AI Shopping: A Holiday Experiment Gone Wrong
Holiday shopping just got “smarter” with AI assistants, but are they really helpful? I set out to find a new smartwatch, eyeing models like the Google Pixel Watch 4 and Garmin Vivoactive 6. Instead, my AI helpers kept pushing watches from years ago, leaving me both amused and frustrated. With OpenAI, Google, Perplexity, and Microsoft launching new AI shopping tools, expectations were high—but the reality was a mix of clever insights and confusing misfires.
Four AI Assistants, Four Different Experiences
I tested ChatGPT, Gemini, Perplexity, and Microsoft’s Copilot for a simple task: find a good Android smartwatch for my Nothing CMF Phone 1. Each AI had its quirks. ChatGPT led with conversational guidance, asking about design and battery preferences, while Gemini offered to contact local stores. Copilot focused on price tracking, and Perplexity leaned heavily on older recommendations. The results highlighted that AI shopping is still evolving—and not always up to date.
ChatGPT’s Deep Dive Shopping Tool
ChatGPT’s new Shopping Research feature stood out for its depth. It generated a personalized shortlist, asked me to rate product suggestions, and compared specs side by side. The final recommendation? The Garmin Vivoactive 5, with handy links to current deals. While impressive, its reliance on slightly outdated data reminded me that even the smartest AI can miss the mark if its training cut-off isn’t current.
Gemini’s Local Store Advantage
Gemini impressed with real-world practicality. Unlike other AIs that only provided online options, Gemini offered to call nearby stores to check stock and availability. This felt surprisingly human and saved me time—but the AI still stumbled when recommending older watch models that no longer matched my preferences.
Copilot’s Price Tracking Power
Microsoft’s Copilot focused on what it does best: tracking deals. It flagged price drops and compared offers across multiple retailers. However, Copilot lacked the nuanced guidance of ChatGPT, making it more of a tool than a true shopping assistant. It’s ideal for deal hunters but less helpful for those seeking advice or recommendations tailored to their lifestyle.
Perplexity’s Hit-or-Miss Recommendations
Perplexity was the least reliable. Its watch suggestions often referenced outdated models or overlooked critical specs I had mentioned. While it sometimes provided useful links, the experience was inconsistent, reinforcing the idea that AI shopping tools are still experimental and far from perfect.
The Verdict: Impressive but Flawed
AI shopping assistants show enormous potential, particularly for research and deal tracking. But they’re not yet ready to replace human judgment entirely. Outdated suggestions, occasional misinterpretation of preferences, and inconsistent accuracy make them helpful but imperfect partners for holiday shopping—or any purchase where up-to-date recommendations matter.
Looking Ahead: AI Shopping in 2026
As AI models continue to improve, smarter and more current shopping guidance is on the horizon. Features like live store checks, personalized comparisons, and dynamic deal alerts hint at a future where AI could genuinely simplify purchases. For now, shoppers should treat AI assistants as powerful tools rather than infallible advisors.
Comment