The launch of 'ChatGPT Health (opent in nieuw venster)' appears, at first glance, to be welcome news. A specialised AI that analyses your blood values and links them to your medical records. Useful. But when we zoom out, we recognise a strategic move familiar from retail: the Amazon model. And that has profound consequences for everyone building on top of AI.
The Strategic View: Platform vs. Product
Amazon grew by allowing others to sell goods, analysing the data, and then launching its own versions of the best-selling products (batteries, cables, basics) under its own label.
OpenAI is copying this playbook. They analyse billions of prompts. Do they see that millions of people are asking questions about their health? Rather than building a better general chatbot, they launch a specialised service that swallows that entire vertical whole.
For the 'scaffold builders' — startups that built a thin wrapper around ChatGPT to offer 'AI for healthcare' — this is the death blow. You cannot compete with the platform on which you are built.
The Pragmatic View: Trust and Risk
That said, two factors will determine whether this move by OpenAI truly succeeds:
1. The Generational Divide on Privacy
In boardrooms, discussions on compliance and privacy can last for hours. But look at the shop floor and consumers under 30. For them, sharing data in exchange for convenience is entirely natural. They trust that it will 'be fine'. Companies that bet on 'customers being reluctant to share their data' are betting on the past. Adoption will come from the bottom up.
2. The Hallucination Problem
Here lies the real boundary for commercial application. An AI that writes a sales email with a minor error is merely careless. An AI that provides medical advice and hallucinates (fabricates facts) is dangerous. As long as OpenAI cannot guarantee 100% accuracy — and they cannot yet — there remains room for specialists.
Conclusion
My advice for leaders and entrepreneurs: do not build a business case on functionality that OpenAI could add tomorrow as a free feature. The 'easy' layers (such as general medical advice, text summarisation, translation) will become platform commodities.
The value no longer lies in broad access, but in the deep niche. Position yourself where it is too complex, too specific, or too risky for the big players.
![Why OpenAI is becoming your competitor — the Amazon strategy](https://substack-post-media.s3.amazonaws.com/public/images/6decc848-7c86-44ac-b718-90cfe9817ff3<em>2048x2048.png" target="</em>blank)