow depth a competitor cannot copy in a weekend. If the only asset is a prompt, there is no moat.
- 3
What is the dependency risk?
How exposed is the business to one model provider's pricing, terms, and roadmap? A single point of dependence is a single point of failure.
- 4
Do the unit economics hold at scale?
When inference volume multiplies, does the margin survive? Many AI products are demos that cannot afford their own success.
- 5
Can the team build the next version?
The current product commoditizes. The real asset is whether the team keeps building faster than the platform catches up.
The writing
Essays on AI-era investing
2026-07-27 · founders
The reckoning is a reframe: how AI startups navigate the shift from hype to scrutiny
The AI funding climate is turning from story and growth to durability and proof. For founders with a real business, that is not a threat. It is a reframe, and here are the four moves that make it.
2026-07-26 · investors
Growth is not durability: reading an AI company in a hype cycle
While the feed argues about whether AI is a bubble, the useful question for an investor is narrower and older, is this company's advantage durable. Here is how to read one when valuations and noise are both high.
2026-06-05 · investors
How to tell a real AI company from a wrapper
Every company is an AI company in the pitch. For investors, the hard question is whether the advantage is durable or a thin layer the next model release will erase. Five questions get you to an honest answer.
2026-07-24 · clients
AI capex is someone else's problem, until it is your unit economics
The headlines are about hundred-billion-dollar data-center spending. The version that matters for your company is smaller and closer: whether a given AI feature earns its own running cost.
What the market is arguing about
Six threads shaping AI investing
Drawn from where the debate is actually happening. Our take on each cuts past the noise to the company-by-company question of durability.
Questions
AI investing, answered
- How do you evaluate an AI startup?
- Beyond growth, assess durability with five questions: does it get stronger as models improve, where is the proprietary advantage, what is the dependency on one model provider, do the unit economics hold at scale, and can the team build the next version. A company that answers all five well is riding the model wave; one that answers none is standing in front of it.
- What is the difference between a real AI company and a wrapper?
- A wrapper adds a thin layer of interface or prompting over a third-party model, with no proprietary data, workflow depth, or system a competitor could not rebuild quickly. A durable company owns something that compounds and gets stronger as the underlying models improve. The Wrapper Test separates the two.
- How is investing in AI different in a possible bubble?
- The bubble debate is a market question and mostly unactionable for a specific deal. The company question, whether this business is durable, is answerable on its own evidence. In a hype cycle, disciplined diligence is the edge, because the market prices stories and durability is what remains when the story ends.
- What should AI founders do to stay fundable as scrutiny rises?
- Reframe from story to durability: move from demo to production, from wrapper to owned advantage, from growth-at-any-cost to real unit economics, and from a loud narrative to a durable one. The correction sorts companies rather than destroying the good ones.
A deal that needs a real technical read?
Independent technical and market diligence on AI companies: what is durable, what is a wrapper, and where the risk lives. A verbal read on the biggest risks comes early.