The launch of the OpenAI GPT Store was considered a revolutionary moment in the field of artificial intelligence. Many predicted that it would disrupt the App Store model and create a new ecosystem of apps. However, since its launch, OpenAI has faced challenges in attracting developers to the platform, and the platform has been inundated with low-quality content. Developers are also finding it challenging to attract users due to limited traffic and poor hands-on experience of the GPT Store.
It is not what it appears to be
At OpenAI’s inaugural developer conference last November, CEO Sam Altman introduced GPT as a tool for accomplishing a wide range of tasks, from programming and learning obscure information to receiving guidance on scientific exercises.
As it happens, Altman’s assessment was accurate. Regrettably, the phrase ‘almost any scenario’ does not accurately reflect reality. Currently, OpenAI’s GPT Store is inundated with spammy, peculiar, and potentially infringing GPTs.
Despite OpenAI’s claim of having over 3 million tailored chatbots at the GPT Store’s launch in January, some developers report that the number of users utilizing the bots has not met expectations. Demochkin, a developer, found that only about 5% of the 36,000 custom chatbots he analysed received between 150 and 500 active users per day. Most chatbots, however, only attract one or two users per day.
This is partly due to the $20-per-month paywall behind the feature; OpenAI requires that only paid versions of ChatGPT customers have access to the GPT Store.
OpenAI should consider implementing a two-way incentive scheme, similar to those used by tech giants such as Apple and Google, to provide developers with a better return. Additionally, OpenAI’s current workload may be hindering their ability to focus on the successful implementation of this new model and addressing spammy content. It is important to establish a mature framework to standardise operations and ensure that efforts are not spread too thin.
In Disarray
Currently, a basic search of the GPT Store shows numerous GPTs that purport to imitate the style of Disney and Marvel productions. Furthermore, the creators of these GPTs openly acknowledge that the underlying service remains OpenAI, and that they are simply able to circumvent AI content detection tools like Turnitin and Copyleaks. This raises concerns about the effectiveness of the company’s auditing department.
One may question the effectiveness of the company’s auditing department. The GPT Store has experienced rapid growth due to low barriers to entry; OpenAI reported having approximately 3 million GPTs in its store as of January. However, this growth appears to have occurred at the expense of quality and OpenAI’s policy terms.
The GPT Store offers several GPTs that are based on popular movie, TV, and video game franchises, but they are not developed or licensed by the franchise owners.
OpenAI has a policy that explicitly prohibits developers from creating GPTs that involve academic fraud. However, the GPT Store currently sells plagiarized content to educators, bypassing AI content detectors with impunity.
OpenAI prohibits GPT developers from creating GPTs that impersonate individuals or organizations without their consent or lawful authorization. However, the GPT Store still contains such GPTs that claim to mimic each other’s tone of voice and personality.
Doomed to fail?
The present condition of the GPT Store implies that the pursuit of monetization could result in further complications. OpenAI has pledged that GPT developers will eventually have the ability to earn money based on the number of GPT users, and may even offer subscriptions to individual GPTs. However, if unauthorized developers profit from GPTs themed around Marvel or Lord of the Rings, how should the original authors respond?
OpenAI has no qualms about launching a GPT Store. The business model of the Apple App Store has been lucrative and enviable in the past. OpenAI aims to replicate this experience by hosting, developing, evaluating, and promoting a wide range of GPT products on its own platform.
Despite actively discussing the importance of governance and safeguards, OpenAI seems to have overlooked these obvious pitfalls. However, it appears that this is not the case, and the current GPT Store is disorganized. Without prompt action, this issue will persist for an extended period.
