Testing The New York Times Cooking AI Recipe Modifier
An evaluation of how AI-powered recipe customization impacts home cooking and user experience.
1AI Integration Controversy
The New York Times Cooking section recently introduced an AI-powered recipe modifier in an iOS-only beta. This feature allows subscribers to customize ingredients and equipment, sparking significant backlash from users. Within four days of the announcement on October 23, the platform received over 172 angry comments, eventually leading to the disabling of comments.
2Functionality and User Control
The tool appears above the ingredient list, enabling users to swap items or adjust for dietary needs. It draws from the Times' own substitution guides, user comments, and expert knowledge. Crucially, the system is designed as a helper rather than a primary chef, with editors having established the initial rule sets for modifications.
3Technical Implementation Details
Modifications are clearly highlighted in red-orange text throughout the recipe, ensuring transparency. Users can toggle between their customized version and the original recipe at any time. If a request deviates too far from the original spirit, the AI redirects the user to a more suitable recipe rather than attempting an impossible transformation.
4Impact on Recipe Quality
Testing revealed that the AI successfully adjusts cooking times and techniques based on ingredient changes. For example, when swapping chicken thighs for breasts, the app correctly reduced simmering time and added butter to the roux to compensate for lower fat content. These adjustments were integrated seamlessly into the step-by-step cooking instructions.
5Limitations and Texture Issues
Despite impressive logic, the AI struggled with specific culinary nuances. In one test, it suggested simmering shredded chicken breast for an additional 30 minutes, which negatively impacted the texture. This highlights the risk of AI-generated instructions potentially undermining the quality of high-end, tested recipes created by professional food writers and editors.
6Versatility in Kitchen Tasks
The tool proved effective for various tasks, including converting waffles to a yeasted version and modifying grain bowls. It successfully scaled ingredients and provided precise instructions for adding vegetables to stews. These capabilities demonstrate how the technology can assist with pantry management and dietary preferences without requiring a trip to the grocery store.
7Preserving Culinary Voice
A notable success of the system is its ability to maintain the original author's voice during modifications. Even after significant ingredient swaps, the poetic instructions of the original recipe creators remained intact. This suggests that the AI is effectively layering functional changes onto the established, high-quality content produced by the Times' editorial team.
8Editorial Team and Strategy
Despite concerns regarding automation, NYT Cooking management emphasizes that they have no intention of reducing their staff of 60 editors and writers. The company continues to hire, viewing the AI as a tool to enhance the utility of their existing content library rather than a replacement for human recipe development and curation.
9Addressing User Feedback
The Times has responded to subscriber concerns by planning to allow users to hide the AI feature within the app. By keeping the tool in an iOS-only beta, the company maintains a measured approach, prioritizing user experience and feedback before considering a wider rollout to other platforms or a full-scale public release.
10The Human-Tech Balance
The resistance from the community likely stems from a fear of losing the human connection that defines the brand. While the AI provides practical utility for substitutions, it remains a subtractive process. The challenge lies in ensuring that these technological conveniences do not dilute the expertise and design intent of professional recipe creators.
11Future of Digital Cooking
This experiment represents a rare instance of technology incrementally improving a product without overwhelming it. By focusing on ingredient substitutions and equipment constraints, the Times is solving real-world problems for home cooks. The long-term success of this feature will depend on balancing these automated conveniences with the essential human touch of culinary expertise.
Original article: Wired — Testing The New York Times Cooking AI Recipe Modifier