Psychological Safety

00:00:01: When was the last time you spoke up against something at work that you considered wrong?

00:00:19: Welcome to the first episode of season two.

00:00:22: Speaking up clearly requires courage, especially when it goes against decisions made at the top authority.

00:00:29: bias sits in all of us and It's not easy to overcome an environments that have been built on hierarchy.

00:00:36: Following authority and capable leaders is a natural human instinct But what about following less capable and even bad ones?

00:00:45: How often do you hear people complaining about bad bosses, And how often to hear great ones these days.

00:00:52: When we look at the representation of wrong kind leaders in articles, posts or books It looks like it has become a thriving genre.

00:01:01: Ruthless and reckless leaders aren't just in news Whistleblower Books Like Sarah & Williams.

00:01:07: Careless People Have Become Best Sellers.

00:01:10: In her case despite the arbitration ruling that keeps her from promoting it, needless to say that matters fight against her became sufficient marketing by itself.

00:01:20: I read a book shortly after publication.

00:01:23: It confirmed what i had already perceived over years with some very disturbing additional details.

00:01:29: There's been enough other well-documented reporting on careless actions of key decision makers at former Facebook.

00:01:35: Only this time it left me with a clear urge to withdraw myself entirely from the world of Metta.

00:01:41: Easier said than done, I tried and won't go into details on how Metta keeps you from deleting your accounts.

00:01:49: The problem is a genuine dependence on their ecosystem.

00:01:52: i try To get people to connect With me through other channels And as i couldn't delete my facebook account?

00:01:58: I just deleted the app everywhere and stopped using It.

00:02:02: when i talked to People about why i'd Done it most agreed that the application of responsibility for consequences linked to Meta's products paired with their long-held mantra, move fast and break things represents utter carelessness at the top.

00:02:18: Still it didn't move any of them to stop using their product.

00:02:22: The next book I read was Empire Of AI an award winning book by Karen Howe about the rise of OpenAI –the company behind chat GPT.

00:02:31: She developed the portrait over several years as an invited journalist until the doors closed for her.

00:02:37: By then, she'd already received enough insights to cover development thoroughly.

00:02:42: The controversy around the company's leading figure Sam Altman has grown ever since the so-called blip in November of twenty-twenty three but brought back a few days later after the majority of employees threatened to leave the company.

00:03:02: The reason for the firing was the board's lack of trust in him.

00:03:06: On Altman's return, the Board got fired and the journey continued from a former nonprofit built-to benefit all of humanity To a for profit Company now targeting an IPO at one trillion dollars.

00:03:19: In April this year, the New Yorker published another sweeping documentation of credibility issues on open AIS leadership.

00:03:26: The article was titled Sam Altman may control our future?

00:03:30: Can he be

00:03:30: trusted?".

00:03:32: You see these aren't just random allegations.

00:03:34: they're ongoing investigations court filings and regulatory complaints.

00:03:39: internal memos obtained by investigators and journalists show how a lot doesn't read like they actually know and control what their releasing into the world.

00:03:51: And these are two of the major players defining the AI arrays, apparently led by people that own boards in former executives don't trust.

00:04:00: with this responsibility reported and documented through various outlets The problem I see was that This entire media coverage produces selective attention And it attributes power to individual people in a way that ignores the responsibility of the people following their lead.

00:04:18: The heroic, as well as bad leader narrative puts CEOs and other powerful figures on a pedestal.

00:04:25: It attributes accountability and power primarily to them... ...and overlooks environment support that got them & organizations where they are.

00:04:35: Researchers call this tendency the Romans' leadership bias.

00:04:39: Our liking for heroes and villains produces a lot of focus on very few people, You may remember me saying that the top shapes the culture of a company and every person joining has to work within constraints on what the organization allows in rewards.

00:05:18: Those constraints are reality for many workplace, but it doesn't mean they're only forces shaping an environment.

00:05:27: When the top leaders are placed on throne hidden in ivory tower shielded from world's reality we create false narrative of dependency.

00:05:36: Every organization can be sustained, influenced and reshaped by the people inside it.

00:05:42: We aren't just bystanders of what's happening.

00:05:45: every person holds some form of agency leaders in particular.

00:05:50: Remember Nokia?

00:05:51: The leading mobile phone company that vanished in less than ten years.

00:05:56: I was a loyal Nokia customer until Apple released the first iPhone in two thousand seven.

00:06:02: The primary narrative around their decline is failed vision, strategy and a problematic operating system.

00:06:09: Most of it attributed to the top management.

