I have been in the industry for 20 years, previously working as an internet product manager. In 2023, due to the global economic downturn, I was laid off and found it difficult to secure a satisfactory job. To support my family, I decided to leverage my skills. I designed my own app, and with a technical background, I taught myself popular programming technologies such as uni-app (a front-end framework) and Python. I wrote 100% of the front-end and back-end code myself and used AI tools to design the visual assets of the product.
In March 2024, I successfully released an application called Mia (com.jacky.her) on the App Store. Mia connects to a large language model to chat with users and provide emotional support. I noticed there were already many similar applications on the market, but most of them were just simple chat boxes. I believe that future product competition will focus on more detailed features based on large models. Therefore, I designed a character prototype, added character mechanisms, and aimed to surpass these competitors. After launch, with limited promotional efforts, I found that although I could barely break even, it was still difficult to achieve scalable profitability, and this model mainly attracted users driven by hormonal impulses.
After a series of considerations and iterations, I prepared to release version 2.0 of Mia in June this year. Compared to 1.0, it includes original storyline progression, no longer just a chat mode, and introduces a gamified card-drawing mechanism to advance character skills. Simply put, it is more like a nurturing game. However, when submitting version 2.0, Apple rejected my application citing "Guideline 4.3(a) - Design - Spam." Several appeals were unsuccessful, leading me to request a phone call with Apple.
Although Apple called me a few days later, I was deeply puzzled by the outcome of the conversation. Here are the main points raised by Apple's review staff:
1.Apple's Review Standards Are Constantly Changing
The reviewer mentioned that the standards are always changing, and the rejection this time was due to "too many similar applications," not a quality issue. However, Mia's 1.0 version passed smoothly, but after adding more unique features, it was rejected, citing the presence of too many similar applications on the market.
2.Lack of Specific Standards and Feedback
The reviewer repeatedly mentioned that my application was "unremarkable," but could not specifically explain what constitutes "unremarkable," nor clearly pointed out in what aspects my application was similar to others. I repeatedly asked for clear standards or next steps for modification but received no specific answers.
3.Apple's Business Strategy
The reviewer finally mentioned that Apple is a private platform, and this is their current business strategy, which does not accept applications like mine. This strategy is not based on quality or uniqueness but is an internal decision by Apple, which is very disappointing for developers.
4.Compared to Other Similar Applications
I researched similar applications on the App Store and found that there are indeed many simple chat applications with rough quality and details. My application features specific character images, unique storyline development, and a gamified card-drawing mechanism, yet it was still rejected. The reviewer suggested I report other applications, but this was of no substantive help.
5.Inability to Disclose Information on Similar Applications
When I asked the reviewer to give examples of applications similar to mine, he said he could not disclose information about other applications. This means that Apple's judgment of my application cannot provide any substantive evidence, and everything is at Apple's discretion.
The above content reflects the actual statements made by Apple's review staff. Apple has recordings of these conversations as well. Although the communication was in Chinese, I have summarized it in English.
I believe the rise of large models will reshape products in many industries. My application is essentially a nurturing game with rich plot and scene settings. Apart from using third-party interfaces for the large model part, all other aspects are independently developed. The interface design and character materials are AI-generated. However, merely integrating large model dialogues led to it being labeled as a spam application. Does this mean that any application utilizing large models cannot pass the review?
As a product manager, I have always been a fan of Apple and Steve Jobs. I started using a Macbook Pro for work back in the iPod era, and I replace my iPhone every year. However, this experience has left me deeply disappointed. Perhaps you do not care about a single consumer or developer, but I cannot accept that nearly a year of effort was dismissed with a single word, "spam."
Goodbye, Apple.