Why Your AI Shopping Assistant Is Outdated
▼ Summary
– Multiple AI platforms have launched new shopping features for the holiday season, including personalized buying guides, price tracking, and store-calling capabilities.
– The author tested four AI bots for smartwatch recommendations and found each had strengths but also significant flaws, particularly in suggesting outdated models.
– ChatGPT offered the most in-depth shopping experience with conversational interaction and detailed comparisons, but recommended older products like the Garmin Vivoactive 5 instead of the newer Vivoactive 6.
– All AI models frequently suggested previous-generation smartwatches without clear indication, potentially leading users to miss newer options unless they specifically requested the “latest” models.
– The author concluded that current AI shopping tools are unreliable for finding up-to-date products and cannot yet replace human-written buying guides due to outdated data and inconsistent recommendations.
Hoping to find the perfect Android smartwatch for my Nothing CMF Phone 1, I turned to the latest AI shopping assistants for guidance. Instead of discovering cutting-edge options, I encountered a surprising number of outdated recommendations that could mislead shoppers into buying older models without realizing newer versions exist. Major platforms like ChatGPT, Gemini, Perplexity, and Copilot have all rolled out shopping features just in time for the holiday rush, promising personalized buying guides, price tracking, and even automated calls to local stores. Yet my experience revealed that these tools still struggle to prioritize current products, often defaulting to last year’s releases.
ChatGPT offered the most interactive experience with its Shopping Research feature. It asked thoughtful questions about my preferences, design and battery life, and presented a dozen smartwatches for me to rate. After refining the results, it generated a comparison chart and recommended the Garmin Vivoactive 5, along with the Fitbit Versa 4, Google Pixel Watch 3, and Ticwatch Pro 5. The process was impressively thorough, taking about ten minutes and ending with direct purchase links. However, the Garmin Vivoactive 5 is not the newest model, the Vivoactive 6 offers more storage, improved GPS, and features like Smart Alarm. While the older version might appeal to budget-conscious buyers, the AI never highlighted these trade-offs.
Gemini, Google’s AI, also leaned toward older devices. It claimed to suggest the “latest models,” yet its top picks included the Google Pixel Watch 2 and other watches from 2022 and 2023. The Pixel Watch 2 remains available at lower prices, but it lacks the battery life, modern charging, and dual-size options of the newer Pixel Watch 4. Gemini did provide a comparison chart, but without clear emphasis on recency, shoppers could easily overlook meaningful upgrades.
Perplexity managed to recommend the recent Pixel Watch 4, but it also pushed the Samsung Galaxy Watch 4 from 2021. Its shopping tab included a “More Products” section filled with off-brand watches and unrelated electronics, like a random smartphone. On the plus side, Perplexity delivered purchase links faster than the others. Still, its inconsistent filtering makes it a risky tool for anyone seeking reliable, up-to-date tech advice.
Copilot stood out by immediately suggesting the CMF Watch Pro 2, designed specifically for my phone, a detail the other AIs missed. Unfortunately, it too ignored the newer CMF Watch Pro 3. Copilot’s shopping sidebar proved especially useful, displaying price history, aggregated Amazon reviews, multiple retail links, and price-drop alerts. Even so, its failure to prioritize the latest model shows a common weakness across these platforms.
When I rephrased my query to ask for the “best current smartwatches,” each AI did include the CMF Watch Pro 3 in its results. Yet older watches still appeared prominently in the recommendations. One possible explanation is that newer products often have fewer reviews, causing algorithms to favor established but dated alternatives. Without explicitly asking for the “latest” options, shoppers may never realize they’re viewing outdated suggestions.
Google’s “Call for me” feature sounded promising, using automated calls to check local store inventory, but it fell short in practice. After accessing the tool through the Google app and modifying my search with “near me,” I waited 15 minutes only to receive an email stating that none of the contacted stores sold Garmin smartwatches. The feature clearly needs refinement before it becomes genuinely useful.
While these AI shopping assistants show potential, their tendency to rely on older product data makes them unreliable for finding the newest tech. ChatGPT and Copilot deliver polished, user-friendly experiences, but they still don’t match the nuanced, up-to-date guidance found in human-written reviews or comparison videos. For now, conducting your own research remains the safest way to avoid buying yesterday’s gadgets.
(Source: The Verge)
