AI is crazy and almost every industry is talking about change.
The education sector is relatively cautious, with AI remaining more focused on entertaining, generating tutorials, criticizing assignments and answering questions.
For example, an art teacher gives a student's artwork to AI to move a still image, and the whole class shouts "wow".
Video from the New People's Evening Post in Little Red Book
At this year's WAIC, Kodak Flight's AI blackboard attracted a lot of people.
One of the most striking features is the use of AI to symbolize abstract mathematical formulas.
Enter a function formula, and the system automatically generates the corresponding image; adjust the parameters, and the image changes with it.
In three-dimensional geometry, space structures that students do not understand can also be transformed from paper diagrams into 3D models that can be rotated, scaled, cut and unfolded through transparency, color marking and perspective switching, intuitive representation of the relationship between points, lines, and faces.
As image generation, 3D modeling, and code Agent gradually enter the same workflow, abstract knowledge can be made into clickable, decryptable, and explorable interactive content. In the past, it required collaboration between designers, 3D artists, and cutting-edge engineers.
X blogger @thebuggeddev recently created a 3D human anatomy application that has attracted the attention of more than 1.6 million netizens.
Atmospheric programming has created 3D human anatomy applications.
The project, called Anatomy Atelier, is an interactive 3D web application for learning human anatomy.
It puts organ models, knowledge cards, and interactions on the same interface so that we can understand the structure of the human body more intuitively through rotation, scaling, and clicking.
Link to the application:
Link to the project:https://github.com/thebuggeddev/anatomy
Currently, the application includes nine types of organs: heart, brain, lungs, liver, kidneys, eyes, intestines, pancreas and skin.
Each organ is equipped with a relatively complete knowledge module. For example, the heart, a page that describes its size, weight, location, method of blood supply and main functions, and contains details such as micro tissues, functional animation, organ contrast, common diseases and related human systems.
We can rotate and scale 3D models, start automatic rotations, click hotspots to see specific structures such as the arteries, heart chambers, heart chambers and membranes.
The page also provides functions such as organ isolation, anatomical observation, layered display, organ contrast and repositioning perspective.
On top of offering simultaneous organ searches, as well as entrances such as Systems, Lessons, Library and Notes, the overall format has come close to a lightweight digital anatomy learning platform.
A developer who makes a teaching tool product
According to the author, he had always wanted to make a really useful product with "ambient programming". He drew inspiration from the 3D animal and plant cell project created by @DilumSanjaya and eventually completed this anatomical application.
The whole project began with a design map. He first used GPT Image 2.0 to create a visual program and then used this image as a reference to generate two-dimensional images of different human organs.
These organ charts were given to Tripo and converted into 3D models that could be used on the web. Once the visual assets were ready, he opened the Codex, wrote a main prompter word based on the design scheme, and combined the prompter words, design charts, and all 3D models to the system.
With the help of GPT-5.6 Sol and Three.js, the Codex quickly completed the first version of the application. Users can view different organs on a web page, observe their structure and location, and learn about the function of each part of the organ through interactive markup.
Codex also generates anatomical illustrations of the location of organs in the human body and creates clickable hotspot markers. Work originally distributed in visual design, front-end development, and content presentation was serialized in a continuous AI workflow.
The first version has been visually shaped, but performance issues are soon exposed.
Each 3D model is close to 120-150MB, with a total resource of about 900MB, and is difficult to deploy directly to the end of the web, running at a latency of about 16 tons.
After several rounds of iteration, the Codex compressed the individual models to about 2-5.5MB, while maintaining the original visual quality. The total assets decreased from about 900MB to 28.6MB, reducing the volume by about 97%. The system also added an on-demand loading mechanism, which allows users to load the corresponding model when viewing an organ.
The whole process was not completed in one go. The author progressed gradually in terms of visual design, organ generation, 3D conversion, web development and performance optimization, with each stage undergoing adjustments. He believes that the operation is not complicated, the key is to dismantle the target and then complete it step by step.
In traditional classrooms, anatomy relies on instructional illustrations and laboratory samples. Students review plans, recall organ locations, and recover their memory during exams. This method lasts for decades, is limited in efficiency, and experiences are exhausting.
Generated AI not only speeds up content creation, but also reduces the barrier to making something really useful. A developer, a computer, a set of prompts can make things that aren't clear in the classroom into a 3D world that can be rotated, clicked, and explored.
The link is as follows: