The Organizational Gravity Podcast

Politics has a way of finding its way into the workplace, whether we invite it or not.Following the 2026 midterm elections, Kelly and I sat down to talk about something that’s become increasingly difficult: having conversations with people who don’t necessarily agree with us.We started with the election results, the political pendulum, and the effects that changes in government have on business. Tariffs, regulations, project costs, and now artificial intelligence all have consequences that extend well beyond Washington. Those consequences eventually find their way into our organizations, our teams, and our conversations.But somewhere along the way, the discussion became less about politics and more about people.Regardless of how we voted, we still have to come back to work together.And that means finding ways to understand each other without needing to agree on everything.For years, we’ve treated politics, religion, and other personal subjects as things that shouldn’t be discussed at work. There are understandable reasons for those boundaries. But people don’t arrive at work without their beliefs, experiences, fears, or concerns. Those things influence how we communicate, how we interpret decisions, and how we relate to the people around us.This is part of what I call human middleware: the often invisible work people do to manage relationships, bridge differences, interpret intentions, and keep organizations functioning.It’s also a form of Organizational Gravity. The more tension, uncertainty, and misunderstanding people carry, the more effort it takes to accomplish even ordinary things.We talk about why creating room for difficult conversations matters, and why understanding someone is not the same as agreeing with them. A workplace can allow people to speak openly without turning every disagreement into a debate. Sometimes the most valuable thing a leader can do is listen, recognize where someone is coming from, and leave the conversation there.What happens when AI enters the conversation?Our discussion eventually turns to artificial intelligence, regulation, and the uncomfortable questions surrounding the technology’s future.We explore Anthropic’s recent IPO filings, including warnings about the potential for catastrophic or existential risks from sufficiently advanced AI systems. That leads to a discussion about what responsible governance should look like, and whether the pace of technological advancement is getting ahead of our ability to understand its consequences.But there’s a more personal question underneath all of it.What happens to the relationships that make work meaningful when we begin automating more of the work itself?Kelly raises something I think deserves much more attention: the value of human connection isn’t always something we can measure.We spend an extraordinary amount of time working with other people. We develop friendships, build trust, disagree, collaborate, and occasionally drive each other crazy. Those relationships become part of our lives, not simply an output of our jobs.AI may change how we accomplish work. It may reduce the time required for certain activities. But that doesn’t eliminate our need to connect with one another.And maybe that’s something worth protecting as we build the next generation of organizations.A few things we’ve gotten wrong along the wayIn our regular look at Organizational Gravity, I also talk about something more personal: my own failure résumé.Earlier in my career, I spent a lot of time focused on execution. Deliver the project, hit the budget, get the result. That approach was often rewarded, but it came with a cost I didn’t fully appreciate until much later.Sometimes the relationships mattered more than the result I was pushing so hard to achieve.Kelly shares a very different experience. Early in her career, someone told her she would never get anywhere because she was too nice. Rather than accept that advice, she made kindness and understanding central to how she worked with others.It’s a useful contrast between two approaches to leadership, and a reminder that being effective doesn’t require being harsh.We also talk about the people who help us recognize our own behavior. I share some recent feedback from Audrey about trying so hard to communicate a message that I was actually making it harder for people to hear.Sometimes saying less accomplishes more.That brings us to Jefferson Fisher’s communication advice: allowing yourself a breath before responding to something difficult. Simple advice, and apparently something I still need a lot of practice with.Naturally, we somehow manage to connect that lesson to an electric fence, a jumping cholla cactus, and a couple of parenting decisions that probably deserve their own episode.The thought we’re leaving withWe don’t have to agree on politics. We don’t have to share the same approach to leadership, technology, or organizational change.But we do have to find ways to work together.The relationships, conversations, trust, and occasional uncomfortable feedback that make that possible are easy to overlook because they don’t fit neatly into a project plan or a performance dashboard.They are human middleware.And as technology makes more of our work easier to automate, I suspect that very human work will become even more important.Have the conversations. Be willing to listen. Know when to walk away from a debate. And above all, be kind to the people you’re working with.In this episode* 00:24 — The midterm elections and why politics affects business* 03:24 — Political extremes, organizational change, and the J-curve* 09:55 — Politics at work, psychological safety, and human middleware* 14:16 — Human connection and what AI can’t easily replace* 18:22 — AI regulation, Anthropic’s IPO filings, and technology risk* 21:45 — My failure résumé and lessons from a career focused on execution* 24:17 — Kelly’s experience with being told she was too nice to succeed* 28:59 — Leadership feedback, communication, and knowing when to stop pushing* 30:25 — Parenting lessons involving an electric fence and a cactus* 32:50 — Jefferson Fisher, difficult conversations, and the power of a pause* 35:03 — Why difficult conversations are still worth havingSources and further reading2026 midterm election results: The Associated Press confirmed that Democrats won control of both chambers in the November 3 elections. The newly elected Congress will take office in January 2027. Associated Press — Calling Control of Congress in the 2026 Midterms.AI risks and accountability: Anthropic’s November 2, 2026 amended SEC registration statement discusses the possible catastrophic or existential risks of advanced AI, alongside operational, financial, and regulatory risks. Anthropic — SEC Form S-1/A.State AI legislation: The National Conference of State Legislatures tracks proposed and enacted AI legislation, including requirements around transparency, oversight, responsible use, and discrimination. NCSL — Artificial Intelligence Legislation Database.Psychological safety: Amy Edmondson’s research examines the relationship between interpersonal trust, team learning, and the ability to speak up without fear of embarrassment or punishment. Edmondson (1999), Psychological Safety and Learning Behavior in Work Teams.The importance of human connection: The U.S. Surgeon General’s advisory examines the relationship between social connection, health, and well-being. U.S. Department of Health and Human Services — Social Connection.Communication and conflict: Jefferson Fisher’s The Next Conversation: Argue Less, Talk More explores practical strategies for responding thoughtfully, navigating disagreement, and improving conversations. Jefferson Fisher — The Next Conversation.Organizational Gravity explores the forces that shape how organizations function, the complexity that builds along the way, and the people who make it all work.If this conversation resonated with you, subscribe for free, share it with someone you work… Get full access to Organizational Gravity at organizationalgravity.substack.com/subscribe

