AI Adoption. The bubble that won’t burst

The Reality of AI Adoption vs. the Hype

A recent study by The Reuters Institute and the University of Oxford has identified a huge “mismatch” between the “hype” around AI adoption and the “public interest” in it.  Whilst there is widespread awareness of Generative AI overall, frequent use of ChatGPT is rare, with just 1% using it daily in Japan, rising to 2% in France and the UK, and 7% in the USA.  Many of those who say they have used Generative AI have used it just once or twice, and it has yet to become part of people’s routine internet use.

Many of those who say they have used Generative AI have used it just once or twice.

The Reuters Institute and the University of Oxford

Living in the AI Bubble

Yet, if you are reading this article, the likelihood is you are all over AI, using multiple tools daily whether for personal or work use.  As a result, given both your own experiences, and likely the experiences of those people that surround you, you might find this study hard to believe.  The truth of the matter is that whether on LinkedIn, your newsfeeds or the pages you browse on the internet, it is all a curated feed and you are living in a bubble. The AI Bubble.  

Has AI Crossed the Chasm?

This is also likely to suggest you are an early adopter so congratulate yourself.  The likes of Chat GPT have only been around since late 2022.  That starts to put things in perspective.  AI adoption is yet to become mass market and is yet to become part of people’s routine internet use.

The key word in the sentence above is ‘routine’, and routines are based on behaviour.  Behaviour change is hard.  After your early adopters, carrying this through into mass market adoption is slow and it requires further maturity to make behaviour change easier. Hype suggests we have already crossed the chasm. But when you cut through the hype and look at the survey results, the reality is adoption is still low.   For example, Gen AI is yet to be seamlessly embedded in peoples every day lives.  It is still a separate activity to engage Chat GPT (Unless you are one of the 5% of consumers using MS Edge).

Lessons from History

To put this further into perspective the table below demonstrates how long it took for global technological advances to achieve mass market adoption:

• Electricity: It took electricity about 46 years to be adopted by 25% of the US population.
• Telephone: The telephone reached 25% of US households in 35 years.
• Television: TV reached 25% penetration of US households in 26 years.
• Internet: The internet achieved 25% penetration in just 7 years.
• Smartphones: Smartphones reached 50% of the US population in under 10 years.
• Social Media: Platforms like Facebook gained millions of users within a few years of launching.

The Accelerating Adoption Curve

One thing you will notice is that the adoption curve of new technology is accelerating faster.  Before Television, advances like electricity and telephones relied on physical connectivity (a network) for adoption.  No cable, no service, regardless of demand so this is an unfair comparison. Television, the Internet, and Smartphones all relied on expensive consumer hardware. This put adoption out of reach for many when they first came on the scene. Again, this makes an unfair comparison.  

So Social Media is the only like-for-like comparison we have to compare. The similarities are that they are both over-the-top services utilising the infrastructure that already exists.  Facebook launched in 2004.  It reached 25% penetration of US households approximately 3.5 years after its launch gaining widespread popularity around 2007-2008.  It expanded beyond college campuses to reach a broader demographic, including adults and families. This rapid growth was fuelled by the platform’s appeal for human communication and increasing functionality.  But it was also fuelled because it played to human psychological behaviours, namely Social Interaction, Validation and Recognition, Identity Construction, and Escapism.

AI’s Epiphany Moment

AI does not necessarily address many of these psychological needs at this stage of development but as services like AI-girlfriends/boyfriends etc come to market, we will start to see adoption based on human psychological needs that will drive behaviour change.  However, what we are starting to see, at pace, is the introduction of AI-driven features in your smartphone.

Some of this is evident in the everyday apps we already use today.  However, an epiphany moment for AI’s impact on consumers will be the introduction of AI at the operating system level.  Apple AI was recently announced to be introduced in the Fall of 2024.  This will have a huge bearing on consumers’ AI adoption whether they know it or not. AI will become native to the many apps and routines consumers habitually stick to.

Writing tools embedded in the keyboard will impact the emails you write and the notes you take. Siri voice controls will become more powerful and relevant in getting things done productively. Intuition from your device will ensure critical messages or calendar appointments are never missed.  Barriers regarding privacy will also be removed.  

Edge Technology: The Key to AI Adoption?

Many of these AI functions impacting your daily life will take place on the device, i.e. on the Edge.  Data will not be removed from your device to centralised processing to carry out many of these beneficial functions.  Instead, in many situations, it’s integrated into the core of your iPhone, iPad, and Mac through on-device processing. So it’s aware of your personal information without collecting your personal information and through SDK integration, this makes it easy for developers to integrate system-level features like AI-powered Siri and AI-enhanced keyboards into your favourite apps, again, without collecting your personal information.  AI becomes democratised and accessible without huge behaviour changes or privacy infringement.

Why the AI Bubble is Unlikely to Burst

So, the upshot is we live in a bubble at present but this bubble is unlikely to burst.  What we can do as citizens of the planet is help the sectors of society that might have behavioural barriers, skill barriers or knowledge barriers to adopting this new technology where it can add value to their lives.  If we are involved in bringing this technology to market, we can think about how we create AI products and services for everyone, that remove barriers to entry, make this technology accessible and above all, make AI trustworthy.

Next Steps: Reducing the Digital Divide

Read part two of this blog which looks at how we can reduce the digital divide for AI adoption, and also the real risks that could widen this gap.