A Shocking Revelation: AI Startup Relies on Humans for 3D Model Generation
The world of 3D modeling has made significant strides in recent years, thanks to the development of AI-powered solutions. These solutions have revolutionized the process of creating 3D models, making it faster, more efficient, and seemingly magical. Whether it’s converting text to 3D, transforming image series into 3D models, or generating 3D models from a single 2D image, AI has taken center stage in the realm of 3D printing.
A recent report by 404 Media, however, has shed light on a startling revelation. It turns out that an AI startup, called Kaedim, has not been entirely truthful about their 3D model generation process. While their website boasted about their proprietary AI tools, which supposedly generated top-quality 3D assets, it has now been revealed that human CAD operators were secretly working behind the scenes to manually fix the models.
This news came as a surprise to many, as users were under the impression that the entire process was automated and driven by AI. Kaedim has now updated their description to reflect the truth: “Kaedim’s machine learning and in-house art team combine to deliver production-quality assets in minutes.” It remains to be seen how well their hybrid service actually works, as it has not been tested yet.
This revelation raises some interesting questions about the reliability and efficiency of AI-powered 3D model generation. In my experience, using various AI tools to generate 3D models has been a time-consuming process. Some tools take only a few minutes, while others can take hours to complete requests. Initially, I attributed this to the algorithms running on busy servers. However, after learning about Kaedim’s reliance on human intervention, I now have suspicions about other AI services with lengthy response times. Could it be possible that they too are using manual labor to fix up the generated 3D models?
Moreover, the quality of the 3D models generated by AI tools has often been subpar. I have come across instances where the generated models contained glaring errors, such as an elephant with two trunks. While such mistakes are undoubtedly disappointing, the fact that AI can generate any form of 3D model is still quite remarkable. It is clear that these AI tools have immense potential, despite their current limitations.
Kaedim’s attempt to provide “game-ready assets” using AI technology in its infancy seems to have backfired. They may have tried to leap ahead before the technology was truly ready, and as a result, had to rely on manual labor to fill in the gaps. Nonetheless, this is yet another step on the path towards more advanced AI-driven 3D model generation.
In conclusion, the report on Kaedim’s deceptive practices has cast doubt on the reliability and effectiveness of AI-powered 3D model generation. However, it also serves as a reminder of the incredible potential of AI in this field. With further advancements and refinement, it is highly likely that we will witness the emergence of highly competent AI-generated 3D models in the near future. The road to AI may have some bumps along the way, but the destination promises to revolutionize the world of 3D printing.
Via 404 Media Share this post:
