OpenAI introduced the personalized chatbot marketplace ‘GPTs’ to create new business opportunities. However, after over two months, GPTs has failed to live up to CEO Sam Altman’s expectations. SimilarWeb data reveals that personalized GPT creations only accounted for 1.5% of ChatGPT website traffic in February, with weekly traffic remaining stable, indicating no sudden increases.
Nevertheless, certain chatbots, such as those claiming to predict stock prices and avoid plagiarism detection, have gained popularity in the marketplace. These chatbots have attracted millions of users, challenging the authenticity of academic papers and answers. Despite OpenAI’s restrictions, several of these tools collectively received at least 3 million views.
Some popular GPTs serve educational purposes, like ‘Consensus,’ which ranks third on the website as a tool for searching and summarizing academic papers. Others include image generators, language translators, and resume builders. However, media analysis suggests that many popular GPTs may violate OpenAI’s policies by offering unverified financial, legal, or medical advice.
Mixed Performance of the GPT Marketplace
One notable app, “Finance Wizard,” has been utilized over 200,000 times and claims to forecast future stock trends based on historical data. Altman has pledged to ensure that personalized GPTs comply with company rules before user access, using a combination of system scans, manual reviews, and user reports to identify policy violations. Additionally, users now have the option to rate and review GPTs. Despite efforts to enforce compliance, media analysis suggests that many popular GPTs may still violate OpenAI’s policies, particularly concerning the provision of unverified financial, legal, or medical advice. This raises concerns about the marketplace’s ability to maintain quality and user trust over time.
Altman has hinted at the possibility of top GPT creators earning income later this year, intending to attract more developers to create personalized chatbots. This approach aligns somewhat with Apple’s App Store model. While SimilarWeb data may be incomplete, as it does not include ChatGPT mobile app traffic, data.ai confirms that the appeal of personalized GPTs remains limited.
