Reading time: 5 minutes

There are too many promoters and not enough clear-eyed analysts in the world today. I have been an exuberant promoter of advances in the science behind AI for the last several years. But the data and conversations with large numbers of executives at firms worldwide have convinced me that the people promoting claims of rapid AI adoption are out of touch or working other agendas.

Some promoters are far too cavalier with the data. Many others could care less about the reality. It’s far easier to succeed in this world saying “rah, rah, rah.” If nothing else, it attracts well heeled advertisers to spend more and more trying to inflate the enthusiasm bubble.

A few weeks ago, an AI newsletter published by a major business news source highlighted a quote that claimed 40 percent of firms are implementing AI. Do you believe that?

  • If you clicked through to read the article behind the headline, you would have learned that 40 percent of attendees at the AI World event in Boston (December, 2018) said their firms are implementing AI already.
  • If you didn’t click through to read the article, you might have assumed the headline was accurately describing the state of the market for AI.

Pity.

The AI World data is good, but it only represents the opinions of people who paid to be at a conference on AI. How would people who didn’t go to that conference have voted? Sample bias is a problem with a lot of marketing claims. Proponents of a position seek out data that supports their position (and their employer’s objectives.)

Other data tells me that

Only five percent are investing heavily

We’ll get to that other data shortly, but first, let’s look at another quote from the same newsletter this week. Citing an article published elsewhere (ostensibly a “C-level executive survey”), the newsletter said:

Companies are accelerating their use of AI as they fear being outflanked by more nimble data-driven competitors. Survey findings included:

87.8% of respondents said the urgency to invest in big data and AI initiatives is greater than it’s ever been.

96.4% of businesses are investing in AI and machine learning.

21.1% of businesses are investing more than $500 million in big data and AI initiatives.

62.2% of respondents report favorable results for their AI and big data investments.

When you read the referenced study, it’s pretty clear that the authors are conflating big data and AI. Not all big data projects involve AI and not all AI involves big data. (It’s also curious that the authors went so far as to include numbers to the right of the decimal point. That suggests naïveté, not precision. It also suggests limited understanding of which numbers are significant and which aren’t.)

Reports claiming huge growth in enterprise AI-technology investments are unbelievable. The more they hyperventilate, the more you should ignore them.

Reality

  • In their 22nd Annual Global CEO Survey with 1378 respondents, Price Waterhouse Coopers (PwC) found only five percent said their firms used AI on a wide scale.
  • Daniel Faggella at emerj has found similar results.

Large numbers of enterprises are experimenting with AI. This has been true for years. They seek to learn more about the potential “magic” inside the hat. Few are strategically investing (particularly in the US.)

There are a lot of investments in small pilots. CEOs have been telling their CIOs that the board wants to know what they’re doing with AI. The answer that comes back? CIOs say

Yes, we’re using AI and no, we’re at no risk because the investment is small and well controlled.

A typical, low risk project might be chatbots installed in the internal end-user help desk portal.

Again, from PwC’s 22nd Annual CEO Survey, they report:

85% of CEOs agree that AI will significantly change the way they do business in the next five years.

PwC claims this is a striking finding. Indeed it is.

Agreement with that statement could mean CEOs expect they will have to get to AI in five years. Why? Because it is not happening now, not in a significant and serious way. But given the din of hype around AI, the CEOs might be expecting to finally see enough business benefits to invest heavily in a few years after addressing their current (non-AI related) strategic initiatives.

This five year stall — we’ll get to it someday — is striking. CEOs are implicitly deferring AI heavy investments far into the future.

The next figure is from the same PwC CEO survey.

Take a good look at the “Global” bar at the top of the chart. Three percent (three, not thirty or ninety) say “AI initiatives are present on a wide scale in our organization” and six percent say “AI initiatives are fundamental to our organization’s operations.” That adds up to nine percent!

Pay no attention to the yellow bar sections “We have introduced AI initiatives in our business, but only for limited uses.” That’s experimenting, prototyping and pilots. As PwC said, “firms sticking their toe in the water.”

Now look at the ‘North America’ bar. Five percent total. F-I-V-E. Got it?

AI Versus Internet

PWC also says

Close to two-thirds of global CEOs see [AI] as bigger than the internet.

Study this chart from the same survey.

In the US, CEOs believe the internet will have a bigger impact on the world than AI. Now compare the US data to China, India, Germany and Japan. CEOs from those four countries feel dramatically different from their US counterparts. They believe AI will have a far greater impact than the Internet.

What does this mean?

We believe that divergence represents the heavy investment US firms have already made in integrating their assets to the Internet while at the same time, feeling under threat from Amazon and others who have figured out how to change the fundamentals of market coordination between various entities.

This is a major threat situation for firms in the US and, to a lesser extent, elsewhere. Many CEO level initiatives are all about the threat these internet-based companies represent. Amazonization is an existential threat. And they don’t see any major quick fixes from AI (or anywhere else) to thwart these new, far more agile competitors who have thrown out the rulebook in many cases.

What’s needed in the market are a range of solution platforms that will exploit many technologies (including but not only AI), helping most entities reinvent themselves for this new age.

(Don’t let this stop your organization from tactically exploiting technologies like AI to enhance your current systems and processes. And don’t assume this makes you safe from ‘apex internet predators’ (apex predators) who are among the leading edge top five percent who are really investing strategically in major enterprise business initiatives in fulfillment of their Amazonization mission.

For more executive perspectives, see CEOs are seeking major business initiatives, not AI or other tech giblets. And I will cover apex predators in a subsequent blog post.

In the meantime, do you see platforms (and ecosystems supporting them) really ready to help firms adapt to and thrive in this new environment?

Let me know.

Disclaimer: This post is my own opinion. It’s not a paid post. I wrote it myself and I have no affiliation with any of the entities mentioned above.

Note: I am intentionally not analyzing the different potential outcomes for the various nations in the second graphic. I’ll leave questions like “China vs. US” to later work. I am also not going to call out the specific newsletters (hype sources) referenced in this post. If we were to catalog all the examples of bad data and bad analysis, there would be scant time to present the right data and analysis. If you want to know who it was this time, ask me and I’ll send it to you.

[Headline and end-notes edited for clarity, 4 Feb 2019.]