The AI-powered tool, ChatGPT, has experienced a noticeable dip in its user base for the third consecutive month as of August. However, there are indicators suggesting that the start of the academic year could potentially reverse this trend.
ChatGPT, OpenAI’s groundbreaking tool, made waves when it launched last November, amassing over 100 million users in a short span. Despite its initial success, statistics from Similarweb reveal that the duration of user engagement on the site slipped from 8.7 minutes to seven minutes between March and August.
June and July witnessed nearly a 10% decrease in active users each month. However, August saw a relative stabilization, with only a 3% drop in global users. Interestingly, US-based traffic recorded a slight increase of 0.4%. This pattern led analysts to speculate that the summer vacation period for students might have contributed to the reduced engagement.
ChatGPT primarily engages college-age individuals, who comprise over a quarter of the adult users, according to Similarweb’s August report. David F. Carr from Similarweb noted, “The percentage of younger users of the website dropped over the summer and is now starting to bounce back.”
A survey by Intelligent.com involving 1,223 students revealed that 30% had used ChatGPT for their academic work during the 2022-2023 academic year. However, this widespread usage among students has raised concerns among educators, prompting some to implement AI detection software to ensure originality in assignments.
The University of Michigan Law School even introduced guidelines discouraging applicants from using AI tools like ChatGPT during the drafting process.
While ChatGPT’s rapid growth marked it as one of the most quickly adopted apps, its perceived decline in accuracy may be contributing to its dwindling user base. A study by Stanford University and UC Berkeley found that GPT-4’s ability to identify prime numbers dropped from 97.6% accuracy in March to a shocking 2.4% in June.
However, the researchers cautioned that further studies are needed to understand these fluctuations, noting that “the behavior of the ‘same’ [large language model] service can change substantially in a relatively short amount of time.”
Moreover, the chatbot market is becoming increasingly crowded, with new competitors emerging regularly. For instance, Anthropic recently introduced a paid tier for its Claude chatbot, directly competing with ChatGPT’s $20/month price point. Whether these factors will continue to impact ChatGPT’s user base remains to be seen.