Why Todoist says less AI can deliver more

Todoist’s leadership is openly stating that they will not chase every new AI trend. Instead, they plan to prioritize features that do not depend on AI, which simplifies the architecture and reduces the need for continuous model updates. This decision reflects a broader caution about the volatility of AI research and commercial readiness.

For teams building or maintaining Todoist, the reduced AI footprint translates into fewer dependencies on external ML services and less infrastructure for GPU-based workloads. The cost of adopting new AI components, such as data pipelines, model monitoring, and compliance checks, will be avoided, allowing resources to be allocated to core product work like UI enhancements or performance tuning.

The trade-off is that any functionality that could have been powered by AI, such as smart task suggestions or natural-language parsing, will either stay static or be omitted. Engineers should expect that existing AI-based features will not receive further investment, and new AI experiments may be deprioritized. This limits the product’s ability to offer cutting-edge AI experiences but keeps the system more stable.

The CTO’s admission of not knowing where AI is headed underscores the difficulty of planning long-term around a rapidly shifting technology landscape. By acknowledging this uncertainty, Todoist signals that its roadmap will be driven by concrete user needs rather than speculative AI capabilities. Teams can therefore focus on measurable outcomes rather than speculative AI performance gains.

Overall, the shift means that Todoist’s engineering effort will concentrate on reliability, scalability, and user-centric improvements rather than on maintaining a complex AI stack. This approach may appeal to developers who prefer predictable, maintainable code over the volatility of AI model lifecycles, while users may notice fewer AI-driven features but potentially a more polished core experience.