IBM AI Study: Would The Research Report Get an A in Statistics 202?
May 9, 2025
No AI, just the dinobaby expressing his opinions to Zellenials.
IBM, reinvigorated with its easy-to-use, backwards-compatible, AI-capable mainframe released a research report about AI. Will these findings cause the new IBM AI-capable mainframe to sell like Jeopardy / Watson “I won” T shirts?
Perhaps.
The report is “Five Mindshifts to Supercharge Business Growth.” It runs a mere 40 pages and requires no more time than configuring your new LinuxONE Emperor 5 mainframe. Well, the report can be absorbed in less time, but the Emperor 5 is a piece of cake as IBM mainframes go.
Here are a few of the findings revealed by IBM in its IBM research report;
AI can improve customer “experience”. I think this means that customer service becomes better with AI in it. Study says, “72 percent of those in the sample agree.”
Turbulence becomes opportunity. 100 percent of the IBM marketers assembling the report agree. I am not sure how many CEOs are into this concept; for example, Hollywood motion picture firms or Georgia Pacific which closed a factory and told workers not to come in tomorrow.
Here’s a graphic from the IBM study. Do you know what’s missing? I will give you five seconds as Arvin Haddad, the LA real estate influencer says in his entertaining YouTube videos:
The answer is, “Increasing revenues, boosting revenues, and keeping stakeholders thrilled with their payoffs.” The items listed by IBM really don’t count, do they?
“Embrace AI-fueled creative destruction.” Yep, another 100 percenter from the IBM team. No supporting data, no verification, and not even a hint of proof that AI-fueled creative destruction is doing much more than making lots of venture outfits and some of the US AI leaders is improving their lives. That cash burn could set the forest on fire, couldn’t it? Answer: Of course not.
I must admit I was baffled by this table of data:
Accelerate growth and efficiency goes down with generative AI. (Is Dr. Gary Marcus right?). Enhanced decision making goes up with generative AI. Are the decisions based on verifiable facts or hallucinated outputs? Maybe busy executives in the sample choose to believe what AI outputs because a computer like the Emperor 5 was involved. Maybe “easy” is better than old-fashioned problem solving which is expensive, slow, and contentious. “Just let AI tell me” is a more modern, streamlined approach to decision making in a time of uncertainty. And the dotted lines? Hmmm.
On page 40 of the report, I spotted this factoid. It is tiny and hard to read.
The text says, “50 percent say their organization has disconnected technology due to the pace of recent investments.” I am not exactly sure what this means. Operative words are “disconnected” and “pace of … investments.” I would hazard an interpretation: “Hey, this AI costs too much and the payoff is just not obvious.”
I wish to offer some observations:
- IBM spent some serious money designing this report
- The pagination is in terms of double page spreads, so the “study” plus rah rah consumes about 80 pages if one were to print it out. On my laser printer the pages are illegible for a human, but for the designers, the approach showcases the weird ice cubes, the dotted lines, and allows important factoids to be overlooked
- The combination of data (which strike me as less of a home run for the AI fan and more of a report about AI friction) and flat out marketing razzle dazzle is intriguing. I would have enjoyed sitting in the meetings which locked into this approach. My hunch is that when someone thought about the allegedly valid results and said, “You know these data are sort of anti-AI,” then the others in the meeting said, “We have to convert the study into marketing diamonds.” The result? The double truck, design-infused, data tinged report.
Good work, IBM. The study will definitely sell truckloads of those Emperor 5 mainframes.
Stephen E Arnold, May 9, 2025
Comments
Got something to say?
-
- Subscribe to Beyond Search
Feature archive
News archive
-
