PwC Uses a Giant Study to Market Its AI Expertise
June 24, 2026
Another dinobaby post. No AI unless it is an image. This dinobaby is not Grandma Moses, just Grandpa Arnold.
I can visualize the process of creating the “data” required to support “AI Reshapes Global Labor Market into Two Distinct Paths, Rewarding Human Skills: PwC 2026 Global AI Jobs Barometer.” The title sounds like a revolutionary pamphlet from a rabid supporter of an MBA program. The content is — well — interesting.
Thanks, MidJourney. Good enough, just like AI on June 18, 2026.
The main idea in my opinion is, “Hire PwC to redesign your firm’s talent strategy.”
The write up does not use direct, clear language. Instead, I highlighted this passage:
At the entry level, AI appears to be increasing demand for more ‘senior’ skills from junior workers. Based on 2.4 million entry-level jobs analyzed in the US, entry-level roles most exposed to AI are now seven times more likely to require traditionally senior-level ‘human-intensive’ skills like leadership, creativity or face-to-face interactions. Job openings for these ‘seniorized’ entry-level roles have grown 35% since 2019, while other entry-level roles shrank 10%.
I am tempted to work through the write up and identify the remarkable jargon and the fuzzy-wuzzy wordsmithing. What I want to do is highlight a few points and then conclude with a clear statement of what AI in the middle of 2026 is. I will exclude the fact that AI marketing baloney is one of the principal outputs of this “next big thing” sector.
PwC-type of AI will handle data aggregation from multiple sources. However, AI is not adept at credibility assessment. AI can monitor routine business transactions and handle some anomaly detection tasks in workflows. Chasing down the issue and figuring out the actors and the motive are not something the AIs my team has test can handle. AI systems can do consulting and meeting preparation grunt work; that is, looking up information, picking out “items” that appear with some statistical significance, and writing a paragraph or punching out dot points for a PowerPoint slide. AIs are not too swift at generating original information. Even the publicized “solved a big math problem” seems to have been a one off, but I expect more mathy stuff as the technology for this type of problem solving is refined. AIs can spit out clusters, link analyses, and “surprise” relationships. However, AIs can do the judgment work that differentiates useful observations from word salad with jargon toppings. AI can identify patterns in large datasets. At this time, a pattern is not what’s needed to handle certain complex but routine challenges; for example, what’s our pricing strategy for the attendees at the conference where our senior partner is giving a one hour lecture? (Senior partners have been known to just offer a deal to land a whale-scale engagement. No AI wanted or even considered. Just kill the whale and take the ambergris.)
The PwC AI jobs report does one thing well: It uses ambiguous but significant framing to create anxiety without destroying a potential client’s hope. If I worked at PwC, I would have included an offer for a free T shirt emblazoned with “I can be a super star with AI’s help.”
Net net: PwC is trying to say that it can remediate an organization’s “human resource” malfunctions. It’s not just automation. It’s a new way to win in the post-AI world. (This assumes that the AI will not hallucinate, compromise organizational security, or deploy defective agentic chains.)
Stephen E Arnold, June 24, 2026
Comments
-
- Subscribe to Beyond Search
Feature archive
News archive
-
