Lamine
Lahouasnia
Director of Gen AI at Euromonitor International, in Singapore.
I build AI that market researchers and their clients can trust, and I write about the gap between an answer that looks right and one that is right.
Ideas I’ve named
Most of my writing names a way AI goes wrong in research, or a way to fix it. Each one links to the piece that introduced it.
- The Plausibility Trap
- An AI answer that looks well researched enough to bet on, and falls apart at the first question about method.From “I Asked AI a Market Research Question. The Answer Was Dangerously Wrong...”
- The Digital Yes-Man
- AI trained to agree with you, delivering confirmation bias as a service.From “AI is Trained to Make You Feel Right. A Market Researcher's Job is to Help You Be Right.”
- The Slot Machine Effect
- Pull the same prompt twice and get a different market. Without reproducibility, it is anecdote, not research.From “Did the Market Change, or Did the AI Just Blink? The Crisis of Consistency in AI Desk Research.”
- The Symmetry Tax
- AI squeezing messy reality into tidy lists of three, deleting the outliers where insight usually hides.From “The Symmetry Tax: How AI's 'Rule of Three' is Costing You Real Insight”
- Insight Engineers
- Researchers who build with AI, not just use it: domain expertise plus strategic fluency.From “Forget 'AI Fluency'. Market Research Needs 'Insight Engineers'.”
- The Action Layer
- The shift from what AI knows to what it can touch: browsers, files and the work itself.From “My 3 AI Predictions for Market Research in 2026”
- Research versus retrieval
- Pay agents to research. Use a data pipe for the facts, instead of paying them to click around for it.From “The Economics of Agentic AI: Research versus Retrieval”
- Institutional skills
- Codifying a firm’s methods once, so every colleague’s AI works the way the firm does.From “The Reliability Gap Has Closed. Why We're Hiring an AI Knowledge Architect.”
Latest writing
All 10 articles- The Reliability Gap Has Closed. Why We're Hiring an AI Knowledge Architect.
Frontier models are now reliable enough to carry institutional processes. The advantage goes to organisations that codify their expertise into governed, reusable skills.
- The Economics of Agentic AI: Research versus Retrieval
Even flawless agents have an economics problem. Pay for research, not for agents clicking around to retrieve facts a data pipe could supply.
- My 3 AI Predictions for Market Research in 2026
2026 will be defined by the Action Layer: what AI can touch, not just what it knows. Three predictions for market research.
- Did the Market Change, or Did the AI Just Blink? The Crisis of Consistency in AI Desk Research.
Ask an AI the same sizing question five times and you can get five different markets. In research, consistency is trust.
What I’m building
Outside work I’m running a small, AI-native games studio. The Moth House is an eerie point-and-click puzzle game in the spirit of Rusty Lake. AI agents triage feedback, build and test from a work board, and I review and sign off each week.
It’s the most honest test I know of where AI genuinely helps, and where taste still has to come from a person. More about the studio