Generative Hardware

What is Generative Hardware, and are we close to productizing it, or are the capabilities far away? What needs to happen to commercialize this as a business offering?

History of Adaptive Interfaces

You're familiar with Gen AI. You type in a message and get a response, in the form of text, image, video, 3D objects or worlds, music, audio, or code.

Now that AI code completion tools have been productized and their accuracy improved substantially, users can ask a question and the machine can respond with custom user interface elements like buttons and text boxes that fit exactly what the user is trying to do.

It's a generative interface.

When Steve Jobs announced the original iPhone 1, he said the biggest problems with Blackberry and other smartphones like it was that the interface was always the same. The same buttons at the bottom half of the device no matter what app you were using. So Apple extended the screen down to the bottom of the device, enabled software to change the interface based on whatever the app wanted to do, and invented multi-touch as the way for humans to interact with it.

The issue, is that it required manual human work to define the interface and it was static until the developer changed the code.

Today, we have the opportunity to dynamically change the user interface based on what the user wants to do at that moment in time. Software, personalized, in real time based on the task at hand.

That is happening today. Inside ChatGPT, you can submit a prompt and it will respond with a dynamic table, chart, graph, or image.

So, it's been commercialized for nearly 1 billion humans using it every week.

What is Generative Hardware?

Generative Hardware takes this concept to the next level, and instead of adapting the interface to the app or the human using it, we now adapt the hardware to the human or machine using it.

For example, imagine a robotic hand that transforms into a screwdriver. Or a table that transforms into a chair.

Here's an example of Generative Hardware courtesy of Big Hero 6's Microbots, from the magic of Disney:

https://www.youtube.com/watch?v=fsVJuN75vzE

That's one type of implementation where a series of small objects "swarm" together in unique ways to solve the current problem. It requires a few things:

• Small or tiny component parts and materials that are strong and light individually
• Strong connection points
• Sensors that detect connection points and the environment it's operating within
• Controller software to determine the movement and connection points
• Energy source and power management
• Movement capabilities

This already exists with swarms of drones. It already exists with 3D printing mechanisms. It already exists with the open-source Robot Operating System. It already exists with Spatial Intelligence. It already exists with low-power computing, solar energy, and batteries. However, we need further miniaturization and self-mobility mechanisms to be built in.

A system integrator, mechanical manufacturer, product, software, and robotics experts could feasibly connect what exists, invent a few things, and begin to iterate on this.

Humanoid robotics represents a static implementation that is sub-optimal for real-world use cases. Opposable thumbs are only a start. The entire robotic system should be "opposable thumbs" with intelligence built in.

Then, of course, you need to spend a lot of time on core values and be very careful about who you trust to build it. Because we put ourselves into our creations.