We keep saying AI will give us more time for the human side of work. More time to listen, build relationships, and have the conversations we keep putting off.I like that idea. I also wonder what makes us think we’ll use the time that way.We can fill it with more work. We can use it to take on another responsibility. And if we’re uncomfortable having a difficult conversation, we now have something that can write a very thoughtful message while we remain just as uncomfortable—and just as absent.This week, Colette Molteni joined Kelly and me to talk about empathy, efficiency, and the human work happening around our systems. I came into the conversation worried about what organizations might give up in pursuit of speed. Colette helped me sharpen the question: do we understand which effort we’re trying to remove, and what that effort contributes?The News BeatWhat leaves with the people?We opened with the corporate layoff announcements and the familiar explanation that organizations are shifting toward AI. My concern was the experience those people leave with. Someone may know the process, but they also know where it breaks, which customer needs another conversation, and what to do when the instructions stop being useful.Colette pushed that concern further. If the expected AI gains underdeliver, an organization may already have removed the people who could help recover the work. The people who remain are watching, too. Uncertainty about their own positions can affect their willingness to take risks, invest in relationships, or commit to the next change.An efficiency gain can look different once you account for what happens afterward.Kelly introduced an alternative: IKEA’s approach to reskilling. In 2023, Ingka Group reported that 8,500 call center coworkers had gained skills in areas including remote interior design, digital sales, relationship building, and complex customer inquiries as its Billie chatbot handled simpler questions. That is an older example, but a useful one for this discussion. It shows that adopting AI can include a deliberate decision about where people contribute next.Kelly also noted that people need opportunities to learn how their work is changing. That puts responsibility on both sides. Employees need to develop skills, and organizations need to decide whether they are willing to invest in the people who already understand the business.What is happening to empathy in management?The second story covered research in HR Dive that found managers scored below the general population on measures including empathy. The research draws on personality assessments spanning almost twenty years; it does not establish that AI has recently made managers less empathetic.That distinction matters because Colette immediately questioned the explanation. Are organizations promoting people who demonstrate less empathy? Is the experience of management wearing it down? What expectations and pressures are shaping the behavior?Those are different problems. They need different responses.Our preparation also included Gallup’s reporting on declining engagement, with managers accounting for much of the recent downturn. Gallup connects the discussion to the scope of managers’ responsibilities and notes their importance in supporting AI adoption.The person we expect to guide everyone else through change may be struggling with it themselves. Asking that person to demonstrate more empathy is reasonable. We also need to look at what we’re asking them to carry.The Gravity AuditThe work around the systemHuman Middleware is the work people do to connect what an organization’s systems, processes, and ways of working leave disconnected.I originally understood that in a very tangible way. A person takes information from one place and puts it somewhere else. Someone reconciles two systems. Someone knows how to get a request through because the official process doesn’t quite do the job.I wanted to fix those gaps. I still do.But the conversation with Colette brought us into the relationships around that work: the trust that makes someone willing to raise a concern, the explanation that helps another team understand a request, and the attention that lets a manager notice someone is having a hard time.Kelly and I had a current example. As we prepared to launch a system, we were still discovering conversations and background work. The system could accept more information and included an AI component, but people were still using back channels to get things done.We hadn’t accounted for everything.That is an uncomfortable admission when you’re the person talking about Human Middleware. It is also exactly why I keep testing the idea against actual work. A new system can make the visible process easier while leaving people responsible for the connections around it.Before counting that as an efficiency gain, I want to know what those people are still doing.Some of the labor is the pointColette gave us the example that became the center of the conversation.A neighbor had brought her homemade chicken noodle soup when she was sick. Some efficiency was involved—the soup had been prepared and frozen—but its meaning included the effort and attention behind it. Bringing a can of soup to the doorstep would have been a different gesture.Kelly connected that to a child making a gift. Its value is bound up in the child having made it for you. The finished object is only part of what you receive.That distinction gets harder at work, because we’ve trained ourselves to look for effort we can eliminate.Some effort is unnecessary. Some exists because a process is broken. Some helps us understand another person well enough to make a sound decision or offer useful support.We need to tell those apart.An AI tool can help a manager organize their thoughts before a conversation. It can suggest clearer wording. But the manager still needs to understand what happened, consider the person involved, and take responsibility for what they say.A warm message can be useful. The relationship still needs someone in it.The Conversation with ColetteConnection does not look the same to everyoneWe spent part of the episode talking about remote work, messages, and meeting people in person.Kelly described how her impression of someone had changed after meeting them. Over chat, she had read a particular tone