Timing. It’s one of the most difficult facets to consider when thinking about the future. We know that convergence across societal, political, economic, science and technological forces is creating many future scenarios. We also know that enablement is happening at an exponential pace. Some believe (present company included) that the coming macro-level tipping point is likely to impact humanity on a scale only experienced twice in human history (hunter-gatherer to agriculture and agriculture to industrial). There will be many micro-level tipping points on the journey towards an automated society – and the timing of those tipping points is impossible to predict.
In the next twenty years, the path to tipping points is influenced by obstacles and accelerants. Timing is therefore dictated by several factors with many unknowns. As scenarios progress, various markers will provide guidance as to their potential paths and timing. Our rehearsals must consider all this, as we look to inform our future strategies.
Will automation lead to technological unemployment or will an aging society drive a fall in working age population? Will it be a hybrid scenario where there are many unfilled jobs due to skill gaps? A more near-term consideration is the investment decision surrounding artificial intelligence. When viewed through an ROI lens, organizations struggle with AI use cases and technological maturity. Is it too early to invest? We can debate whether ROI is the right lens (it’s not) but the question is one of timing. Is the technology mature enough to deliver real value across multiple use cases? When will it reach a point where inaction threatens viability? First, given the pace of innovation, inaction is not an option. Sitting on the sidelines is the fastest path to irrelevance. Second, rehearsing must be the mindset going forward. There are simply to many building blocks converging, as the combinatorial nature of innovation today accelerates the path of both science and technology.
One marker is the growth in computing. If we leverage this marker in evaluating the tipping points associated with artificial intelligence, it provides guidance as to the potential path. Let’s use a visual from Ray Kurzweil to look at potential tipping points.
Overlaying his visual on our emerging futures curve, we can look at the potential path:
Artificial narrow intelligence: this early phase of AI represents the narrow, use case specific application of cognitive computing. In this phase, AI is applied to a specific use case – like the medical application of IBM Watson. It must be set-up for each use case, and anything learned cannot be used in a general sense. Although still in its early stage, we are at the tipping point.
Artificial General Intelligence: by 2029, Ray predicts that we will have reached the artificial general intelligence phase, where AI has reached human intelligence and is no longer tied to specific applications. At this stage of maturity, the intelligence of a machine can successfully perform any intellectual task that a human can, leveraging all that it learns across all domains.
Artificial Super Intelligence: this is the phase that has so many in the industry worried. AI at this stage has reached a level of intelligence smarter than all of humanity combined, with the potential to realize many science fiction scenarios. This phase has a projected tipping point of 2040 – but many believe that it occurs much later. This stage is often linked to technological singularity, which is the hypothesis that the invention of artificial super intelligence will abruptly trigger runaway technological growth, resulting in unfathomable changes to human civilization.
We used one marker (growth in computing power) to look at timing. That does not consider the many potential obstacles or accelerants likely to influence the path of artificial intelligence. Does quantum computing accelerate the timeline? Does the growing ethical debate slow it down? These are all factors to consider in any future thinking exercise.