I was born in Japan and grew up playing Nintendo games.
For me, the first computer I really felt close to may not have been a personal computer at all. It may have been a game console.
The Famicom controller had only a directional pad and a handful of buttons.
And yet, pressing one button could make a character jump. Pressing the directional pad could move something through an entirely different world on the screen.
A small movement of the hand immediately changed what was happening in front of me.
As a child, I did not think of this as an “interface.”
It simply felt natural and fun that something I touched with my hands could directly change what happened on the screen.
When input changes, the experience changes
Games continued to explore many different forms of input.
Analog sticks.
Motion controllers.
Touchscreens.
With the Wii, moving the controller itself became part of the game.
With the Nintendo DS, touching the screen directly created entirely new kinds of interaction.
Even when the computer underneath is fundamentally similar, changing the way we interact with it can create a completely different experience.
An input device is not simply a tool for sending commands.
It can change how we understand, experience, and relate to a computer.
That feeling is one of the starting points behind PitsBoard.
The mouse changed our relationship with computers
Looking back at the history of computing, changes in interfaces have repeatedly changed computing itself.
In the 1960s, Douglas Engelbart and his team demonstrated many ideas that would later become fundamental to modern computing, including the mouse, hypertext, and windowed interfaces.
Instead of interacting with a computer only by typing symbolic commands, the idea emerged that a person could point to something on a screen and manipulate it directly.
Graphical user interfaces later developed further, and computers such as the Apple Macintosh helped bring icons, windows, and mouse-based interaction to a much wider audience.
I do not think this was simply a matter of making computers easier to use.
It changed the relationship between people and computers.
A computer was no longer only a machine that required users to remember and type commands.
It increasingly became a space where people could see objects, point to them, and manipulate them.
Smartphones shortened that distance even further.
The pointer disappeared.
Instead of moving a cursor to an object, we started touching the object itself.
To scroll, we move a finger.
To zoom, we spread two fingers apart.
The relationship between action and result became more physical and direct.
And yet, PC interaction has not changed that much
Desktop computers are interesting in this respect.
Computing power has changed dramatically over the past several decades.
Displays have become higher resolution.
GPUs can process enormous amounts of information in real time.
Software has become vastly more capable.
And now AI systems can increasingly understand natural language, images, and even relatively ambiguous human requests.
But in many desktop applications, we still move a pointer with a mouse, click small icons, open menus, and memorize keyboard shortcuts.
Changing brush size in Photoshop.
Moving through a video timeline.
Rotating a model in 3D software.
These are completely different actions in terms of meaning, but from the perspective of the body, they are often reduced to very similar movements.
Move the pointer.
Click.
Or press a memorized key combination.
There is no question that the keyboard and mouse are extremely effective general-purpose input devices.
But their generality also means that many very different software actions are translated into a relatively narrow set of physical movements.
Why did we need so much GUI?
Traditional computers had an important limitation.
They were not very good at understanding human intention directly.
A person might say:
“Make this photo feel a little softer.”
“Make this part slightly less noticeable.”
“Find a shot that connects naturally from here.”
Another person can usually understand the general intent behind these requests.
Traditional software could not.
Instead, software designers had to translate human intentions into precise, machine-readable operations.
Buttons.
Menus.
Sliders.
Checkboxes.
Dialog boxes.
Complex applications may contain hundreds of functions, each of which needs a clearly defined way to access and control it.
Smartphones often took the opposite approach.
Rather than exposing every possible function, interfaces became simpler by reducing options or limiting what could be done in a particular context.
Both approaches made sense in a world where computers could not deal well with ambiguity.
AI is beginning to change that assumption.
AI may reduce the need to give exact instructions
One of the most interesting things about AI is that it can accept a certain amount of ambiguity.
A person no longer needs to specify every parameter explicitly.
“Make it a little brighter.”
“Shorten this paragraph.”
“Clean up this area naturally.”
AI can increasingly interpret instructions like these.
There are still major issues around accuracy, consistency, and predictability.
But the broader direction seems clear to me.
People may gradually need to spend less effort translating their intentions into the exact internal structure expected by a computer.
If that happens, the meaning of being “fast at using a computer” may also change.
Memorizing large numbers of keyboard shortcuts.
Typing commands extremely quickly.
Knowing the exact location of every function inside a complex menu hierarchy.
