On February 11, 2021 Elon Musk told Joe Rogan that (with respect to software design) “all user input is error.” This was in reference to the lack of turn signals and drive direction controls for the next version of the Model S vehicle. Ultimately, the goal is for there to be no steering wheel (or “yoke”) for the user to engage with at all. A sensor-fed computer does not fatigue. Self-driving vehicles have the potential to supersede the capabilities of even the best human driver. Sensors can pierce through weather (radar), see in the dark (infrared red) and even see around obstacles (ref). So how does this relate to Intellectual Property (and specifically software patents)?
Avoiding Mental Steps
Software architects and developers should seek to minimize user engagement. To perform any function, how many times must your user interact with input devices? When the software has the intelligence to anticipate the user interaction, you have created a practical application and minimized the possibility of user error by input. Inventors consume themselves with creating an entirely new technology rather than improving what exists.
A Mobile App Example
Consider a QR code check in to a gym. In this example, the user must engage the device as follows:
- Open the application;
- Press the “hamburger” menu icon on the upper left;
- From the slide-loaded “Navigation menu” select “Membership Card.”
Did the developer consider that displaying the QR code was something done every time the user would enter the club? Instead of embedding intelligence into the mobile application, the developer put the burden on human interaction with the application. Unless the software application is a game, users do not want to repetitively interact with an application. Consequently, in the design of software, all user input is error. In our example above, the QR code should show at the application launch or homes screen:
Making the application smarter
Why should the user even have to interact with the device in the first place? The mobile app could run in the background and automatically check you in when you walk into the club? The “Contactless Check-In” option again requires manual activation. A well-informed developer might note polling the GPS might drain the battery but intelligent handling of that is easy. Poll every 5 minutes, if within the radius of a club, poll at a shorter interval, etc… By automating this, you reduce both user distraction for the user and also the health club employee. For a software patent, how the concept is implemented is critical. Ease of use, conserving battery life, gathering appropriate and pertinent data, etc.
Automation and Software Patents
In 2014, the United States Supreme Court reined in patents on abstract ideas in Alice v. CLS Bank. In doing so, a wide net caught many software applications and found them unpatentable. The United States Patent & Trademark Office provides guidance through numerous publications but inventors and patent attorneys still struggle. A key aspect of finding unpatentability is something called a “mental step.” This means it requires a human to make a decision or perform a task. In other words, the software design requires user input (which may lead to error). The further we remove user input from the software design the more intelligent and automated the software becomes.
Tesla produces more than simple cars. They are computers on wheels and run advanced software . Driving (or riding in a car) is common and repetitive. Changing gears, switching turn signals, braking and accelerating are repetitive tasks. Humans are capable of performing these functions but are subject to distraction, incompetence, medical ailment and even intoxication. When a computer can perform a task better than a human, but we nevertheless require human engagement there is a problem. The ultimate goal for an automated system and acknowledge that “all user input is error.”