Intellect vs. intelligence: The difference matters in AI

Intellect is the faculty of reasoning and objective understanding. Intelligence is the ability to acquire knowledge and skills. Both of these definitions were yielded to me by my laptop, a general purpose machine, which is itself an ancestor of the Electronic Numerical Integrator and Computer (ENIAC). The ENIAC was the first programmable machine, built at the University of Pennsylvania in 1946. Up to that point, all tools were single-purpose — that is, designed to do one thing. Hammers, calculators, steam engines, and bow and arrows are all examples of single-use tools. The rise of computers made it so we can now set our tools to adapt themselves.

The rise of computing inspired researchers with various tales of techno-utopias, visions of abundance, and endless productivity. Alas, this was all to devolve into techno-Taylorism and the mere optimization of industries. The new machine mostly made our stuff faster, smaller, and wireless. Then the internet, endless sensors, and exponentially fast computing were all joined together. Again, we started wondering about the future. We were building machines beyond our understanding, and as such, it was becoming difficult for us to extrapolate what they might do and how our relationship with these tools might evolve.

This quick journey through time roughly gets us to today, where we’re mostly concerning ourselves with automation, artificial general intelligence, and robot overlords. Before we carry on with our journey into the future, I want to reflect on the nitty-gritty of humans, machines, and intelligence. I would like to indulge in a quick argument about the inevitable, organic nature of business: Anything that can save money for a business will happen. And anything that can be broken into sets of instructions can be done by a machine. As machines become exponentially more capable in the narrow manipulation of intellectual and physical property, it follows that anything that can be broken into sets of instructions and saves money (with no value lost) will be automated.

Now that we’ve covered that inevitability, let’s consider design. Intellect is objective. It is the domain of tasks that don’t require redundancy, exploration, or judgment — the hallmarks of human thinking. Tasks to include under this pillar would be driving a truckload of flowers from Amsterdam to London, semantically scanning a million medical journals for an answer, or counting apples in an orchard. In today’s technological landscape, this is where machine learning and statistical computing make significant progress. This is because these are largely linear and highly structured mechanistic flows.

Intelligence, on the other hand, is conditioned in its subjectivity. It is the canvas to which we bring our mental, intellectual, and personal backlog. This is where exploratory innovation happens; where machines are our tools, not our replacements. It is the poetry that connects phenomena with frame.

Intelligence is the innovative idea the truck driver had while picking up empty plant pots in London, after listening to a podcast about the circular economy on the road. It is ideas that consolidate processes, break down silos, and make a resource more interoperable. Post-Taylorist thinking is multidimensional, unstructured, and very hard to plan for. Intelligent tasks cannot be automated, not until we build an artificial general intelligence at least.

This push and pull between intellect and intelligence is natural. When was the last time you were certain your gut instinct was right, only to objectively learn later that you were wrong?

Conversely, we all make inexplicable, on-the-fly choices when speaking to our spouses, meeting with our bosses, or playing sports. I don’t see any of us taking a calculator out when kicking a football. Let us consider what we do that is intellectual and what is intelligent. The first will be automated, either by us or by our (human) competitors. It is the intelligent design of automation that will lead the next generation of businesses.

Nitzan Hermon is a principal at Studio VV6, a communication studio, and technology consulting company. He is also a researcher and designer in AI, human-machine augmentation, and language processing.