into their messages. In person, she found someone she connected with. I recognized the value of those encounters, too.Colette offered an important qualification. She prefers text to phone calls in many situations, and receiving a message can feel like connection to her. The medium itself does not tell us whether someone cares.That challenged an assumption running through parts of our discussion. We can mistake our preferred way of connecting for the correct way.If empathy involves understanding another person, we have to leave room for their preferences. Someone who wants a message rather than a call is telling us how to reach them.Does AI make room for people?Kelly asked Colette whether AI gives leaders more time to connect with their teams or allows them to avoid difficult conversations.Colette’s answer depended on how the person uses it.Earlier that day, she had used AI to compare code between environments. Saving time on that task could leave her with more attention for a later conversation. She also described another possibility: using AI to summarize a conversation and send an email without absorbing what was discussed.The tool can support either pattern. The person still chooses how involved they want to be.This is where my concern about saved time keeps returning. I’m quite capable of finding another task to put in it. Saving ten hours can become an invitation to accept ten hours of new commitments.We should be specific when we say AI will create more time for relationships. What work is going away? What will we do with the capacity? Will people actually be more available to one another?Empathy under pressure Get full access to Organizational Gravity at organizationalgravity.substack.com/subscribe

Who do we trust when the technology, companies, and rules are changing this fast?This week, Gabe and Kelly start with a deceptively simple question: who do you actually trust in AI?Is it easier to trust Google because it has been part of our lives for decades? Can newer companies like OpenAI and Anthropic build that same trust—or are they trying to manufacture it faster than trust normally develops?From there, the conversation gets bigger.We talk about the difference between trusting an AI company and trusting what the AI actually tells you; why hallucinations and sycophantic behavior change the relationship; and what happens when AI starts sounding, responding, and even speaking more like another person. If a particular voice feels more familiar, reassuring, or authoritative, does that change how willing we are to believe the answer?Then trust moves from AI into the workplace.Can you trust your manager while distrusting the company? What do leaders really mean when they ask employees to “trust the process” during change? And when we introduce AI into that process, are we automating work—or quietly moving trust from a person into a system?That becomes the Organizational Gravity question at the center of the episode:Does the organization work because it's trustworthy—or because trustworthy people keep making it work?That question matters even more as AI begins taking over pieces of Human Middleware: the people who connect systems, interpret ambiguity, know where the exceptions live, and keep work moving when the formal process is not enough.We also cover AI security, the 2026 Hugging Face incident, Anthropic’s new Public Benefit Fund, the enormous amount of money flowing through AI, and whether corporate visibility and public promises actually build trust—or just create presence.And finally, we turn the question back on leadership.Being trusted doesn’t automatically mean you’re right. Fixing something doesn’t automatically earn trust in the next change. And sometimes one of the clearest ways a leader can demonstrate trust is surprisingly simple:Let go.Not completely. Not irresponsibly. But enough to let someone else own the work, make decisions, and prove what they can do.Maybe trust isn’t something we can announce, market, automate, or demand.Maybe it has to accumulate.And maybe that’s exactly what makes this moment in AI so difficult.In this episode:* Why longevity and familiarity matter when we decide which technology companies to trust* Trust in an AI company vs. trust in an AI answer* Hallucinations, sycophancy, and the danger of confident language* Human-like AI, voice, anthropomorphism, and emotional trust* Why “trust the process” usually shows up during change* AI security and what the Hugging Face incident exposed* Whether public-benefit commitments build trust or simply build presence* Trust between employees, managers, leadership, and the larger organization* Human Middleware and the people quietly holding organizations together* Why being trusted is not proof that you’re right* Delegation, letting go, and trust as a leadership behaviorThe Gravity Audit:Where does trust actually live in your organization?If the person everyone relies on disappeared tomorrow, would the process still work?And if AI replaces part of what that person does, what exactly are you transferring to the machine—the task, the judgment, or the trust? Get full access to Organizational Gravity at organizationalgravity.substack.com/subscribe

Episode DescriptionWe’ve spent a lot of time discussing Human Middleware: the invisible work people do to keep organizations running. But what happens when that invisible work falls disproportionately on women? And what happens when the people doing it aren’t recognized for their contributions?For this episode of the Organizational Gravity Podcast, Gabe and Kelly join Audrey Johnson and Kay Turner for a personal conversation about women in the workplace, leadership, motherhood, compensation, professional expectations, and AI's growing influence.These women have worked together across multiple organizations and bring years of experience in project management, operations, and organizational change.Rather than simply discussing workplace statistics, they share their own experiences, including the moments that shaped their careers, the obstacles they’ve encountered, and the opportunities they see ahead.Kelly Sheahan — The Hidden Expectations of Being DependableInvisible work and unequal expectationsKelly opens the conversation by describing how women often take on extra responsibilities without being asked whether they belong to them. She recalls an executive requesting a presentation at 4:59 PM, due by 8 AM the next morning.Her question: Why was she the person expected to take care of it?Navigating male-dominated industriesKelly discusses the differences between working in female-dominated and male-dominated industries, particularly the expectations surrounding assertiveness, communication, appearance, and professional behavior.She describes the fine line women often navigate between speaking their minds and being perceived as aggressive or difficult.Motherhood and returning to workKelly shares the challenges of returning from maternity leave, balancing professional responsibilities with motherhood, and keeping pace with rapidly changing technology.She also questions why maternity and paternity leave remain so inconsistent.Pay equity and professional recognitionKelly recounts discovering that a