I’ve been impressed over the years by how Elon Musk consistently seems to achieve his goals, be it with Tesla, Solar City or Space X, so I wouldn’t be surprised to see his new startup, Neuralink, produce as promised. But that scares me.

I’ve spent the last two decades in the field of user experience (UX), mainly conducting research to make sure that human-computer interaction is successful. In that time, I’ve seen repeatedly that there always needs to be a backup plan: an opportunity for a user to reroute after mistakes and confusion. Even a generally great interface just won’t make sense for some people, and without a plan B, task success would be otherwise unachievable.

An interface should always have backup plans and redundancies

What I find when I do usability testing of web and mobile resources is that invariably there is always some situation for which the product creator firmly believes the interface provides a single, clear, logical flow for users. Yet users don’t process as expected, and so they take a wrong turn or become confused.

When this happens, I see that some users can auto-correct—albeit with some frustration—but others get hopelessly lost and either tell me that they give up or else end up going around and around in circles until I suggest that we stop the task and move on to something else.

When I give interface recommendations to a client, I often suggest tweaks to the expected main pathway to limit potential opportunities for confusion. But I also often suggest places where related links and alternative pathways can be added. In the event that a user does take the wrong pathway, there is an opportunity for them to reroute using one of those alternatives.

Crashing the interface

Sometimes when doing usability testing, users’ surprising behavior will actually crash the interface—they simply did something so unexpected that there was no anticipated correction. Or maybe, they took much longer than anticipated as they struggled to understand and certain behind the scenes action timed out.

When the interface crashes today, it’s often simply a matter of restarting the app or the browser. But what happens when the interface that crashes is in a person’s head? Is it simply a matter of rebooting? Or does the person become so dependent on that interface that they crash too? What if the crash sends the interface into some sort of unstoppable loop of distraction? I would wonder what might happen if they are driving at the time—but if Tesla and other companies are successful by then that will be automated anyhow. More simply, let’s say that the crash happens if they are climbing the stairs. Do they fall back down?

A pause button

In today’s world when users take a wrong path, after determining that they are stuck, they can pull their hands away from the keyboard or screen, pause, reflect and perhaps ask for help. Is that possible with a hardwired neural interface that quite literally moves at the speed of thought? Is there going to be an opportunity to throw up one’s hands and press pause?

Every user is different

Ultimately what makes usability testing so much fun and keeps us user researchers on our toes is that every user is different. Every participant for a study comes in with a different cultural, technological and educational background. At risk of being cliché, each user of a product is a unique snowflake, and the product needs to account successfully for this wide range of individuality.

When we talk about a brain interface, we add an additional layer of complexity. Beyond just thoughts, beliefs and behaviors, at their core, each person’s brain is physically and biochemically different, effectively with different programming—a unique combination of learned and hardwired—as unique as can be. While even identical twins can share the same DNA, no two people have the exact same combination of DNA and life experiences that shaped this brain wiring.

With today’s user research, we are focused on analyzing the impact of visible behaviors on a stand-alone interface. Putting (valuable) accessibility research aside for the moment, as well as the ever growing popularity of voice, most interactions happen with the hands in some way. And hands interact with an interface similarly.

The complexities of the brain would seem to require a whole new layer of considerations and a whole new set of backup plans. What if a person who is “wired” for more anxiety feels so much anxiety that it overwhelms the mental interface commands? What about an obsessive thought that clouds out the proper neural commands? Even without diagnosed OCD, we all can become obsessed. Is there a backup plan? Pop a Xanax and try again? That doesn’t sound quite right.

The varieties of (within-person) human experience

With voice commands – be it to Siri, Alexa or Cortana – I can imagine someone waking up in the morning and trying to talk before taking out their nightguard. Doesn’t work – okay – they take out their nightguard, reduce the slur, and try again. How does that work with a brain interface?

We’re human: We get tired, we get sick, we may drink a little too much, we periodically take medications that may affect our thinking, and we sometimes have “off days” for no apparent reason. How will error-correction programming account for the same person thinking differently on different days due to internal and external factors. At what point will a persons’ thoughts be “normal” and at what point will their thoughts be different enough to confuse the interface? How can that difference be corrected for?

Let’s have early discussions about neural interface redundancy

I have lots of questions, and not enough information to even fully speculate about solid answers. But no matter what, there must be interface redundancy, and there must be a backup plan when considering incorporating a digital layer into the human brain. That backup plan can’t only be focused on system issues and dealing with potential technology risks or even sabotage; it also needs to address what happens when users make mistakes—because they will – we all will – so Neuralink better be ready before those clinical trials begin!

 

Image of wired brain: LexxIam / BigStockPhoto