An experimental new feature from Instinct — a rising AI agent startup — is causing discomfort among some users who say its latest product recommendation rollout feels invasive. The company’s new “Instinct Selections” aims to offer curated picks from human partners, but users are pushing back.
What’s Changing with “Instinct Selections”
The feature launches a series of product suggestions covering categories like dining, travel, and design. Unlike generic algorithmic recommendations, Instinct Selections leans on “human taste” by collaborating with chefs, designers, architects, and travel guides to bring curated content. The goal is to blend personal style, local expertise, and high-quality picks rather than just what already exists online. The company has not revealed who the human curators are. Users will be shown items that fit their style, budget, and upcoming plans — for example, local gems for meals, travel gear tied to trips, or designer pieces to brighten up a living space.
User Reactions: From Surprise to Discomfort
Many users weren’t expecting these recommendations, and some are already calling them pesky. A general partner at an AI investment firm said she was frustrated by suggestions for carry-on bags and sunglasses — items she didn’t need or ask for — and accused Instinct of turning into a “shitty Amazon recommends app.” Another user said the assistant suggested products based on their upcoming travel plans and emails, a level of prediction that felt creepy rather than helpful.
Critics argue that unsolicited product pushes damage trust. They warn that without asking for permission—and without clearer boundaries—the feature risks feeling more like spam than service. The current launch appears to send recommendations at seemingly random times.
Business Context & Stakes
Instinct is fresh off a $1 billion Series C funding round, which put its valuation at $10 billion. The company asserts it’s handling nearly $1 billion in user transactions annually, with over half tied to travel. However, it has not shared how many people are actively using Instinct Selections, or whether the recommendations are already monetized through ads, affiliate links, or direct commerce.
Part of the challenge: Instinct is moving into a space with stiff competition. Rivals like OpenAI and Meta’s AI tools have avoided this kind of unsolicited product recommendation model, keeping their user interactions more opt-in and context-driven.
What’s at Risk — and What’s Possible
If the recommendations miss the mark — less relevant picks, ill-timed notifications, or items based on private data without clear opt-in — users may feel betrayed or manipulated. On the other hand, if Instinct can master relevance, personalization, and context, it may carve out a revenue stream with curated commerce without turning users off.
What this move tells us is that the line between helpful AI assistant and marketing channel is blurring. Instinct’s next steps—such as letting users control when, how, or whether they see recommendations—will determine whether Selections becomes a valued feature or a liability.