ACI: the new type of AI that brands need to know about

Opinion

ACI has the potential to be a game-changer in AI by revolutionising the way brands serve their customers.

We talk about AI an awful lot. That’s because a lot of people fear it.

The fear is driven by visions of a humanoid intelligence that walks, talks, dances, drives and is prone to considering its own place in the universe. It’s driven in part by Hollywood’s weird, almost fetishistic interest in apocalyptic outcomes, spawning films like The Matrix, The Terminator or new blockbuster The Creator [although, spoiler alert: here it turns out the humans are the bad guys].

This happens because the debate over AI is unduly focused on artificial general intelligence (AGI). The ‘general’ in artificial general intelligence heralds the birth of an AI that can potentially match human capability across multiple tasks. We could even witness the arrival of AGI as a single AI that can execute a complete array of human tasks by rapidly adopting, adapting, and applying new learning to master every life skill.

The problem with these AGI-fantasies is that they are a distraction.

It’s particularly unhelpful for media and advertising professionals, who should really turn their attention to ACI.

Because ACI — artificial capable intelligence — has the potential to radically alter consumer behaviour.

The benefits of ‘Narrow AI’

As chronicled in AI guru Max Tegmark’s book Life 3.0, AI experts simply cannot agree on when AGI might arrive — or the criteria for officially declaring its arrival. Whilst some estimate it could be as early as the 2040s, others state we will be lucky to witness it before 2100, if at all.  Those eagerly awaiting Blade Runner replicants are likely to continue to be disappointed.

In fact, the undue focus on AGI is eclipsing a more useful angle in the present. Available to us in the now is Narrow AI. Unashamedly limited in its skillset, Narrow AI offers us bespoke AI solutions for specific tasks only.

The last decade has seen the crossing of certain thresholds. In 2015, Lee Sedol, genius champion of fiendishly difficult board game ‘Go’ was thrashed by an AI called AlphaGo, leading to a very public sulk. In 2021, Google claimed their driverless Waymo cars had clocked up 20 million miles on public roads. In 2022, many brands are already using rudimentary chatbots to triage customer enquiries via pre-programmed decision-tree style answers.

But these are all examples of Narrow AI, also known as Weak AI. Each AI named above can only do one thing: AlphaGo couldn’t tell you the opening hours of a supermarket. Supermarket chatbots can’t drive a car. And driverless cars can’t beat Korean super-geniuses at ‘Go’.

The strength of Narrow AI is in its dedication to one task: never aspiring to be a jack of all trades, but rather the master of one.

For brands and businesses, the opportunity lies in bespoke, channelled, focused solutions to consumer needs, like chatbots that answer very specific questions on mortgage eligibility, limited speech recognition on phone helplines, rudimentary “if-you-liked-that-then-you’ll-like-this” product recommendation engines.

Most significantly, according to one Silicon Valley luminary, Narrow AI is about to get an upgrade that brands and businesses are going to want to know about.

From Narrow AI to ACI

In his book The Coming Wave, Deep Mind co-founder Mustafa Suleyman states that whereas Narrow AI is dedicated to doing one task well, and AGI is the moment where AI becomes able to master multiple tasks, ACI is the moment AI will be able to handle a collection of differing but related tasks within a category.

To illustrate, consider the rigmarole of booking a holiday. Today, we’re much more likely to be independent travellers, purchasing our flights, booking our hotels, and arranging airport transfers with a portfolio of operators, and managing the synchronisation of our itineraries ourselves. The industry even has a term for it: Fully Independent Travellers (FITs).

FITs willingly take on the heavy lifting of scrolling through pages and pages of results to check availability, reading restaurant reviews, and comparing hire car prices, before packaging them up into a bespoke holiday for themselves. However, if one element of the holiday runs into trouble, for instance flights need to change, then like holiday Jenga, everything falls apart. The independent traveller must start again and rebuild the plan to rebook the elements.

ACI solves that issue by ostensibly becoming a Virtual Travel Agent. Its skill lies in understanding the linkages between a seemingly disparate set of tasks, and not only is able to perform a deep dive on hotel prices, like a Narrow AI, but also collate and align results of different, external tasks found elsewhere, for instance tickets for tourist attractions and, totally separately, restaurants nearby.

Whilst some package tours still comprise all-inclusive end-to-end experiences, the ability to customise might be restricted. But an ACI could potentially be free to curate a package using the widest selection of disparate options surfaced on the ‘open web’, presenting them back in one unified dashboard. Crucially, when one element of the holiday changes, a Travel ACI would be able to adjust and realign all the pieces of the trip to get it back on track, even though they appear to the AI as independent variables.

Health appointments, buying cars, reclaiming our time

Extrapolating out to a wider application, branded ACIs could become a powerful tool for businesses looking to provide end-to-end services that ordinarily might require advanced coordination between different elements of the business, or indeed, external partners working as subcontractors.  In short, it’s a way of turning your brand into a “one-stop-shop”.

To illustrate, we’ve already mentioned travel as an example. Health and beauty might be another, with appointments and treatments arranged and synchronised and aligned. Events might be another instance, with an ACI coordinating venue hire, arrival of the delivery trucks, catering set-up and soundchecks.

Purchasing a car might be further application, unifying the experience of searching, test-driving, purchasing, insuring, and servicing. Of course, some brands offer these services digitally now, but ACI would enable a whole new turbo-charged level of coordination and frictionless-ness.

The learning here is that consumers are often flicking between tabs on Chrome trying to line up the independent elements of their plan.

Because the value is in the savings they make and the tailored nature of the solution, they give up their time in exchange for a better overall result, rather than plump for an inflexible off-the-shelf solution.

But if a consumer could reclaim their time AND have a bespoke experience, product, or service, with just a few inputs at the front of the process and be digitally chaperoned throughout to track the progression and respond to changes, then this may be a more appealing prospect.  Brands would effectively become more like service providers.

This is what ACI may be able to deliver. And this is why brands should find out more.

Read more articles by Rowley here.