AI in the Workplace

The show's compiled thesis on AI's arrival in work, recompiled from the full ai-in-the-workplace corpus (170 segments) on 2026-08-14. Every claim cites its source segment at its public agents-site URL.

What AI changes about work

AI is a profound, internet-scale shift. Oliver Yonchev says AI will revolutionise every profession over the next 20 years in a shift comparable to the internet, and that ChatGPT-4 passing the bar exam 90% of the time means "these models are the worst they will ever be"; Al notes that after 1,000 days ChatGPT hasn't caused an unemployment tsunami but is eliminating data-entry-style jobs as people move up a step (https://agents.truthliesandwork.com/segments/38-oliver-s-take-on-ai-and-the-future.md, https://agents.truthliesandwork.com/segments/232-1000-days-of-chatgpt-transforming-work-or-a-cheat-tool.md).
AI is an amplifier, not a replacement. Jeffrey Fermin views AI as an amplifier, not a replacement, and that employee relations will always need a human in the loop; Michael compares LLMs to the spreadsheet — making existing employees more effective rather than killing jobs; Jodie Hill says her firm integrates AI to become more profitable so staff work fewer hours (https://agents.truthliesandwork.com/segments/285-ai-as-an-amplifier-not-a-replacement.md, https://agents.truthliesandwork.com/segments/64-michael-on-the-future-of-work.md, https://agents.truthliesandwork.com/segments/73-integrating-ai-into-service-delivery.md).
Jobs are made of tasks, and some tasks are more automatable. Andrew Palmer cites Ethan Mollick's "impossibly bundled jobs" — AI will nibble away at some tasks while front-office human interactions and judgement are defences; the World Economic Forum predicts 92 million jobs obsolete by 2030 but 170 million new ones created (https://agents.truthliesandwork.com/segments/277-jobs-are-made-of-tasks-and-some-tasks-are-more-automatable-than-others.md, https://agents.truthliesandwork.com/segments/192-21-job-titles-predicted-to-disappear-by-2030.md).
It reshapes what work means. Che Ugwuala says society will have to rethink what it means to be human and to have a job as AI automates much of current work; Danny Wareham warns AI tools display dark-triad behaviours (narcissism, manipulation, confident fabrication), making them worrying for informal therapy (https://agents.truthliesandwork.com/segments/LIVE-2026-03-25-dystopia-or-utopia-what-does-automation-free-us-for.md, https://agents.truthliesandwork.com/segments/243-ai-the-dark-triad-and-the-future-of-work.md).

The hidden adoption of AI

Workers use AI covertly. Al reports over 40% of professionals used ChatGPT at work, 70% without telling their bosses; Microsoft research finds 52% of workers using AI are hesitant to disclose it, fearing they appear replaceable — the "hidden AI revolution" (https://agents.truthliesandwork.com/segments/33-chatgpt-s-new-capabilities.md, https://agents.truthliesandwork.com/segments/99-chatgpt-the-hidden-ai-revolution.md).
It can add load, not just remove it. Al reports an Upwork survey finding 77% of employees say AI has added to their workload and 47% don't know how to achieve the expected productivity gains; Leanne defines "botsitting" as the labour of making AI usable — feeding content, supervising output and cleaning up after it (https://agents.truthliesandwork.com/segments/254-story-2-the-rise-of-ai-xiety.md, https://agents.truthliesandwork.com/segments/316-word-of-the-week-botsitting-and-botshitting.md).
Adoption is slower than the hype. Al argues AI is not just hype but cultural adoption is slow, and jobs will change in unexpected places rather than be eliminated quickly; Rob Kalwarowsky cites an MIT study that 95% of AI projects didn't add value (https://agents.truthliesandwork.com/segments/157-the-ai-hype-cycle-and-what-s-really-changing.md, https://agents.truthliesandwork.com/segments/257-ai-disruption-friction-nonsense-and-data-privacy.md).

AI, bias and fairness

AI can automate bias, not fix it. Nora Burns warns AI could worsen hiring bias through a garbage-in-garbage-out dynamic; Ryan Sherman notes Amazon's recruiting algorithm kept selecting men and video-interview scores shifted with appearance changes; Al and Leanne call AI video interviews like HireVue flawed, citing discrimination risks for neurodivergent candidates over eye-contact scoring (https://agents.truthliesandwork.com/segments/161-how-ai-could-make-hiring-bias-worse.md, https://agents.truthliesandwork.com/segments/49-trend-7-technology-and-ai.md, https://agents.truthliesandwork.com/segments/97-ai-in-recruitment-a-flawed-future.md).
Design can counter the bias. Kate Young of Sapia says her tool cleans job descriptions for bias, uses a human-in-the-loop layer, gives every candidate the same questions for fairness, and doesn't do automated pass/fail — cutting time-to-hire by 70–80%; job analysis matters even more in the age of ChatGPT because measuring what matters reduces bias (https://agents.truthliesandwork.com/segments/287-how-ai-condenses-8-hours-into-90-minutes.md, https://agents.truthliesandwork.com/segments/287-why-it-s-not-a-chatbot.md, https://agents.truthliesandwork.com/segments/287-is-job-analysis-still-relevant-in-the-age-of-chatgpt.md).
Monitoring AI can backfire. Leanne cites a study of 27,000 European workers finding AI performance monitoring increases stress, reduces teamwork and raised workplace accidents by 70% (https://agents.truthliesandwork.com/segments/178-story-3-ai-performance-monitoring-productivity-boost-or-micromanagement-nightmare.md).

