The robots are coming, the robots are coming!

Honda's Asimo Robot, the most advanced humanoid robot.

Honda's Asimo Robot, the most advanced humanoid robot.

Everywhere we looked in 2015 there seemed to be another headline about how artificial intelligence (AI) is going to take over the world and eradicate the human race. The fear that the media is playing off even has its own buzzword: AI anxiety.

If you're not up to speed with the potential of AI, it's worth investigating to form your own conclusions on the impact on the human race. The basic premise is that the ability to combine sufficient computing power (hardware) with intelligence (software) to rival the human brain will be possible, for sub-$1,000, somewhere between 2035 and 2055. What that means for the human race and our society is the subject of much debate. We at Intersection X won't be so audacious as to present an opinion on that grand topic, but rather, we'll focus on what it means for a subset of the human race: productizers!  

Let's familiarize ourselves with the concepts to get started. Perhaps the most considered article (and easily our favorite of the various pieces we've read) is from the fine folks at WaitButWhy. We're going to lean on a nice framework they outlined to help us answer the question "Why should productizers care about AI?"

The WaitButWhy folks talk about three categories of AI: Narrow AI, General AI and Super AI.  Narrow AI is the category of AI that can automate tasks of a very focused nature. General AI builds on Narrow AI and could, in the future, replicate the general intelligence of humans. The theory says that there comes a time when the general intelligence of AI will improve to the point that it surpasses human intelligence and it becomes Super AI. Super AI is what's getting lots of people wrapped around the axle.

As productizers, we like to tout that we can imagine the future of products and help bring them to life. But in the context of AI, we're very shortsighted; we rarely think more than one to five years down the line because our goal is to deliver revenue as soon as possible! That means Super AI, at 20-plus years away, isn't relevant for productizers today. What about General AI? Well, that's a good question. At Intersection X, we are lucky to be exposed to leading-edge AI work being carried out by a range of clients, and everything we have seen suggests General AI is not on the radar in the next 24 months.

So that leaves us with Narrow AI. Let's discuss an example to get a sense of its power and how it's already impacting daily life unbeknownst to us. When I emailed an investor friend of mine to meet for lunch, he added his admin, Amy, to the conversation to find a time to meet. Amy and I emailed back and forth and in a few minutes found a mutually workable time to meet. I was marveling at Amy's efficiency when I noticed the domain of her email address: After some quick Googling, I discovered that Amy is an AI executive admin powered by the startup Nothing about the experience had suggested to me that Amy was a computer; it was simply her email address that tipped me off. I was blown away. For reference, this happened in January 2014. Yep, well over a year ago... eeek!.

Let's look at another Narrow AI example from Building System Planning (BuildingSP), a client of ours. If you're sitting in an office right now, look above your head. Can you see all those pipes and ducts hanging from the ceiling? Did you know that a group of highly paid CAD drafters and engineers had to sit at a computer and laboriously route the water, the electric conduits, HVAC pipes and ducts through the building?  It's not uncommon for each routing (water, HVAC, etc) to be done by a separate draftsperson and for it to take months to complete for one building. BuildingSP is leveraging the power of AI to eliminate the human time needed to route the pipes. Imagine the architect simply describing the requirements - "Route the water and electrical ducts, give the water pipes precedence and optimize the routing for cost. And make sure everything is within 10-inches of the ceiling". Thirty seconds later BuildingSP's artificial intelligence has generated the optimal solution. They're revolutionizing the building and construction industry, which will lower the cost of building for everyone.  And BuildingSP is delivering this value today.

Ok, so Narrow AI is living and working in our midst today. Back to our question though: Why should productizers care about AI? Well, what's consistent in both of our examples is that the user experience has changed. More specifically, the user interface has (almost) been eliminated. We no longer need to learn how to use an application to achieve our goals. We simply tell the system what we want. Let that settle in for a second. You no longer need to invest time in order to learn how to use a product. Powerful, right?

The skeptics amongst you are probably pointing out that "telling" an application to do something doesn't always work so well. We've all been the angry customer shouting at our bank's phone-based system: "I said connect to an agent, grrr!". The reason that experience is so common is that speech recognition traditionally requires large specialized teams, huge data sets and a big investment to get to usable accuracy. Only the likes of Google and Facebook have that mix of resources. However, it probably won't surprise you to hear that AI is helping here too. Capio, an Intersection X client, creates AI-powered, self-learning voice recognition and conversational systems. They can get to usable accuracy in 48 hours, and their models get better over time by testing themselves against human-provided reference data. What once required a huge time and money investment is now becoming available to the (software-developing) masses.

It's time to put the nail in the coffin of user interfaces! AI is eliminating them and allowing users to get to the value they seek much more quickly. That's a huge shift for the products of the future. Essentially, we're no longer constrained, or we're at least significantly less constrained, by the limitations of clunky, non-organic interfaces like a keyboard or touch screen. It's a little difficult to fathom the possibilities of what might emerge from this transition.

So, it's time to ask yourself, what would your product be like without a user interface?

P.S. Just in case you're not freaked out enough already by the power of AI, we've created a list of the articles we've read to keep the paranoia going. Enjoy!