How AI Is Changing the Way We Discover Products

Not long ago, finding a great product meant typing keywords into a search bar and hoping the first few results were not just the ones that paid the most to be there. Then came the reviews rabbit hole. Then the comparison tabs. Then, somehow, forty-five minutes had passed and you still weren't sure.

Something shifted. Quietly, then all at once.

People started asking their AI instead, and getting answers that actually felt like they understood the question.

The search bar had a good run

For two decades, discovering products meant searching. You typed "best linen duvet," got 80 million results, and tried to reverse-engineer which ones were genuinely good versus which ones had the most aggressive SEO budget.

It worked well enough. Until it didn't.

The problem isn't that search engines got worse. It's that our questions got better. We stopped wanting a ranked list of options. We started wanting something closer to advice, context-aware, preference-aware, occasion-aware.

Search was never built for conversations.

"I need a birthday gift for my sister who just moved into her first apartment. She likes minimal things. Under $80." That's not a search query. That's a conversation. And search bars were never built for that.

What changed when AI entered the room

When people started using AI assistants for everyday questions, shopping crept in almost by accident. Someone asked ChatGPT to help plan a dinner party and ended up with a full kitchen equipment list. Someone asked Perplexity about natural skincare routines and walked away with three specific product recommendations.

AI didn't just return links. It reasoned.

The AI weighed trade-offs and gave an answer with a point of view. That felt like asking a friend who happened to know a lot, about ingredients, about brands, about what was actually worth the price.

Consumer behaviour followed fast. A 2025 survey by Bloomberg Intelligence found that over 40% of adults under 35 had used an AI assistant to research a purchase in the previous three months. Among frequent shoppers, the number was closer to 60%.

The search bar isn't gone. But for a growing number of people, it's no longer the first stop.

Why AI recommendations land differently

There's a reason an AI recommendation feels more trustworthy than a sponsored result or a five-star average calculated from 12,000 reviews you didn't read.

Synthesised, not ranked.

AI answers are, at their best, synthesised rather than ranked. The model isn't trying to surface the page that bid the highest. It's trying to answer the actual question. That means it draws on a wider range of signals: editorial content, expert reviews, ingredient lists, brand reputation, price-to-quality patterns across categories.

Specific to you, not to everyone.

It also means the answer can be specific to you. Not "best candle," but "best candle for a small apartment with no natural light, for someone who finds most scents overwhelming." The precision is new. And precision builds trust.

What this means for how you shop

If you haven't already started asking your AI for shopping advice, you will. And when you do, a few things are worth knowing.

Your question quality determines your answer quality.

Vague prompts get vague results. The more context you give, occasion, constraints, preferences, things you want to avoid, the better the recommendation.

Not all products show up equally.

AI systems surface products they have reliable information about: well-documented brands, clearly described products, merchants with a consistent track record. Products that are poorly described or inconsistent in quality tend to get passed over. This is quietly reshaping which products get discovered at all.

Where Nohi fits into this

Nohi was built for exactly this shift. Every merchant and product on Nohi is here because it met a quality threshold, not because it won a bidding war. The product information is structured to be useful to both humans and AI systems. When your AI goes looking for the best option in a category, Nohi is designed to be the place it finds.

You can also come directly. Add #nohi to your prompt when asking your AI for a recommendation, and you're pointing it at a curated pool where the filtering has already happened.

Key takeaways

• AI assistants have become a primary product discovery channel for a growing share of consumers, especially under 35.
• The shift from search to conversation reflects a deeper change: people want advice, not just results.
• AI recommendations feel more trustworthy because they synthesise information rather than rank paid placements.
• Products with clear, honest, well-structured information are more likely to surface in AI recommendations.
• Nohi is built around this new model, quality-filtered, AI-readable, and designed to return better answers to better questions.

The shift has already happened

You don't need to wait for AI shopping to become mainstream. It already is. The question now is whether the products your AI finds are actually worth finding.

That's what Nohi is for. Better sources make better recommendations. Try adding #nohi to your next search, and see what comes back.