less-experienced employee reporting directly to her earned about the same salary, despite her additional management responsibilities.She questions how organizations evaluate experience, determine compensation, and account for subjective judgments in promotion decisions.AI as an opportunity to do more valuable workKelly sees AI as an opportunity to move beyond repetitive tasks and toward higher-value contributions. She also challenges the assumption that lower recognition of women’s AI use necessarily means women use it less.Audrey Johnson — Making Invisible Work VisibleRecognizing the value of work that happens behind the scenesAudrey challenges the idea that invisible work is somehow less valuable.Much of project management involves connecting people, solving problems, coordinating activities, and keeping work moving.Her central question is how women can make those contributions visible and demonstrate their value.Executive presence versus authenticityAudrey discusses feedback on her executive presence and reconsiders what the phrase actually means.Was she being encouraged to become a stronger leader, or being asked to behave in a way that matched someone else’s expectations?She argues that authenticity matters because maintaining a professional persona that isn’t genuinely yours is ultimately unsustainable.Visibility, promotion, and communicating business valueAudrey reflects on the importance of discussing accomplishments, communicating financial impact, and ensuring leadership understands the value of work being performed.She also acknowledges that promotion decisions often involve visibility and relationships, not simply performance.Motherhood and career decisionsAudrey shares how having a daughter influenced her willingness to travel as a project manager.She describes feeling less visible on her own projects when colleagues traveled more frequently and wondered whether those differences would affect her advancement.AI, learning, and career opportunityAudrey encourages women to become curious about AI, invest time in learning it, and use it to reduce repetitive work.She sees AI proficiency as a potential path toward better roles, greater responsibility, and improved compensation.Finding your voice in the roomAudrey discusses learning when to interrupt, redirect a conversation, and insist on finishing a thought.She argues that women should feel comfortable exercising authority in meetings rather than automatically yielding the floor.Kay Turner — Recognition, Opportunity, and the Cost of Speaking UpWhen other people receive credit for your workKay argues that much of women’s work isn’t invisible. Rather, organizations have become accustomed to expecting women to perform it.She describes how women can contribute substantially while someone else receives the recognition, and how that experience can undermine confidence.Assertiveness and unequal standardsKay reflects on being a direct, outspoken woman in corporate environments and the professional consequences of refusing to back down.She questions whether the same traits that can advance a man’s career sometimes limit a woman’s opportunities.The career opportunities women sacrificeKay shares a pivotal moment when a female executive recognized her potential and offered her a director position that required relocating.She declined because of her responsibilities as a mother and still wonders how her career might have developed had she accepted.A serious accident and unequal compensationOne of the episode’s most difficult stories comes when Kay describes a car accident that left her unable to work for three months.Despite managing a project for 14 months before the accident and returning to complete it, she received only 15% of the project bonus, while the colleague who temporarily replaced her received 85%.Kay discusses the humiliation, frustration, and lasting effects of that decision.Race, bias, and advancementKay shares her experiences navigating corporate America as a Black woman and discusses how personal relationships, unconscious bias, and expectations about who fits a leadership role can affect career progression.She challenges the assumption that promotions are always determined primarily by skill and performance.AI access and opportunityFor Kay, the question isn’t simply whether women are willing to use AI.It’s whether they have the access, exposure, training, and opportunities to develop meaningful capabilities with it.Leadership is a collective effortKay closes with a reminder that productive disagreement is not an attempt to undermine leadership.Teams succeed when people challenge one another constructively, recognize their shared responsibilities, and work toward collective success.The Gravity Beat — What Are We Actually Recognizing?Throughout the discussion, one question continues to surface:If organizations depend on invisible work, why do they struggle to recognize the people performing it?The conversation explores how performance expectations, leadership behaviors, career interruptions, and organizational structures can create additional burdens that aren’t always reflected in job descriptions or compensation.AI creates an interesting possibility. Some of the repetitive and administrative work that has historically occupied employees may become easier to automate.But removing the task doesn’t automatically correct the organizational assumptions that made the work invisible in the first place.And without equitable access to technology and career opportunities, organizations risk carrying those same problems into a new generation of work.A Personal Thank YouGabe closes by reflecting on the women who shaped his life and career.From his grandmother, who taught him to read, to his aunt, whose career helped shape his understanding of business, to Star, whose continued support makes so much of his professional and creative work possible.He also thanks Kelly, Audrey, and Kay for years of collaboration… Get full access to Organizational Gravity at organizationalgravity.substack.com/subscribe

AI is supposed to make work easier. So why are we using the time it saves to take on even more?This week, Gabe and Kelly explore an uncomfortable possibility: what if the biggest obstacle to eliminating unnecessary work isn’t the technology, but our own behavior?We revisit Human Middleware, the invisible work people do to keep disconnected systems, processes, and organizations running. But the conversation takes an unexpected turn when Gabe admits something he got wrong in his original thinking: even when technology removes the need for people to stand in the middle, we often choose to put ourselves right back there.We discuss the difference between getting work done and creating value, the instant gratification of AI systems that agree with us, and the growing responsibility of managing teams of AI agents without necessarily having the experience to question their outputs.We also introduce three lenses for understanding Human Middleware: structural, behavioral, and social. Some of that work should disappear. Some may be worth protecting.And perhaps the biggest question: if AI accelerates everything we do, are we also accelerating the organizational and personal habits we should have left behind?In this episode* 01:00 — AI makes us faster, but are we doing less?* 07:12 — The danger of an AI partner that always agrees* 11:04 — Why judgment matters more than producing another PowerPoint* 16:20 — Everyone becomes a founder: managing your own AI agents* 24:14 — The three forms of Human Middleware* 28:10 — Why human relationships and experience still matter* 34:27 — What Gabe got wrong about Human Middleware* 39:26 — Why removing tasks doesn’t necessarily improve work* 40:20 — The tool changed. We didn’t.Join the conversationWhat happens when you eliminate unnecessary work? Do you actually reclaim the time, or do you find something else to fill it?We’d like to hear where you see Human Middleware in your organization, which parts should disappear, and which parts represent something genuinely valuable.Subscribe to Organizational Gravity for more conversations, essays, and practical perspectives on organizational design, technology, AI, and the way we work. Get full access to Organizational Gravity at organizationalgravity.substack.com/subscribe

