DeepTech Summary: The author of this article, Steve Blank, is very well-known in the Silicon Valley startup community and is often called the "Father of Lean Startup." He wrote "The Four Steps to the Epiphany" and is the originator of the Customer Development methodology.
Eric Ries' "The Lean Startup" is built upon his theories. He has taught entrepreneurship at Stanford, UC Berkeley, and Columbia University, and the U.S. National Science Foundation's I-Corps program is also based on his methodology.
Recently, Steve Blank had coffee with a founder he had invested in and discovered that after six years of hard work, the founder was completely unaware that the world outside had changed.
As a result, he wrote this article with a very direct core point:
If your company has been around for more than two years, there's a high probability that your business plan is already outdated. AI is reshaping development speed, team size, pricing models, and competitive barriers. Founders still running on the 2024 script may not make it to the next round of funding.
For readers who are either entrepreneurs or interested in the tech and venture capital community, this firsthand observation from across the ocean is worth reading.
Below is the full translation of the article.
If your company has been around for more than two years, many of your initial assumptions have likely become invalid.
You need to pause whatever you are doing, whether it's coding, product development, hiring, or fundraising, and take a look around. Otherwise, your company will die.
Anxiety Triggered by a Cup of Coffee
I just had coffee with Chris. Chris is a founder I invested in six years ago, and since then, he has been working tirelessly on:
2) In an existing market,
3) with a unique business model.
Chris is now preparing to launch the first round of large-scale funding. I looked at his investor deck and found an issue: While he has been head down working hard these past few years, the outside world has completely transformed.
The proprietary system software barrier he spent five years building is becoming less and less unique. The emergence of Ukrainian autonomous drones and ground vehicles has led to dozens, even hundreds of companies with larger teams and more funding doing the same thing.
Chris has been fighting for customer adoption in his niche market (which indeed needs disruption, but the incumbents are still holding strong). Meanwhile, the autonomous technology demand in an adjacent market has exploded: defense.
Over the past five years, VC investment in defense startups has skyrocketed from zero to $200 billion annually. His product is a perfect fit for logistics support and medical evacuation in a contested environment. But he is unaware of these opportunities in the defense market.
Chris's team has indeed done an impressive job of system integration (deeply integrating with an existing flight platform, making his solution different from most competitors), and the business is there, but it's no longer the business they originally envisioned.
After talking to Chris, I realized: for most startups that have been around for over two years, the business plan is outdated, and the technology stack and team composition are likely obsolete.
If you haven't looked up recently, here's what you've missed.
What Has Changed
VC money has heavily shifted towards AI. By 2025, AI projects took two-thirds of the total VC investment. This means that if you are not working on something AI-related, you are competing for a smaller funding pool. Non-AI startups must answer one question: why can't a better-funded AI-native competitor directly eat your market?
For software founders, AI has completely rewritten the old formula of cost, speed, and manpower. With tools like Claude Code or OpenAI Codex for Vibe Coding, an MVP (Minimum Viable Product) can be done in a matter of days, even hours, no longer requiring months. This also means that an MVP alone no longer demonstrates your team's capabilities.
These Tools Are Changing the Composition of Development Teams: The number of engineers has decreased, and the type of engineers has changed, with the emergence of "Business Process Engineers" and "Deep-Tech Engineers."
What used to require a whole development team can now be done by a few people, sometimes even by one person. Data used to be a differentiator and a moat, but now foundational models (such as ChatGPT, Gemini, Claude) are commercializing open data sources.
The concept of Agile Development itself needs to be rethought.
The old bottleneck used to be: Can we afford to build and release this product? The new bottleneck is: Do we know what to test? Can we reach users fast enough to learn? Agile is no longer a linear process. An AI Agent can run multiple tasks in parallel at the same or even lower cost.
You can now simultaneously test multiple versions of the same business, or even test different business directions at the same time. You can run five pricing models, ten marketing messages, and twenty UX flows simultaneously. And the "user interface" may no longer be a screen; the testing goal may have become: Find the prompt that gets the AI Agent to deliver the desired outcome.
The bottleneck is no longer engineering capability but has shifted to judgment, insight into customer expected outcomes, and distribution.
