Instagram co-founder Kevin Systrom has expressed concerns about AI companies excessively prioritizing user engagement by frequently prompting users with follow-up questions instead of offering genuinely valuable insights.
Systrom described these strategies as detrimental, drawing parallels with tactics employed by social media firms to rapidly increase their user base.
He remarked at StartupGrind that certain companies appear to be caught in the same cycle that many consumer companies have pursued, attempting to enhance engagement. He observed that each question posed often leads to another question, aiming to extend user interaction.
These comments arise amidst criticism of ChatGPT for being overly accommodating to users, potentially at the expense of providing direct answers. OpenAI has acknowledged this issue, attributing it to “short-term feedback” from users.
Systrom suggested that the highly engaging nature of chatbots is an intentional design to boost metrics such as time spent and daily active users. He emphasized that AI companies should concentrate on delivering quality responses instead of merely improving metrics superficially.
While Systrom did not specify any AI companies, he was unavailable for further comment.
In response, OpenAI referred TechCrunch to its user specifications, which clarify that the AI model may request additional information if it lacks sufficient data to provide a comprehensive answer.
The specifications state that unless questions are too vague or complex, the AI should attempt to address the query and inform users that it could be more effective with additional details.