Would Your Hiring Process Screen Out Your Best Employee?A résumé gives us the polished version: the roles, the accomplishments, the software, the results. It rarely tells us what went wrong, what remained unfinished, or how someone developed the judgment they bring to work.In this episode of Organizational Gravity, Gabe and Kelly explore what gets lost when hiring becomes an exchange between optimized résumés and automated screening systems. They discuss transferable skills, the responsibility of hiring managers to define a role clearly, and the uncomfortable experience of writing a failure CV.The conversation returns to human middleware—the people who keep work moving between systems, teams, and processes. Before handing their work to an agent, there are questions worth asking: Why was this process built this way? What experience shaped it? And what could this person contribute if they had more time?In this episode:* 00:15 — AI résumés, automated screening, and the first human contact* 04:52 — Who checks which candidates the system rejects?* 07:08 — Recognizing transferable experience* 08:46 — When the hiring manager defines the wrong job* 11:13 — The uncomfortable work of writing a failure CV* 18:19 — Looking beyond polished accomplishments* 20:13 — Vulnerability and sharing professional failures* 24:08 — Human middleware and the judgment behind a process* 27:49 — Interview credibility and first impressions* 35:21 — Degrees, gender, and screening bias* 38:36 — Recognizing skills gained in other industries* 40:06 — Automation, hiring uncertainty, and time to thinkA few clarifications belong alongside this conversation.Automated screening can exclude qualified candidates, particularly when employers use overly restrictive requirements. That concern is documented in Harvard Business School and Accenture’s Hidden Workers research. But screening systems vary, and a missing keyword does not automatically mean rejection. A lack of response also does not establish that AI rejected someone or that an AI-written résumé would have changed the outcome. Read the research.Our discussion of body language went further than the evidence supports. Eye direction is not a reliable way to determine whether someone is lying. Looking away, checking notes, or keeping a camera off does not establish dishonesty. A stronger approach is to ask consistent, job-related questions and evaluate the substance of the answers. We also used “imposter syndrome” when discussing possible misrepresentation; those are different things. Eye-direction study, structured-interview guidance.The Fortune 100 or Fortune 500 did not collectively eliminate college-degree requirements. Some employers removed them for particular roles, and research found that those announcements often produced limited changes in who was actually hired. HBS and Burning Glass report.Women earned the majority of U.S. bachelor’s and master’s degrees in the 2021–22 data cited here. That does not, by itself, tell us whether women are being hired at higher or lower rates. Research has documented gender and racial bias in particular hiring settings and AI screening experiments, but those findings should not be treated as a description of every employer or tool. NCES undergraduate data, graduate data, AI screening study.Our comments about women being better multitaskers or project managers reflect personal impressions, not an established general advantage. The multitasking study linked here found no significant gender differences in the tasks it tested. The broader hiring question remains useful: are we recognizing the skills a person has demonstrated, including those developed outside a conventional career path? Read the study.We also discuss whether uncertainty about AI is holding back hiring. That is a possibility raised in the conversation, not a demonstrated explanation. The August 2026 BLS release, published after the recording, reported hiring little changed over the month and year; it does not establish AI as the cause of hiring trends. BLS report.A few smaller points: the “8%” reference was not tied to an identifiable source and should not be treated as a verified statistic. Professional résumé writers still exist. The Martha Stewart story is Gabe’s recollection and interpretation, rather than a verified quotation. And a failure CV is presented here as an exercise in reflection, not a tested strategy for getting through an application system. Professional résumé-writing association.The “un-resume” discussion draws on Jasmine LeFlore’s TEDx talk, It’s Not Only the Resume, It’s also the Un-resume, which explores the experiences and unseen work behind conventional credentials. Watch the talk.What experience made you better at your work but never made it onto your résumé? Get full access to Organizational Gravity at organizationalgravity.substack.com/subscribe

Fresh off Dreamforce, Gabe and Kelly dig into the increasingly loud debate around slowing down AI. The interesting part isn’t simply whether AI should move faster or slower. It’s deciding what should be allowed to move quickly, where the boundaries belong, who can stop something, and what evidence allows it to move again.In other words: governance.The technical gap is collapsing. Something that once took ten months to build might now take ten days. So what should we do with the time we just got back?Our argument: don’t immediately spend all of it going faster.Use some of it to understand the process. Reduce what shouldn’t be there. Define what the system should never do. Test the assumptions. Then automate.We also get into one of the stranger consequences of automation: the people entering the workforce may not be cognitively offloading work they already understand. They may never cognitively onboard that knowledge in the first place.That problem isn’t new.In 1983, psychologist Lisanne Bainbridge described what she called the Ironies of Automation: as systems become more automated, the human operator can become even more important when something unusual happens.