AI, leadership and human skills

AI breaks the knowledge hierarchy. Al argues AI makes a junior employee as capable as a leader, so leaders must earn their place through influence and trust; leadership becomes relationship-building, listening and flexibility — the "soft skills" reframed as core skills (https://agents.truthliesandwork.com/segments/192-more-from-sophie-leadership-in-an-ai-world.md).
Human connection is the durable differentiator. Matt Poepsel argues 1960s–70s leadership frameworks are outdated and the long-term competitive differentiator is the human parts AI can't replicate; Tom Emery says organisations that humanise digital change best will succeed, and AI can handle simple HR tasks but not complex human issues requiring judgement and connection; Charli Nordone says storytelling and asking AI questions beats letting it write for you (https://agents.truthliesandwork.com/segments/265-enlightened-leadership-and-the-role-of-human-connection-in-the-age-of-ai.md, https://agents.truthliesandwork.com/segments/321-humanising-workplace-change-in-the-age-of-ai.md, https://agents.truthliesandwork.com/segments/LIVE-2026-05-01-authenticity-and-the-ai-vs-ai-loop.md).
Leaders should be "happily confused" and human-first. Andrew Palmer advises leaders to be "happily confused" about AI and experiment; Owen Fitzpatrick says frame AI adoption by painting a vision of employees becoming indispensable experts rather than promising easier jobs; the show notes banks are hiring behavioural scientists for AI because the biggest risks are behavioural, like overtrusting systems (https://agents.truthliesandwork.com/segments/277-be-happily-confused-advice-for-leaders-who-don-t-know-how-to-feel.md, https://agents.truthliesandwork.com/segments/323-leading-through-ai-and-future-change.md, https://agents.truthliesandwork.com/segments/270-story-3-why-banks-are-hiring-behavioural-scientists-for-ai.md).
Automation can strip the human out of work. Peter Dorrington says automation strips simple tasks but leaves more stressful creative firefighting, removing employees' decompression time and driving burnout; Sean O'Shea warns choosing a chatbot over a colleague eliminates the "messy moments" that drive team performance (https://agents.truthliesandwork.com/segments/98-ai-and-automation-the-double-edged-sword.md, https://agents.truthliesandwork.com/segments/309-cold-open-the-ai-connection-warning.md).

AI and the future of management

AI will reshape management, messily. Andrew Palmer worries AI will be used to slim the middle-management layer, reinforcing existing problems; Vince Sanderson predicts 2026 will bring a messy phase as AI replaces distinct management skills, frustrating managers and team members who lose human connection; Kristien Turner predicts a severe management-skills crisis by 2030 as firms shed middle managers holding key soft skills (https://agents.truthliesandwork.com/segments/217-ai-and-the-future-of-management.md, https://agents.truthliesandwork.com/segments/263-prediction-4-ai-will-break-management-before-it-fixes-it-vince-sanderson.md, https://agents.truthliesandwork.com/segments/311-the-coming-skills-gap-and-the-gen-z-shift.md).
The best AI adopters are the best managed. Leanne argues companies getting results from AI aren't the most tech-savvy but the best managed; Andrew Palmer notes early generative-AI uses like internal chatbots are mundane, and coaching is the most developed experiment — an AI sales coach that often misreads calls still drove a 50% jump in sales productivity (https://agents.truthliesandwork.com/segments/277-generative-ai-vs-the-ai-that-came-before.md, https://agents.truthliesandwork.com/segments/277-intro-the-ai-sales-coach-that-gets-things-wrong.md).
AI literacy is becoming a leadership skill. Al notes Google's Generative AI Leader certification signals leaders now need to know how to use generative AI; Becky Simms predicts candidates will treat AI access as a dealbreaker, asking what AI tech stack a company will provide; Al reports Google will let candidates use AI in interviews to evaluate AI fluency, since 75% of its new code is AI-generated (https://agents.truthliesandwork.com/segments/276-story-2-google-s-generative-ai-leader-certification.md, https://agents.truthliesandwork.com/segments/263-prediction-7-ai-will-become-a-dealbreaker-for-candidates-becky-simms.md, https://agents.truthliesandwork.com/segments/308-google-embraces-ai-in-job-interviews.md).