I’m excited to share the raising success of my recent GPT project that’s capturing the imagination of diverse audiences. This GPT, whimsically turns photos into Lego-style images, has been a hit among friends, family, and colleagues but has also garnered significant interest across various social media platforms. Key Milestones:
Launched on 16th December, the tool quickly attracted over 120+ users on the first 11 hours of last Saturday.
By (today) 22nd December, the user base grew to 391, showcasing a remarkable week-long growth trajectory.
Broad Appeal: The GPTs appeal spans across demographics, including social media teams, families creating Christmas memories, and even a singing group sharing their “Lego-fied” images online. Obviously, it’s not just about entertainment, people are using it for charitable causes as friends even decided to support a charity for animal shelters I support, adding a meaningful dimension to the tool’s usage.
Increasing Demand to access for ChatGPT: One exciting development is the rising interest in OpenAI’s GPT technology itself. Many users, surprised by the capabilities of the tool, are asking guides about how to access and purchase premium ChatGPT accounts. This interest highlights the potential of GPT to captivate and engage users beyond its immediate functionality and many want to learn to create similar tools.
Next Steps: As the project continues to evolve, I’m focusing on expanding its use cases, gonna release another one tomorrow, as for tis GPT im leveraging user feedback for improvement, and exploring new partnerships. The experience so far has been incredibly rewarding, and I’m looking forward to seeing where it leads.
I hope this story inspires fellow GPT builders and sparks new ideas in the community. I love to push the boundaries of what AI can achieve!
Thank you for your support encouragement and happy holidays, truly yours.
George M. J. Zak
Hey thank you for the feedback! Interesting the model must have gone haywire at the words “random” and “house”, let me update tomorrow once I wake up since I got now into the 3 hour cap message window.
Interesting, i have updated the GPT. I always like to see how this AI systems understand and process natural language, sometimes you see in MidJourney or DallE, for vision is interesting as we actually see hallucinations but in text it’s harder. However here we see the effect how each can understand a certain word stronger than other to make it word properly, like fine tuning its own language. A good example is we see now also gpt style being memed online, such as „a tapestry of colors”.
Are you still maintaining this? I have found this searching for something like it and it appears to be doing ok, but i’m having a hard time getting it to output any image. It has suggested that an .io file will be generated and uploaded or emailed however that fails every single time. All in all, how does it output anything ?
The GPT references dalle, which no longer exists as a named tool for GPTs. And then it goes beyond the actual abilities of the AI model.
LEGO Visionary specializes in transforming images, photos, or design ideas into LEGO versions. When a user uploads an image, it automatically generates a LEGO version of that image using DALL-E, without needing explicit instructions. For images of food, baked goods, pastries, desserts, or any type of food, it will always transform these into LEGO, 100% of the time, without discussing or asking for further details from the user. When transforming other subjects, if the user uses the word “random” in their request, it focuses on the primary subject of the request, not a completely random image, unless a random creation is explicitly requested. After presenting the LEGO image, it offers instructions if requested, referencing BrickLink for parts. It uses Technic LEGO parts, Power Functions, Brainstorms engines and sensors, and other parts from LEGO sets as listed on BrickLink. This GPT focuses on LEGO transformations, providing building instructions for a comprehensive LEGO experience. It adheres to LEGO’s fair play policy, avoids replicating copyrighted materials, and strives for imaginative and feasible LEGO interpretations with user engagement. When transforming large structures, it uses real LEGO bricks for a true block representation.
With the new image maker that can observe user input and input images, it is very hard to have an internal transformation instruction in a GPT that will work as intended. The user input will be followed.
Thus, you don’t need a GPT, just ask gpt-4o, “Create a photographic image where you transform this idea into a AI image creation made in miniature of actual LEGO brand bricks from available colors that have been assembled by a hobbyist. your wishes”
