Hi Mijael,

Thank you very much for your thoughtful comment and for taking the time to share your findings. I found your perspective very interesting, especially because it approaches a similar issue from the consumer side rather than the investment side.

What you describe does seem highly relevant to the mechanism I discuss in my piece. In the article, I focus on how LLMs may reinforce attention bias in investment recommendations by systematically favoring firms that already receive more coverage, discussion, and market visibility. Your observation suggests that a similar pattern may also operate one step earlier, through consumer recommendations that repeatedly direct users toward the same dominant platforms.

At the same time, I think there is also an important endogeneity aspect here. Although it can look like a kind of unpaid marketing channel, these firms are also often the ones that became publicly traded precisely because they were already highly successful, widely adopted, and deeply embedded in consumer activity. In that sense, the models may not only be creating attention, but also reflecting and amplifying patterns that already exist in the real economy. That feedback mechanism is what makes the issue especially interesting.

I also think your point about these recommendations functioning as an invisible user acquisition channel is very compelling. If models consistently steer users toward the same firms and platforms, that could have meaningful implications not only for competition and platform growth, but also for how attention gets translated into economic outcomes.

Thank you again for comment and for sharing such a thoughtful connection to my work.

Best regards,
Toghrul