Donkeys, Not Unicorns

2026-02-20 · Source: Towards Data Science · Field: Business & Management — Entrepreneurship & Start-ups, Corporate Strategy & Leadership · Depth: Intermediate, medium

Summary

Yariv Adan, General Partner at ellipsis venture, argues that the traditional venture capital model is fundamentally changing due to the commoditization of AI-driven software and services. While AI introduces massive economic opportunities by unlocking previously inaccessible value, it also creates significant commoditization risk, making many areas uninvestable for venture-backed startups. This commoditization stems from users building ephemeral apps with AI, an explosion of competitors due to lower entry barriers, and AI-driven changes in distribution, where switching costs approach zero. Big Tech companies are also moving up the stack, leveraging their vast user bases and development velocity. Adan proposes an alternative "donkey" model for entrepreneurs: building a portfolio of many small, niche, passive-income-generating businesses, automated by AI agents, rather than chasing a single "unicorn" startup.

Key takeaway

For entrepreneurs evaluating startup opportunities in the AI era, you should critically assess whether your business possesses a genuine, defensible moat against commoditization. If your venture lacks proprietary data or unique expertise, consider adopting the "donkey" strategy: building a portfolio of AI-automated, niche micro-businesses. This approach allows you to generate sustainable revenue and retain full ownership, offering a smarter path than chasing a high-risk, venture-backed unicorn in a crowded market.

Key insights

AI's commoditization erodes traditional startup moats, necessitating new entrepreneurial models beyond venture-backed unicorns.

Principles

• Differentiation and defensible moats are critical for sustained success.
• Commoditized magic unlocks value but introduces disruptive risk.
• Zero switching costs diminish brand loyalty and first-mover advantage.

Method

Automate ideation, market research, user research, prototyping, and analysis to build and manage a portfolio of bootstrapped, niche micro-businesses, with AI agents handling operations.

In practice

• Explore niche markets underserved by large players.
• Automate business operations using AI agents.
• Distribute risk across multiple small ventures.

Topics

• AI Commoditization
• Venture Capital
• Startup Moats
• AI Agents
• Portfolio Entrepreneurship

Best for: Entrepreneur, Investor, AI Engineer

Related on AIssential

• AT&T Ventures’ Head Vikram Taneja On The New Rules of Seed-Stage Defensibility — AI shifts seed-stage technical risk from buildability to defensibility, demanding early focus on moats and distribution.
• Get To Know The Latest Class Of Ultra-Fast Fundraising Unicorns — AI companies are rapidly achieving high valuations and securing substantial funding rounds at unprecedented speeds.
• Why things will eventually fall apart — The AI market, especially for LLMs, lacks moats, leading to commoditization, price wars, and unsustainable valuations.
• Building an AI Startup Is Easy. Selling It Is Hard — When product building cost is near zero, distribution, not the product, becomes the critical competitive moat.
• Ben Horowitz on AI Infrastructure, Economics and The New Laws of Software — AI fundamentally rewrites software's competitive rules, enabling rapid development but eroding traditional moats and demanding new infrastructure.
• You will never own the AI. You can still own what it cannot. - CTech — Ownership of frontier AI models is concentrating, shifting value to those who build on top of them.

Open in AIssential →

Editorial summary, takeaway, and curation by AIssential. Original article published by Towards Data Science.