In a recent LinkedIn post, 🌀 Luca Rossi sounds a note of caution regarding the widely discussed productivity gains from artificial intelligence, suggesting that many perceived improvements might be illusory. Rossi highlights a critical gap in current research: the heavy reliance on self-reported data, which can be misleading.
According to 🌀 Luca Rossi, while individuals might *feel* more productive when using AI tools, actual output doesn’t always align with this perception. He points to a study (the METR research) as a potential counterexample, which, using control groups, indicated that participants who believed they were faster with AI were, in reality, slower.
“Beware AI productivity “placebo”… most data is just… self reported. And engineers can indeed *feel* more productive, but the only study (the METR one, AFAIK) using proper control groups found people *thought* they were faster while actually being slower.”
The ‘Fake State of Flow’ with AI
🌀 Luca Rossi draws an analogy to the experience of getting stuck in a bug-fixing loop. He explains that the quick feedback cycle of making a change and immediately checking the result can create a ‘fake state of flow.’ This state feels engaging and productive because it’s on autopilot, but it doesn’t necessarily lead to genuine progress or deep engagement.
This phenomenon, Rossi argues, can be replicated when working with AI. The interaction can feel highly productive, yet the actual work output might be less efficient than perceived. As 🌀 Luca Rossi puts it:
“In other words: AI can trigger the *feeling* of progress without real output.”
Strategies for Genuine AI-Assisted Productivity
To combat this potential ‘AI productivity placebo,’ 🌀 Luca Rossi outlines a set of practical work hygiene strategies aimed at fostering genuine engagement and efficiency:
1. Maintain Active Engagement
🌀 Luca Rossi emphasizes the importance of actively understanding AI’s processes. This involves providing feedback and steering the AI’s direction rather than passively accepting its outputs. As 🌀 Luca Rossi advises:
“Stay engaged — understand what AI is doing, provide feedback, steer actively.”
2. Avoid Multitasking During Critical AI Work
The post strongly recommends against multitasking when deeply engaged with AI. Rossi suggests two approaches: either allow the AI agent to run autonomously while attending to other tasks (like a meeting), or maintain focus in short, iterative bursts. He argues that frequent context switching is detrimental to performance, especially when working with AI.
3. Incorporate Regular Breaks
Finally, 🌀 Luca Rossi advocates for structured breaks, suggesting a Pomodoro-like technique (e.g., 5 minutes of rest every 30 minutes). This practice is intended to prevent cognitive fatigue and mental ‘brain rot’ by forcing a reset, thereby maintaining a higher level of focus and effectiveness.
By implementing these strategies, professionals can aim to harness the true potential of AI tools, ensuring that perceived productivity translates into tangible results, according to 🌀 Luca Rossi’s analysis.
📝 About This Content
This article is based on insights shared by 🌀 Luca Rossi on LinkedIn.
📅 Originally posted on February 9, 2026 | View original post on LinkedIn →