These skills are not going to disappear overnight.
But I am not sure they will remain the central measure of comfortable and efficient computer use.
Instead, interaction may become more about combining different ways of expressing intent.
Quickly indicating what you want to act on.
Selecting an object intuitively.
Expressing degree or direction physically.
Using language only when language is useful.
PitsBoard also comes from this hypothesis.
Language is not the answer to everything
AI has made natural language an extremely powerful computer interface.
If we can simply say:
“Open this.”
“Summarize this document.”
“Find the file I worked on yesterday.”
then many traditional GUI operations may become unnecessary.
But I do not think a future where everything is controlled through language would necessarily be ideal.
Imagine wanting to reduce the volume just slightly.
Move an image a few millimeters.
Adjust a color while looking at the result.
Scrub backward and forward through a timeline until something feels right.
Turning all of these continuous actions into language would often be less natural.
“Move it a little further right. No, slightly back.”
In many situations, moving something directly with the hand is simply faster.
Human intention is not always formed as language.
We adjust while looking.
We search while touching.
We think while moving.
These non-verbal forms of thought are also important parts of how we interact with tools.
Input can become even more direct
PitsBoard is initially being developed as an input device for desktop applications.
I think the most realistic place to begin is with software and computing environments that already exist, and explore different ways of interacting with them.
But the idea behind PitsBoard — connecting physical action more directly with digital state — is not limited to desktop computers.
In the future, I think intuitive input will become even more important with devices such as VR headsets and AR glasses.
A graphical interface built around sitting in front of a display and moving a mouse cursor may not always be the most natural model in spatial computing.
Look at an object.
Move a hand.
Touch.
Press.
Turn.
Use language when necessary.
In many situations, combining these forms of input may be more natural than relying on a single one.
This is also why I am interested in eventually connecting PitsBoard with VR and AR systems, eye-tracking tools, and other input technologies.
PitsBoard does not need to become the device that handles every kind of input.
It could instead become one physical part of a larger interaction system that combines gaze, voice, and physical action.
Combining body, gaze, and language
I do not expect the next generation of computer interfaces to converge on a single form of input.
We probably will not control everything with voice.
We probably will not control everything with our eyes.
And we probably will not control everything through touch.
Each method has different strengths.
Gaze can indicate which object we are interested in.
Physical movement can communicate how much or in which direction.
Language can express what we want, even when the request is ambiguous.
AI can interpret the space between them.
If that becomes possible, we may not need to decompose every human intention into hundreds of predefined buttons in the same way we do today.
This is still only a hypothesis.
But the more flexible software becomes through AI, the more freedom I think we may regain on the input side as well.
Computing as an extension of the body
The goal of PitsBoard is not to define one final interface for the future.
I would rather create a place where different possibilities can be tested.
If something needs to rotate, give it a dial.
If something needs continuous adjustment, give it a slider.
If a direction needs to be selected, create a physical form that communicates direction.
And when useful, combine those interactions with AI, voice, or gaze.
The physical shape does not need to remain fixed.
It can change depending on the application and the interaction.
The screen can show only the information that matters in that moment.
I do not want PitsBoard to become just another shortcut device.
There are already many excellent products for assigning keyboard shortcuts to physical buttons.
What interests me is something more fundamental.
How can the functions of a computer connect more naturally to the human body and attention?
Not separating seeing from touching
Physical controllers have one major advantage.
You can often understand position and shape through touch alone.
Graphical interfaces have a different advantage.
They can change dynamically depending on context.
PitsBoard tries to use both.
But I do not want the user to constantly look down at the device.
Ideally, attention stays on the thing being created or manipulated on the screen.
The hand understands the interaction through shape and physical feedback.
The eyes see the result.
And when necessary, AI can help interpret ambiguity.
The ideal state is something closer to using a familiar tool, where the interaction method itself fades into the background.
With a pen, we do not think about the coordinates of the pen tip.
When riding a bicycle, we do not consciously calculate the angle of the handlebars.
We move, and something happens naturally at the other end of that movement.
I think computers could move a little closer to that.
Not necessarily by adding more features, but by reducing the small distance between human intention and computer action.
Instead of memorizing how to operate the system, perhaps we can increasingly look, touch, move, and understand.
And perhaps using a computer itself can become a little more enjoyable again.
PitsBoard is an experiment in exploring that possibility.