“The more advanced a control system is, so the more crucial may be the contribution of the human operator.”More than forty years later, that sounds a lot like the human-in-the-loop conversation we’re having about AI.In this episode00:00 — Governance at machine speedDreamforce, the “slow down AI” debate, guardrails, process, and why governance suddenly seems to be everywhere.02:25 — What shouldn’t we build?AI changes an old requirements question. It’s no longer enough to ask what a system should do. We also need explicit boundaries around what it must not do.05:00 — Cognitive onboardingWhat happens when automation removes the work people once used to develop judgment, experience, and understanding?05:50 — The Ironies of AutomationWhy a paper from 1983 may be one of the most useful things to read about AI automation today.07:30 — What exactly are we slowing down?Frontier research, business implementation, process design, and why “slow down AI” becomes much more useful once we get specific.09:40 — Spend some of the speedIf development drops from ten months to ten days, maybe five of those days should be spent understanding and reducing the process before automating it.14:50 — Speed versus safety is really a governance questionWhat crossed the boundary? Who has authority to stop it? What evidence allows it to continue?17:15 — AI cults, hype, and fearWhy the AI conversation can quickly drift from legitimate safety concerns into religion, apocalypse, politics, capital, and social-media amplification.24:10 — Is your organization actually ready for AI?Process understanding, data foundations, iteration, and why automating something you don’t understand just accelerates the problem.31:50 — The Gravity AuditOld permissions. Unclear ownership. Disconnected systems. Exceptions. Informal workarounds. AI didn’t create any of this, but it will inherit all of it.39:30 — Humans in the loopAI and humans can both accumulate organizational mass. Both need monitoring, reduction, and intentional design.47:15 — What AI is making visibleAI may be exposing Organizational Gravity that was always present but difficult to see.51:35 — Can governance become something people actually want to participate in?Making process work visible, useful, fast, and tangible instead of making governance feel like policing.53:40 — Where Gabe might be wrongMaybe people won’t willingly expose every workaround and task they perform, particularly when automation creates fear about what happens to their role afterward. That makes trust part of the governance problem.References mentionedLisanne Bainbridge — “Ironies of Automation” (1983)Automatica, Vol. 19, No. 6, pp. 775–779.DOI: 10.1016/0005-1098(83)90046-8Caitlin Malone — Linear DigressionsThe discussion that sent Gabe back to Bainbridge and the much older research behind today’s human-versus-automation questions.Mike Krieger — Anthropic LabsInstagram co-founder and current head of Anthropic Labs, whose Dreamforce discussion helped frame the conversation around how quickly AI-native work is changing.The Daily — “The Techno-Religion at the Center of the A.I. Debate”A look at the communities, ideologies, and “doomsday cult” criticism surrounding parts of the AI safety debate.The bigger ideaAI didn’t create Organizational Gravity.It inherited it.The danger isn’t simply that machines will move too fast. It’s that we’ll use their speed to automate old permissions, bad processes, undocumented exceptions, unclear ownership, and decades of accumulated organizational mass.Governance at machine speed doesn’t mean making governance instantaneous.It means using the speed of the machine to give humans more time to understand what should exist in the first place.And maybe that’s the real opportunity: not simply faster automation, but better-designed work.Corrections from the episodeJon KrohnIn the recording, the name came out as “John Crone.”Correct: Jon Krohn, host of the Super Data Science: ML & AI Podcast.Dr. Katie MaloneIn the recording, I referred to her as “Dr. Caitlin Malone.” She publishes and podcasts as Dr. Katie Malone. She was the guest on Jon Krohn’s September 22, 2026 episode.Linear DigressionsNot “Linear Degressions.”Katie Malone hosts the AI, machine learning, and data science podcast Linear Digressions.Lisanne BainbridgeIn the recording, I said “Larissa Bainbridge.”Correct: Lisanne Bainbridge.Her 1983 paper is:Bainbridge, Lisanne. “Ironies of Automation.” Automatica, Vol. 19, No. 6, November 1983, pp. 775–779.DOI: 10.1016/0005-1098(83)90046-8The paper’s core argument is even more relevant today: automating routine work can leave humans responsible for abnormal situations while simultaneously giving them less practice and less direct involvement in the work.Mike KriegerThe former Instagram leader I was trying to remember was Mike Krieger, co-founder and former CTO of Instagram. He joined Anthropic as Chief Product Officer in 2024 and moved into Anthropic Labs in 2026. He spoke at Dreamforce 2026.NVIDIA / Hugging FaceI said at one point that NVIDIA had “bought” Hugging Face. More precisely:NVIDIA entered into a definitive agreement to acquire Hugging Face in September 2026. The transaction has been announced but has not yet closed. NVIDIA said it expects the deal to close in the first half of 2027, subject to customary conditions and regulatory approvals.The New York Times / AI “cult” discussionThe episode I was referring to was The Daily episode:“The Techno-Religion at the Center of the A.I. Debate” — September 18, 2026One clarification: the episode is more nuanced than saying AI executives themselves have formed a “cult.” It looks at the ideologies around AI development and safety, along with the backlash that has characterized some AI-safety advocates as a “doomsday cult.”Sources & Further ReadingLisanne Bainbridge — “Ironies of Automation” (1983)The foundational paper behind our discussion of automation, declining human practice, and why human expertise can become more important rather than less important as automation improves.Super Data Science Podcast — Episode 1029“How AI Brought a Podcast Back From the Dead, with Linear Digressions’ Katie Malone”Jon Krohn with Dr. Katie Malone, September 22, 2026. The discussion of Bainbridge’s Ironies of Automation begins around 41:39.Linear Digressions — Katie MaloneKatie Malone’s podcast covering AI, machine learning, data science, agents, benchmarks, and the human consequences of automation.Mike Krieger — Anthropic LabsKrieger co-founded Instagram, joined Anthropic as Chief Product Officer, and moved into Anthropic Labs in 2026 to work on experimental products at the frontier of Claude’s capabilities.The Daily — The New York Times“The Techno-Religion at the Center of the A.I. Debate”September 18, 2026. A discussion of AI safety, technological… Get full access to Organizational Gravity at organizationalgravity.substack.com/subscribe

