Limitations of AI Shopping Assistants in 2023

Recent evaluations of AI shopping assistants, specifically ChatGPT, Gemini, Perplexity, and Copilot, reveal a mix of capabilities and shortcomings in their ability to assist consumers in finding current tech products, such as smartwatches. The assessments were based on a user’s experience searching for a smartwatch compatible with the Nothing CMF Phone 1.

The Verge

ChatGPT emerged as the most conversational and comprehensive, utilizing its Shopping Research feature to engage users by asking about preferred features like design and battery life. It provided a list of recommendations, including the Garmin Vivoactive 5, Fitbit Versa 4, Google Pixel Watch 3, and Ticwatch Pro 5, along with a comparison chart. However, it failed to suggest the latest model, the Garmin Vivoactive 6, which offers significant improvements.

The Verge

Copilot also performed well by suggesting the CMF Watch Pro 2, tailored for the Nothing CMF Phone 1, but overlooked the newer CMF Watch Pro 3. Its shopping sidebar was noted for its utility, providing price history and aggregated reviews. Both AI models struggled to prioritize the most current products, often recommending older models due to a lack of recent reviews.

The Verge

The user experience with these AI assistants highlighted a critical issue: the reliance on outdated product data. For instance, Google’s “Call for me” feature, which automates inquiries to local stores, ultimately yielded no results for Garmin smartwatches after a 15-minute wait. This inefficiency raises concerns about the practical utility of these AI tools in real-time shopping scenarios.

The Verge

Critics argue that while AI shopping assistants may appear beneficial, they often fail to deliver relevant and timely recommendations. The tendency to suggest older models can lead consumers to miss out on newer options. The lack of specificity in queries, such as not indicating the need for the latest products, exacerbates this issue. As a result, many users may find themselves better served by traditional buying guides and reviews created by human experts.

The Verge

The current state of AI shopping assistants indicates that while they are making strides, they are not yet reliable enough to replace human-curated buying guides. The data they utilize often lags behind market offerings, making it challenging for users to make informed purchasing decisions. Until these systems can provide more accurate and up-to-date recommendations, consumers may prefer to conduct their own research when shopping for tech products.

The Verge

• “If you don’t know what you’re looking for, it could be easy to end up with an outdated product, or even just miss out on newer options the AI doesn’t realize exist.” — User Experience Reviewer

The Verge

• “They seem to be pulling from product data that’s a couple of years old at this point, while also confidently giving you recommendations.” — User Experience Reviewer

The Verge

• “For now, I’ll do my own smartwatch research.” — User Experience Reviewer

The Verge