There are many things we look forward to at the end of each year – quality time with family and friends, holiday traditions, fruitcake1, and, of course, the many year-in-review articles. As we eagerly anticipate what’s to come in 2022 for AI, we take a moment to reflect and synthesize what was learned this year.
It may come as no surprise to you that each report reviewing the adoption of artificial intelligence displays upward trends, showing an increase in AI use among all industries. However, the stories told by the details of each of the major reports differ. While there are many organizations that compile reports on the use and outlook of AI, for our review we will limit our sample to O’Reilly2, KPMG3, IBM4, PwC5, and Deloitte6.
AI Adoption in 2021
We’ll go into detail about the findings of these reports, but unsurprisingly, all indicate an acceleration in the adoption of artificial intelligence among all industries surveyed. While not all industries have reached a significant use of AI in production, an impressive number of industries now have a productive level of AI maturity – and activity and interest in it has only increased.
After reviewing the data from these 5 major researchers, our analysis and synthesis of the data tells us that roughly 25% of companies have mature AI products in production in 2021.
Let us explain…
Larger Companies Invest More, Smaller Companies See ROI
While the O’Reilly and PwC studies revealed that a wider range of companies of various sizes and in diverse industries are investing in mature AI, the IBM and KPMG findings indicated slightly lower maturity of AI across the spectrum but with higher maturity levels occurring among companies with larger than $1B in annual revenue.
It may come as no surprise that larger companies with more revenue are investing more into their AI initiatives and are leading the charge in rolling out AI capabilities across their enterprises. What is interesting is that smaller companies (by headcount) are more likely to report their AI implementations as succeeding – with 88% claiming that AI is “at least moderately to fully functional” within their organizations, as opposed to 75% of larger companies claiming the same (IBM). Similarly, smaller organizations are more likely (67% of small vs. 57% of large companies) to report that AI initiatives have returned more than what they initially expected (KPMG).
Retail & Financial Services Lead in AI Maturity
With 40% and 38% of respondents, respectively, in the Retail and Financial Services sectors stating that their companies have AI in production (O’Reilly), these are the clear industry leaders in AI implementation. It’s not shocking to see that large industrial manufacturing firms also have a strong foothold in AI, but we were somewhat surprised to learn that healthcare and Life Sciences are the industries lagging. For larger companies across the Industrial Manufacturing, Financial Services, Tech, and Retail sectors, over 80% of respondents have AI capabilities that are at least moderately functional, if not fully functional, at scale (KPMG).
Benchmarking AI Adoption
Looking beyond mere adoption, Deloitte applied some simple benchmarking measures across all of the companies it surveyed that have AI implemented in their organizations. The benchmarking exercise categorizes AI adopters into four quadrants based on the number of AI applications they have deployed and the outcomes they’ve achieved:
- Starters (low deployment, low level of outcomes achieved)
- Pathseekers (Low deployment, high level of outcomes achieved)
- Underachievers (High deployment, low level of outcomes achieved)
- Transformers (High deployment, high level of outcomes achieved)
In the upper two quadrants sit the Pathseekers and the Transformers – companies that have undoubtedly invested in AI within their organizations (in either or both quality and quantity). However, if you find yourself sitting with the 17% of companies who are Underachievers, you’re showing a lot of heart, but your execution may be what’s off. Connect with us – we can help with that.
Recruiting Skilled AI Talent Still a Challenge
It is unsurprising that professional services firms would cite finding skilled talent as the top challenge to adopting AI. Supporting that viewpoint, however, we have that a whopping 39% of respondents in the IBM study cited talent as the biggest barrier to adoption, and this challenge was raised in studies executed by Deloitte and PwC as well.
While having a qualified and talented team in place is necessary for a strong foundation for any AI initiative, there are other roadblocks that Executives need to identify and address as well. The findings of the O’Reilly study places the talent challenge in context and illustrates the breadth of issues executives must be aware of:
- Lack of skilled people or difficulty hiring (19%)
- Lack of data or data quality issues (18%)
- Difficulties identifying appropriate use cases (17%)
- Company culture doesn't recognize need for AI (14%)
- Technical infrastructure challenges (12%)
If you’re a seasoned professional when it comes to process improvements and technological innovation, those 5 key areas of concern will not come as a shock to you. Hiring staff skilled in the new technology is always a challenge, at first, but eventually resolves between hiring eager new graduates and cross-training existing staff. Ultimately though, company culture, internal processes, and change management will play integral roles in your organization’s success with Artificial Intelligence.
What’s All the Hype About?
No article on technology adoption is complete without a look at the relevant Gartner Hype Cycle. With innovative technology arising constantly, how can one discern between “the hype” and a product that is commercially viable and, ultimately, useful for organizations and society?
As we close out 2021, we can see that hip new technology like Transformers, Knowledge Graphs, and Edge AI are nearing the “Peak of Inflated Expectations” and, AGI7 is still but a glimmer in our collective imagination.
We’re pleased to see, though, that core technologies such as Deep Learning, NLP, and General Machine Learning are descending from their hype peaks and settling into the realm of real use. As this occurs, we may hear less about them, but will likely see increased productivity from them before they reach the “Plateau of Productivity.”
So, as you consider the viability and potential for AI in your organization, keep in mind that you aren’t alone out there. The pace of adoption within your industry may be rapid even if you haven’t begun to look at it. The value of AI clearly exists – and a sound strategy is important, but as always it is execution that really makes the difference. To ensure that your company launches and executes against your AI opportunities seamlessly and effectively, connect with our team. We’ll get you started on the right foot.
- I actually like fruitcake, I think its reputation is unearned.
- O'Reilly, AI Adoption in the Enterprise 2021
- KPMG, Thriving in an AI World
- IBM, Global AI Adoption Index 2021
- PwC, AI Predictions 2021
- Deloitte, State of AI in the enterprise, 4th edition
- Artificial General Intelligence — true thinking machines
