Have we crossed the AI chasm yet?

Technologists love to talk about ‘crossing the chasm’ in relation to the development of any new platform, product or service as it moves from its developmental stages towards wider levels of mass adoption or failure.

The term itself comes from Geoffrey Moore’s famed 1991 book of the same title, where the chasm represents the gulf between early adopters and the perhaps more skeptical pragmatists who hold back in order to assess where any new innovation is likely to be headed.

This begs the question then: have we crossed the artificial intelligence (AI) chasm yet?

Movies to mainstream

After its initial appearance in several 1980s sci-fi movies, AI went largely silent while the IT industry went through various reinventions of itself moving from client-server models to mobile device ubiquity and then ultimately towards cloud computing and the present day.

During that (roughly) last quarter-century’s developmental phase, we managed to bring down the cost of data storage, develop massively advanced approaches to analytics and even broach the hitherto unimaginable world of quantum computing.

As these forces have now started to coalesce, we have been able to apply AI to real-world industry use cases. So has AI finally crossed the chasm and become part of our everyday lives?

Industry use cases

It feels like the answer could be yes, AI (or at least some version of it) is becoming very real.

Reading through the IT industry’s newswires, it has become clear that AI will be increasingly applied to chatbots designed to serve online customers’ needs across a variety of different industries.

Terms like ‘digital onboarding’ have sprung up where AI ‘agents’ help log users into new systems and sign us up for services that will themselves be driven by increasingly AI-enriched intelligence. For the record, these agents are actually just pieces of software, but we humans like to give inanimate objects personalities and make them feel tangible, right?

Ideation nation

If we really are over the AI chasm, then it’s not just chatbots and web pages that we’ll be experiencing AI advancements on, it’s deeper levels of business strategy and planning. Companies have been pointing AI engines at their business models to create what they like to call rapid ‘ideation’ (yes, it’s really a word now) and prototyping to test new methods of working and shake up their supply chains.

AI across the chasm is a place where firms don’t necessarily need a Research & Development (R&D) department. Instead, they can direct AI to help develop entirely new revenue streams based on assessments of market demand and an intricate understanding of potential customer demographics.

Predictive business

With our feet firmly planted on the other side of the chasm, we can start to use AI to perform predictive analytics and run business models with a far more granular level of data intelligence. That means more data, about more things, impacting more aspects of the business every day.

For example, a baked bean manufacturer used to run its business based on consumer demand and the price of beans and other core ingredients. Its more sophisticated business model would perhaps include seasonal market fluctuations, supply chain information and factors related to the amount of advertising and marketing it engaged in.

Now, across the AI analytics chasm, our baked bean manufacturer (or any other random business you care to imagine) can start to also monitor consumer sentiment on social media, health trends, Natural Language Understanding (NLU) processing to filter through customer call center records and perhaps even the air quality levels impacting workers in its factory.

What AI does next

What AI does next is a question of how we train the engines that drive it, obviously. As we expose AI engines to new layers of our business structures, it will go through machine learning (ML) cycles that allow it to get the intelligence we demand.

What we as businesspeople need to do is to understand what AI feeds on, what it likes and what it is good at. When we realize that AI is good at automated ‘grunt work’ of a repeatable nature, then we apply its benefits more directly to real-world operational business models

It’s probably safe to say that we are across the AI chasm, but in many places, we’re only just managing a foothold on the other side. In a world where AI engines still need to differentiate and classify between basic physical objects that a three-year-old child has already mastered, there may be a while before we compartmentalize and capture all those basic elements of understanding about our world.

AI has crossed the chasm, but keep your climbing gear on for now if you can please.