Every company that adopts AI gets better copy, better design and better campaigns than it had before. So does every competitor, with similar tools trained on similar data. The result is a market full of polished, interchangeable communication.
This article looks at the controlled evidence for that effect, why it matters more when agents do the choosing, why verifiable truth is becoming a legal requirement, and four practical steps to keep your edges. Opinions are labelled as Andrea d’Agostini’s view.
01The Sameness Problem: When Everyone Has the Same Mind
The evidence
In 2024, Anil Doshi and Oliver Hauser published a preregistered experiment in Science Advances. 293 writers wrote short stories; some could request story ideas from GPT-4. 600 separate evaluators rated the results1. Writers with access to up to five AI ideas produced stories rated 8.1% higher on novelty and 9.0% higher on usefulness than those without AI. The gains were largest for the least creative writers1. But the AI-assisted stories were measurably more similar to each other. The authors describe it as a social dilemma: individuals are better off, while collectively a narrower range of novel content is produced1.
Similarity between stories: higher ↑ (individuals better off; collectively a narrower range).
293 writers; 600 evaluators; writers with access to up to five AI ideas. Source: Doshi & Hauser, Science Advances (2024) 1
In plain words: AI raises the floor and narrows the range.
Why this matters commercially
Every company that adopts AI gets better copy, better design and better campaigns than it had before. Each one feels it has improved, and it has. But competitors improved in the same direction, with similar tools trained on similar data. The market becomes a sea of good, polished, interchangeable communication.
Agents sharpen the problem. When thousands of businesses describe themselves in the same polished words, an agent can only choose on hard data: price, availability, reviews, consistency. Everything else becomes noise.
A second, quieter risk sits underneath. Research published in Nature in 2024 showed that AI models trained recursively on AI-generated data degrade, a process the authors call model collapse2. Original human input is not only valuable to buyers. The machines need it too.
Truth is becoming a requirement
The US Federal Trade Commission’s final rule of 14 August 2024 bans fake consumer reviews and testimonials, explicitly including AI-generated fake reviews, and also bans buying or selling fake indicators of social media influence such as bot-generated followers or views3. A tribunal in British Columbia held Air Canada responsible for a refund policy its chatbot had described to a customer4. Verifiable truth is moving from a marketing choice to a legal and operational requirement.
What to do about it
- 01Start from your own point of view. Write down what you believe that competitors do not. Use AI to express it, not to invent it.
- 02Use your real material: your cases, your numbers, your stories, your mistakes.
- 03Keep your edges. The specific, human parts of a brand are what an algorithm smooths away.
- 04Check for sameness. Put your website next to your top five competitors. If you could swap the logos and nobody would notice, you have a problem.
A note on emotion
Outrage has been one of the most efficient ways to buy attention online; a 2017 study in PNAS found that moral-emotional language makes social media posts spread further5.
Download the PDF. Give it to your agent.
Have your agent study our white papers alongside you.
Free download
Want the full version?
Leave your email and we’ll send it.
Thanks, you’re on the list.
We’ll send it to .
Adapted from the Monarchy white paper “You are not talking to a human” (2026) by Andrea d’Agostini. Sources are listed below. Statements marked as views are opinions.
Sources
- Doshi, A. R. and Hauser, O. P., “Generative AI enhances individual creativity but reduces the collective diversity of novel content,” Science Advances 10, eadn5290 (2024). https://www.science.org/doi/10.1126/sciadv.adn5290
- Shumailov, I. et al., “AI models collapse when trained on recursively generated data,” Nature (2024). https://www.nature.com/articles/s41586-024-07566-y
- US Federal Trade Commission, “Federal Trade Commission Announces Final Rule Banning Fake Reviews and Testimonials,” 14 August 2024. https://www.ftc.gov/news-events/news/press-releases/2024/08/federal-trade-commission-announces-final-rule-banning-fake-reviews-testimonials
- Moffatt v. Air Canada, 2024 BCCRT 149 (Civil Resolution Tribunal of British Columbia). https://www.canlii.org/en/bc/bccrt/doc/2024/2024bccrt149/2024bccrt149.html
- Brady, W. J. et al., “Emotion shapes the diffusion of moralized content in social networks,” PNAS (2017). https://www.pnas.org/doi/10.1073/pnas.1618923114
