Musical accompaniment for this week’s newsletter. I was talking to Dan Riley of RADICL the other day for an upcoming podcast episode when (not surprisingly) the topic of AI came up. I feel a little bit like a broken record in this newsletter because it feels like it’s all about AI all the time. And it is. And this relentless conversation feels unavoidable.
I’m also a firm believer that AI adoption is not as pervasive as we in the tech community pretend it is. Fresh Gallup research points out that the frequent use of AI in the workplace continued to rise in Q4. This data shows AI use at work is no longer niche — about 46% of employees use it at least occasionally, and frequent use keeps rising even as overall adoption plateaus. Nonetheless, daily AI users still account for only 12% of the workforce. Leaders are way ahead of the rank-and-file, which tells you everything about where this is heading: AI is becoming a management tool before it becomes a worker tool.
What about agentic AI? Salesforce’s 2025 research with global CHROs shows that currently only 15% of organizations have fully implemented agentic AI, but adoption is poised to explode; CHROs project a 327% increase in agent adoption by 2027 (from 15% to around 64% adoption at that level).
What I think is being lost in the shuffle is the broader impact. There’s a big shift in moving from AI as a helpful autocomplete to AI as a semi-independent actor. It’s a different psychological contract. When an agent runs tasks in the background, makes decisions, and persists beyond direct supervision, it stops being a tool and starts being infrastructure. Or maybe even staff.
AI strategist Craig Hepburn nails the emotional undercurrent of this shift. His Clawdbot agent wasn’t just software — it started to feel like a colleague that didn’t sleep, didn’t forget, and didn’t really leave when shut down. When an agent is running around your apps and inbox, the blurry line between what you did and what your agent did starts to feel a little Black Mirror. That’s the agentic AI era in a sentence: autonomy without accountability feels creepy fast.
Hepburn also shared another valuable insight about AI agents: You are better off treating them as employees with their own accounts, systems, even machines. This sounds like a better approach, but be careful: Now agents have their own social network — because 2026 — and Polymarket thinks there’s a 70% chance an AI agent sues a human. What a time to be alive.
The workplace question isn’t “will agents replace jobs tomorrow?” It’s “how do we design trust, control, and human purpose into systems that increasingly behave like coworkers?”
Trust — and lack thereof — is the recurring theme in these conversations. And it’s bigger than just AI. It’s pretty much everything in the world of work and the world itself.
What else is going on this week?
So, the Wall Street Journal just discovered that Answer Engine Optimization is a thing and realized what everyone in marketing has been muttering for a year: Brands are now optimizing for answers, not rankings. Call it AEO, GEO, whatever — it’s the same game, just played inside ChatGPT, Gemini, Perplexity, and whatever AI sidebar Google ships next.
And the stakes are Big Time. Referral traffic from AI is already meaningful (more than 40% for some businesses, but generally far lower based on my experience). And those users convert harder. So, of course, people are trying to rig the slot machine.
OK, what does all this mean? I highly encourage y’all to read Lily Ray’s reflections on SEO, AEO & AI search. As she notes, AI search is real, but the fundamentals haven’t magically changed. Authority still wins. High-quality original content still matters. The web is still the training data. If your site is garbage, no amount of optimization fairy dust will fix it. And as Ray mentions, there is an astounding amount of AEO snake oil and BS out there. Caveat emptor.
Ray’s take dovetails with what we are seeing at Rep Cap. There is a flight to quality content over the commodity AI slop that the market is awash in. When you combine quality content with quality earned media (PR/AR), GTM ops excellence, and rigorous AEO, you have a solid foundation for what’s coming next. If you can be a source the models trust, you are most of the way there.
G2 decided to run the table on review sites after acquiring Capterra, Software Advice , and GetApp from Gartner. If you sell enterprise software, this is a massive GTM wake-up call. Basically, G2 just bought the front door to your category. This deal consolidates four of the biggest buyer-intent engines in B2B software under one roof — something like 6 million reviews and 200M+ annual buyers in one dataset.
This consolidates the biggest buyer-intent and review channels in B2B software into one platform — meaning software discovery is now more centralized and likely more pay-to-play. For Work Tech marketers, it’s also an AEO move: reviews aren’t just social proof anymore, they’re training data for AI-driven recommendations — right as AI agents start answering “what tool should I use?” without ever visiting your website. See the item above.
It’s also a sign that Gartner is retreating to its research/advisory ivory tower as AI disruption is coming like the Night King’s hordes for their margins and their business.
AI product announcements are everywhere these days, but Businessolver’s launch announcement for its anticipatory benefits approach actually deserves your attention. I was fortunate enough to get a demo recently, and it’s truly game-changing functionality, not AI for AI’s sake. The difference is the true value it delivers to employees, helping them better use and maximize their benefits. There’s plenty of value on the employer and the benefits broker side. Find out more in this NelsonHall post from DeeAnna Warrington.
In the latest WorkTech Weekly episode, I talked with Claire McTaggart of SquarePeg about how keyword screening flattened resumes, why embellishment became rational, and why asking recruiters to stop fraud is like asking a pilot to fix an engine mid-flight. Oh, and where the best restaurants in Left Miami are. Plot twist: AI could actually help — not by replacing judgment, but by restoring lost context. Today’s hiring crisis isn’t a people problem. It’s an incentives problem. Change those, and the outcomes follow. LISTEN NOW