AI Agent Will Rewrite Every Software Category
The AI Agent will transform every software category, including the one you're working on.
Today's software applications operate by presenting information to users and then waiting for users to act through dashboards, alerts, workflow tools, and reports. However, customers buy software to get a job done, not to look at more screens. Enabling the job to be done is what the AI Agent (orchestrated through tools like OpenClaw) will autonomously achieve.
What does this mean?
If your product currently guides users on "what to do next," an AI agent will eventually take that next step for the user. If a competitor's product automatically completes tasks while your product still waits for the user to click a mouse, you will lose your competitive edge.
The next-generation applications will not just display information on the screen; they will act like an employee: resolving tickets, scheduling meetings, qualifying sales leads, and automatically replenishing stock. As products transition from "software is the interface" to "software is the outcome," pricing will also shift from per-seat charges to per-outcome charges: for every ticket resolved, every meeting booked, and every lead closed.
(The pursuit of Product/Market Fit will evolve into the pursuit of AI Agent/Customer Outcome Fit. The Minimum Viable Product (MVP) will become the Minimum Performable Outcome (MPO). I will expand on this topic in the next article.)
Hardware Is Not Spared Either
For hardware founders, the change is equally profound. Hardware is still constrained by physical laws, capital, the supply chain, and manufacturing cycles—you can't skip metal cutting, prototyping, or chip fabrication.
But AI can help you weed out bad ideas faster. Now, you can simulate more design variations before manufacturing a physical prototype, create a digital twin, stress test assumptions earlier and cheaper. The result is an acceleration in learning and discovery (sometimes toward faster failures), and in a startup, faster failure is an advantage, not a drawback.
Once AI is embedded as part of the system, the product itself changes. Adding an AI backend to a camera turns the camera into a monitoring system, vibration sensor, or machine failure prediction system. Robots become factory workers. The moat is no longer just the hardware itself but what the hardware can sense coupled with what AI can do with that data in terms of decision-making and action.
Sunk Cost Trap
Companies founded before 2025 typically optimized their tech stacks for a world where software development was expensive and bespoke. Agile development and DevSecOps made us lean, but they operated in silos, and team sizes were structured accordingly.
Companies that spent years building a "proprietary code and feature moat" are now finding AI is commoditizing most of their tech stack. This has put venture-backed startups in an awkward position: their business model may be partially or completely outdated.
When you are heads down building a product, searching for Product/Market Fit, these changes may not always be visible.
Your tech stack, product features, user interface, team size—all these sunk costs can become reasons you are unwilling to pivot: How can we throw away years of work? Our VC backed us in this direction. The customers still want the UI. The team believes in this roadmap. Our clients are not ready yet.
(Chris is a classic case. He built something truly impressive, likely competitive, but the business model around it needs to change.)
Some sunk costs are actually assets: deep domain knowledge, customer relationships, proprietary data, hard-earned regulatory approvals, physical integrations. These are worth holding onto. Chris's flight platform integration falls into this category.
The truly debt-like sunk costs are: large engineering teams built for a slow software cycle, seat-based pricing models, product roadmaps built around features rather than outcomes. These are the so-called "Dead Moose on the table," the problem is obvious, but no one is willing to address it.
The surviving founders are those who can look at what they've built and ask themselves: If I were to start a new business today using today's tools, in today's market, what would I actually do?
When you are already funded in a particular direction, this question is uncomfortable. But compared to investors telling you they are not going to fund the next round, and then you shutting down with an outdated plan, this discomfort is nothing.
Summary
· You can't run the 2026 race with a 2024 (or earlier) script. Funding, tech, business models have all changed. Agile development is turning into parallel development.
· The pursuit of Product/Market Fit will turn into the pursuit of AI Agent/Customer Outcome Fit. MVP will turn into MPO (Minimum Path to Deliverable Outcome).
· A sunk cost mindset will lead to your bankruptcy.
· Defendable moats may still exist in: proprietary data, deep understanding of customer outcomes, regulatory lock-ins, or becoming a Program of Record.
· If you can sleep soundly, it means you haven't yet figured out what's happening.
· A surviving founder will step out of the office, see the big picture, pivot, and course-correct.
