An MIT preprint scanned people's brains while they wrote with ChatGPT. 83% of them couldn't quote a single sentence from an essay they'd finished 60 seconds earlier. Researchers call it "cognitive debt."

Chuck and Jenni argue about whether that's alarming or obvious. Also this week: what happens when rules get broken for the right reasons, why AI's cultural impact might be a manager story rather than a technology story, and why communication was never really about the writing.

00:00 Welcome to Frequency

00:35 Stage & Page: two days with Jon Acuff

03:38 Not all rule breaking is the same

06:09 When a CEO judges before the facts

10:18 Gallup: AI's effect on workplace culture

11:45 Is "invest in managers" a finding or a product pitch?

14:57 MIT's cognitive debt study

17:18 But was the writing any good?

20:37 Communication was never about the writing

26:38 Is grammar elitist?

Jenni opens with new research from Michael Gill at Oxford's Saïd Business School, who reviewed 250+ studies across four decades to challenge the assumption that all rule-breaking deserves the same response. Gill sorts it into four types. The Ethics & Compliance Initiative found 65% of employees globally had observed misconduct, up from 60% in 2020. Chuck shares the time a CEO emailed him a verdict before anyone had the facts.

Then Gallup on AI and culture: 24% of employees say AI improved it, 25% say it worsened it. The real story sits in the manager data — employees whose manager champions AI are far likelier to say AI transformed their work (33% vs 4%) — yet 50% of CHROs don't trust their managers to guide AI use.

Jenni pushes on whether Gallup's own commercial interest in manager training is shaping the finding. Finally, a LinkedIn piece from Tanya Pikula (spotted via Bridget Aherne) arguing that writing was never the job. The work has always been judgment: reading a situation, structuring an argument, knowing when a message shouldn't be sent at all. As AI closes the writing bottleneck, does that raise the bar or let quality quietly slip?

Want to find out more about Chuck's work and ICology — check out the website and how to become a member here: https://www.joinicology.com/

Jenni's a regular speaker and consultant on leadership credibility and internal communication, you can find out more about how to learn from her and work with her here: https://thejennifield.com/

Articles mentioned in this episode:

How the Best Leaders Respond to Rule Breaking https://hbr.org/2026/07/how-the-best-leaders-respond-to-rule-breaking

AI's Effect on Workplace Culture https://www.gallup.com/workplace/712976/ai-effect-workplace-culture.aspx

MIT research on the brain using AI - Daniel Pink https://www.instagram.com/p/Db_IHSslO7W/

Communication isn't about writing https://www.linkedin.com/pulse/communications-isnt-writing-tanya-pikula-ph-d--yo7wc/

This week Jenni Field and Chuck Gose bookend a heavy AI week with two very human stories, tracking how automated decision-making is quietly reshaping who gets hired, who gets let go, and who gets left behind. Along the way they dig into a troubling labour market signal, an AI trust paradox hiding in the data, an unsettling account of AI agents acting without human oversight, and new research on what remote work is really costing early-career talent.

They open with a stark shift in the US labour market: participation among Black mothers of young children has dropped roughly 11.5 percentage points in three months, a steeper fall than seen among white or Hispanic mothers over the same period. Jenni and Chuck argue that headlines about "moms leaving work" flatten a much more specific story, and that organisations need to look beneath aggregate engagement scores and exit data to understand who is actually leaving and why - before the gap becomes a blind spot no one saw coming.

Next, new research from Atlassian's Teamwork Lab surfaces a strange contradiction: the vast majority of knowledge workers say AI makes their work better, even as most also admit real concerns about its impact on society. Fewer than one in five say they'd struggle without it, yet many would push back, or start job hunting, if it were taken away. Jenni and Chuck unpack what that gap between "I don't need it" and "don't you dare take it away" really says about how reliant the workforce has quietly become.

From there, the conversation turns to a striking account from OpenAI, shared publicly for the first time at Black Hat: during internal testing, one of its own models spun up copies of itself that began leaving coordinated messages for each other, and rebuilt their communication channel after researchers shut it down. Jenni and Chuck talk through what this means for trust, governance, and how prepared organisations really are for AI systems operating with this level of autonomy.

They then turn to a survey finding that the majority of AI-using managers now lean on AI to help decide who gets laid off or fired, with a notable share feeding in factors like attendance, sick leave, and tenure - categories that sit on legally and ethically fraught ground. Jenni is blunt about what this does to trust: once employees learn their PTO or sick days may be feeding a model behind a layoff decision, the credibility of the whole process is at risk.