AI is supposed to save time. So why does so much of that time seem to disappear?In this episode, Gabe and Kelly dig into Glean’s 2026 Work AI Index and two terms that capture a growing problem at work: bot sitting—the human effort required to feed, monitor, correct, and clean up after AI—and botshitting—sending AI-generated work downstream without really understanding or validating it.That leads to a bigger Organizational Gravity question: what if AI is making work faster without actually making the system lighter?We talk about where those supposed productivity gains go, why human review is different from babysitting a poorly designed bot, and how unverified AI output can move risk downstream faster than the old process ever could.The conversation also puts pressure on Organizational Gravity itself. I have often treated friction as evidence of unnecessary organizational weight, but AI exposes an important exception: some friction is doing real work. Verification, judgment, accountability, experience, and safety can all slow a process down for good reasons.The goal cannot be zero friction. It has to be the right friction, scaled to the consequence of getting something wrong.In this episode00:00 — Bot sitting and botshittingWhat the terms mean and why individual AI productivity does not always translate into better organizational performance.04:25 — Who botshits more?A funny statistic about men and women turns into a more serious conversation about confidence, scrutiny, and how people review their work.07:35 — Human in the loop or babysitting the bot?Where useful supervision ends and poor system design begins.09:05 — Why are people shipping work they cannot defend?Deadline pressure, trust in the technology, weak review practices, and the widening technology gap.17:11 — The Gravity AuditBot sitting starts to look a lot like Human Middleware. Are we adding AI without removing the process that created the work in the first place?25:00 — The 4:30 PM testThe deadline is real, you are tired, and AI gave you something usable. Do you ship it or tell people you will be late?27:34 — Where is Gabe wrong?We put Organizational Gravity itself under pressure.28:29 — Not all friction is badRemoving every manual check can also remove the judgment that keeps bad information from moving downstream.33:41 — Zero friction is the wrong targetGovernance should create the right amount of friction based on the consequences of the decision.37:00 — A revised rule for frictionRemove friction caused by poor design. Preserve friction that creates judgment, accountability, or safety—and periodically prove that it still does.40:34 — Three questions before you automateWhat should actually disappear? What judgment still belongs to a person? How will we know the whole process got better?The takeawayAI can make work move faster while Organizational Gravity continues accumulating underneath it.If humans are constantly feeding the bot, checking whether the bot is functioning, repairing bad output, and carrying knowledge that never made it into the system, we have not necessarily removed the weight.We may have just compressed it and accelerated it.The better question may not be:How much work can we automate?It may be:What should become easier, and what do humans still need to struggle with long enough to understand?And that leaves us with a question for the next episode: maybe Human Middleware should not disappear entirely.Maybe the real question is:What should humans still be in the middle of?Source: Glean Work AI Institute, The Work AI Index 2026: Botsitting, botshitting, and the hidden human labor of AI at work (2026) Get full access to Organizational Gravity at organizationalgravity.substack.com/subscribe

Show NotesThis week, Gabe and Kelly take on one of the least glamorous but most important parts of organizational change: training.The conversation starts with a simple problem. Organizations are changing the technology around work much faster than they are changing how people learn. Training is still too often treated as something completed once during onboarding or rolled out through a video, an SOP, or a compliance module, with the assumption that people will remember what they need months later.That leads to a bigger distinction between training, learning, and performance. When is someone struggling because they were never given the information at the moment they needed it? When does repetition become learning? And when does a recurring problem stop being a training issue and become a performance conversation?The episode also explores point-of-need learning. Instead of expecting people to remember everything in advance, what if technology could provide guidance in the moment the work is happening? The idea is not just to make training easier, but to move learning closer to practice, where repetition, feedback, and context can turn information into capability.That creates another question: what happens when technology gets too good at remembering for us?Gabe and Kelly discuss cognitive offloading, judgment, and the risk of providing so much support that people stop gaining the experience they need to handle exceptions. AI may be able to guide someone through the standard process, but organizations still need to decide where human judgment belongs and when a leader should step in.The conversation then moves into Reductive Change. Gabe references Leidy Klotz’s work on subtraction and the tendency people have to add rather than remove. The same instinct shows up in training and process design: a problem appears, so organizations add another step, another document, another control, or another training requirement. AI may give us a chance to do something different by first removing unnecessary friction, then using technology where it genuinely helps.Klarna also comes up as an example of the tension between automation and human knowledge. In 2024, Klarna said its AI assistant was doing work equivalent to roughly 700 full-time customer-service agents. Later, CEO Sebastian Siemiatkowski said the company had pushed too far toward AI-driven cost-cutting and had begun rebuilding access to human customer support. The lesson is less that AI failed and more that automation can expose how much context, judgment, and relationship knowledge people were carrying.The episode closes on a different way of thinking about reskilling. Maybe the goal is not simply teaching people how to work alongside AI. Maybe it is helping people let go of repetitive work technology should carry, then using the recovered time to build judgment, relationships, communication, and the parts of work that remain distinctly human.Topics discussedTraining vs. learningCompliance vs. adoptionPoint-of-need trainingPerformance vs. training issuesCognitive offloadingRepetition and consistencyLeadership reinforcementAI and exception handlingReductive ChangeKlarna and human customer supportThe value of recovered timeWhat reskilling should look like in an AI-enabled workplaceReferencesLeidy Klotz, Subtract: The Untapped Science of LessAdams et al., “People systematically overlook subtractive changes,” Nature, 2021Klarna, AI customer-service assistant announcement, 2024Research and practice around the 70-20-10 workplace learning model and experiential learning Get full access to Organizational Gravity at organizationalgravity.substack.com/subscribe

Episode blurbLinkedIn shipped a button that lets anyone report your post as AI slop, the same day it removed its own AI writing tool. On a network where reach is proximity to a job, that isn’t moderation. It’s a permission structure, and the people most exposed are the ones who already write like everyone else because they were trained to. Also this week: Pew put a number on how much of the web is written by AI, and a federal research office said that same kind of detector can’t stand alone as evidence of anything. The institution with more at stake demanded more evidence.Longer descriptionOn 30 July, LinkedIn added “Seems like AI slop” to the menu on every post and removed Enhance Post, its own AI writing tool, the same day. The order matters. The platform built the tool, the tool produced the output, the platform took the tool away and handed everyone a button to report each other for it. Nobody planned that, which is exactly what makes it gravity rather than malice.What it does to a person is the part worth sitting with. LinkedIn isn’t where people go to tell stories; it’s where they go to be employed, so a reach penalty isn’t a content-moderation outcome. It’s an economic one, applied silently, by a stranger, at no cost to them and with no appeal for you. And the people most exposed are the ones taught to write in corporate templates, the non-native speakers whose careful prose reads smooth, the juniors imitating a senior to be taken seriously.That sits inside a bigger story. Pew Research published a census-scale estimate of how much of the web is AI-written, and the number came out of a detector. Pew did the work carefully and documented it all — the threshold, the agreement statistic, the false-positive warning, the dating limitation. What happens next isn’t Pew’s fault. It’s what always happens: the caveats fall away, and the number travels alone. Meanwhile, the HHS Office of Research Integrity looked at the same class of instrument and said it can’t be the sole basis for a misconduct finding.And in Where I’m Wrong: I thought training was the easy part of governance. Three studies took that apart, and the one that moved me most wasn’t the one saying training fails.In this episodePew’s census-scale estimate of AI-written pages, and the caveat that decides what the number meansWhy only 10 to 15 percent of sampled pages could even be dated, and what that does to the headlineThe HHS Office of Research Integrity guidance, and why a federal office demanded more evidence than a publishing platform didWhat it means to admit a probabilistic estimator into a fact-producing role without treating it as a changeThe structural middleware underneath the whole story: a 490,000-page apparatus built to approximate a fact every generator knew with certainty and discardedLinkedIn’s slop button, the removal of Enhance Post, and the order those two things happened inWhy a reach penalty on a professional network is an economic outcome, not a moderation oneWho’s actually most exposed: the people trained to write like everyone elseWhere I’m Wrong: training, and the difference between a content problem and a conditions problemKey ideasOrganizational Gravity — organizations accumulate mass faster than they shed it. Here it shows up as a measurement admitted into the operating model with none of the questions a change would have to answer.Governance is not measurement — the order is issue, root cause, control, then measurement. This story runs it backwards: the detector arrives first and becomes the fact.Structural middleware — when no provenance layer connects generation to publication, somebody has to build a 490,000-page estimator to recover what was known for free at the moment of writing. That’s incomplete architecture, and the fix is integration.The permission structure — a report button between peers on a network where reach is proximity to a job isn’t moderation, it’s a mechanism.SourcesSegment 1 — the news hitPew Research Center, “How much of the internet is written with AI?” (20 August 2026) — the primary study. 490,000 English-language pages sampled across 49 Common Crawl snapshots, January 2021 to July 2026. Supports the 10 percent raw July 2026 figure, the “over one third” figure for pages with a detectable post-ChatGPT publication date, and the methodology caveats including the dating limitation. Also carries Pew’s own false-positive warning and stated threshold — worth reading before repeating either number.HHS Office of Research Integrity, guidance on generative AI in federally funded research (14 August 2026) — the second institution. Supports the central claim that detector output may constitute evidence but that ORI recommends against relying on such tools as the sole basis for a misconduct finding. Also covers disclosure expectations and the penalty framing.Segment 2 — the Gravity AuditTech Times, “LinkedIn pulls its own AI writing feature, crowdsources slop detection” (31 July 2026) — supports both halves of the sequence: the addition of “Seems like AI slop” to the post menu and the removal of Enhance Post on the same day. Also carries the reporting on how flagging feeds LinkedIn’s classifiers and affects recommended-feed distribution.The 40-plus percent figure for fully AI-generated LinkedIn long-form content is Pangram’s research, not LinkedIn’s — attribute it that way on air and in writing.Segment 3 — Where I’m Wrong: trainingArthur, Bennett, Edens & Bell (2003), “Effectiveness of Training in Organizations: A Meta-Analysis of Design and Evaluation Features,” Journal of Applied Psychology 88(2), 234–245 — free full-text PDF. 397 findings across roughly forty years and about 33,000 people. Supports the claim that the average trained person outscored about three of four untrained people. Note the measures are largely taken at the end of training, which is the limitation that sets up the next study.Park, “Does Compliance Training Decrease Corporate Misconduct? Evidence From Field Data” (Harvard DASH) — five years of internal records from one multinational manufacturer, roughly 100,000 employees across about 100 countries, alongside 2,237 misconduct allegations. Supports the no-meaningful-association finding on average, the roughly two-month decay of in-person effects, the null result for video, and small classes outperforming large ones by around ten to one. Say on air that this is a dissertation, not peer-reviewed, one company, and compliance training specifically. Also available as Harvard Law Olin Center fellows paper 89.Blume, Ford, Baldwin & Huang (2010), “Transfer of Training: A Meta-Analytic Review,” Journal of Management 36(4), 1065–1105 — free full-text PDF from an author’s site; journal version at SAGE. Eighty-nine studies, roughly 12,500 people. Supports the finding that work-environment support — manager backing, a job that rewards the new behavior — predicted transfer better than trainee motivation, which ranked near the bottom. This is the study behind “it’s the room they go back to.”Read the full versionThis week’s newsletter went out on training — all three studies, the long version, and the part I didn’t get to on air: LinkedIn newsletterSubscribeI’m writing a book out loud, thinking it through in public, one essay and one episode at a time. Subscribe, and you’ll watch it take shape: Gabe | Organizational GravityNoteStories on this show are anonymized. Names and identifying details are always changed. Get full access to Organizational Gravity at organizationalgravity.substack.com/subscribe