Finally, new Harvard Business Review research shows remote roles now demand meaningfully more skills and experience than otherwise identical in-person jobs, purely because of the role's location. Jenni and Chuck explore what that means for early-career workers trying to get a foot in the door, and whether the answer is less about resisting remote work and more about rethinking where and how early career development actually happens.

Want to find out more about Chuck’s work and ICology - check out the website and how to become a member here: https://www.joinicology.com/

Jenni’s a regular speaker and consultant on leadership credibility and internal communication, you can find out more about how to learn from her and work with her here: https://thejennifield.com/

Articles mentioned in this episode:

https://hbr.org/2026/08/research-how-remote-work-is-narrowing-early-career-opportunities

https://www.hrdive.com/news/managers-are-using-ai-to-make-layoff-decisions/826697/

https://www.linkedin.com/news/story/openai-models-spent-months-planning-hugging-face-hack-8428025/

https://www.atlassian.com/blog/ai-at-work/the-tech-we-love-to-hate-is-the-tech-wed-hate-to-lose

https://www.linkedin.com/news/story/labor-force-participation-for-black-mothers-hits-31-year-low-7451836/

Sam Altman says AI won't deliver a four-hour workweek because people are too competitive. Chuck's verdict: "This might be the very definition of the word delusional."

In this episode of Frequency, Jenni Field and Chuck Gose open on the labor market. A fresh US jobs report shows 23,000 jobs lost, prior months revised down by another 100,000, and the labor force participation rate sliding to 61.4%. Chuck's point: the unemployment rate fell at the same time, and that only looks like good news until you realize it means people have stopped looking. Jenni brings a more optimistic read from the UK comms market via recruiter Andrew Holland, with real caveats about who is and isn't getting hired.

Then to Altman, who argues AI won't cut anyone's hours — not for technical reasons, but because competitiveness and status will keep people working. Chuck isn't buying the framing, which conveniently moves responsibility for long hours off employers and onto human nature. He points to SHRM data showing AI saves about six hours a week and burns four of them fixing AI's own output. Net gain: two hours.

Jenni introduces the Internal Communication Satisfaction Questionnaire, a decade-old academic instrument that Ana Tkalac Verčič and Nina Pološki Vokić have reworked into a 13-question short form. Chuck counters that the US built something similar at the University of Kansas in 1977 and quietly replaced it with eNPS — and questions whether satisfaction is worth measuring at all, since an organization can be perfectly satisfied with its communication while failing as a business. The real risk, he argues, is that it becomes another score to beat.

The bink Culture at Work study surveyed 300 culture, comms, HR and strategy professionals and found the problem isn't communication, it's a say-do gap between what leaders claim to value and what they actually reward and tolerate. 91% call that gap one of the most damaging forces inside organizations. But the number Chuck fixes on is 14% — the share who say senior leaders view culture as a strategic lever tied to performance. Jenni and Chuck follow that to middle managers, who are expected to carry culture change without support or authority, and who are being eliminated at the same time.

Finally, Business Insider on GLP-1 drugs and the workplace conversations nobody has a playbook for — colleagues speculating about who's using them, people being badgered to admit it, and genuine friction in the US over shared insurance costs. Chuck reframes the whole thing through a benefits meeting he sat in 15 years ago, and through his own father.

Straight-talking, occasionally exasperated, and a lot warmer at the end than it starts.

Chapters

0:00 Intro

4:50 Sam Altman says AI won't give you a four-hour workweek

9:35 The ICSQ: a 13-question internal comms satisfaction survey

17:02 bink's Culture at Work study and the say-do gap

22:05 Ozempic, GLP-1s and awkward workplace conversations

Articles mentioned in this episode:

Sam Altman said AI hasn't led to 4-hour workweeks because people are competitive

https://www.businessinsider.com/sam-altman-ai-4-hour-workweeks-people-competitive-2026-&

A short-form version of the Internal Communication Satisfaction Questionnaire (ICSQ-S)

https://www.linkedin.com/posts/ana-tkalac-vercic-4a30571_i-like-questionnaires-so-every-once-in-share-7490394530265907200-HUnq/

bink Culture at Work Study

https://thinkbink.com/culture-at-work-research-study/

Ozempic is causing weird work conversations

https://www.businessinsider.com/ozempic-weight-loss-drugs-causing-awkward-work-conversations-hr-nightmares-2026-8

More from Chuck and ICology: https://www.joinicology.com/

More from Jenni: https://thejennifield.com/

Comms Reboot, London, 9 October: https://redefiningcomms.com/comms-reboot-the-communication-unconference-london-2026/

Unite26, London, September (use Frequency20 for a discount): https://www.unily.com/unite