00:06:11: And while those aspects clearly mattered... ...the more interesting discovery was presented by researchers from INSEAD In two thousand fifteen.

00:06:20: they published a study based on seventy six interviews with Nokia's Top & Middle Managers Engineers and External Experts.

00:06:29: They wanted to understand why a company with that kind of market dominance had failed to respond to the iPhone.

00:06:36: What they found was problematic company culture, A Culture Of Fear Driven By The Top But Kept Alive by People Underneath.

00:06:44: The top managers did have terrifying reputation.

00:06:48: Threats Of Fireings And Emotions Were Common.

00:06:51: Out Of Fear They Failed To Face Reality Themselves.

00:06:54: Despite seeing what was happening, they failed to acknowledge that they could not compete with what Apple was building.

00:07:01: The N- ninety seven Nokia's flagship response to the iPhone Was a total fiasco in terms of product quality but They kept pressuring middle managers To drive small improvements to their existing products without revealing how serious things actually were?

00:07:17: The manager is facing That pressure and operating In a climate Of fear did What they were told And they went silent.

00:07:24: One of them described to the researchers how information never flowed upwards.

00:07:29: The top management was directly lied-to, people said they were being told by their supervisor to manipulate data and presentations for better impression.

00:07:38: everybody knew things were going wrong but the thinking was why tell Top Management about this?

00:07:44: It won't make things any better!

00:07:47: Another middle manager suggested to a colleague But his colleague answered that he didn't have the courage, He had a family and small children.

00:07:57: This is what a culture of fear driven by the top produced in layers beneath – calibrated silence as survival behavior.

00:08:05: The engineers could see that Nokia's technology was not going to compete.

00:08:10: They just did'nt feel safe say so.

00:08:13: So Nokia died as phone manufacturer because climate of fear didnt allow people tell their top leadership what they would have needed to hear, to change direction.

00:08:23: The researchers themselves named the practice that could've prevented this.

00:08:28: The company's leadership should've encouraged and role-modeled more authentic and psychologically safe dialogue internal coordination And feedback mechanisms To understand the true state the Company was in.

00:08:42: Later one of the researchers also wrote That when he taught the Nokia case to hundreds Of senior executives Many of them privately told him that the same harmful fear climate also existed in their companies.

00:08:56: And most importantly, they said it was impacting performance.

00:09:01: Now let me give you a different example and Let's talk about Google Project.

00:09:05: Aristotle was initiated In two thousand twelve to answer The question what makes A team effective at google?

00:09:12: Expectations were That the Answer would be all About talent.

00:09:16: The project ran for two years, looked at one hundred eighty teams took hundreds of interviews and tested over two-hundred fifty variables.

00:09:24: What Google found was not what they expected.

00:09:27: Talent mattered far less than they had assumed.

00:09:30: a team with less individual talent but the right way of working together outperformed a team of high performers that didn't have it.

00:09:38: They identified five factors that consistently distinguished effective teams from less effective ones.

00:09:44: dependability Members of effective teams reliably complete quality work on time.

00:09:50: Structure and clarity, team members understood what was expected and how to deliver it meaning the work mattered to people doing it.

00:09:59: Impact – Team members could see their work contributed.

00:10:03: And most important dynamic underneath these four ranked number one for effectiveness psychological safety.

00:10:11: Interestingly, this was something Google couldn't name until one of the analysts on The Project came across a paper published in nineteen ninety-nine by Harvard professor Amy Edmondson.

00:10:23: Thirteen years earlier she had called it psychological safety and she defined as... and that the team is safe for interpersonal risk-taking.

00:10:42: So, The Strongest Predictor For Team Effectiveness at Google was exactly this climate.

00:10:47: where these conditions existed teams outperformed comparable ones significantly Where they didn't.

00:10:54: talent alone couldn't compensate.

00:10:56: so Google's research ought to have settled a question about the value of psychological safety.

00:11:05: It came from inside one of the most successful, innovative and talent-obsessed organizations on this planet.

00:11:12: And it told us that The Way Team's work met us more than who is on them.

00:11:16: Amy Edmondson published her foundational work over twenty five years ago.

00:11:21: She spent decades researching it, writing and talking about it... ...and making the case to organizations That This Is Fundamental for Performance.

00:11:29: After Google's research was published a concept became more popular.

00:11:33: It made its way onto culture statements, into leadership decks and HR town halls across the business world.

00:11:40: But a phrase on his slide isn't the same as an established practice.

00:11:45: So The Value Proposition has been there for over twenty-five years.

00:11:49: How we've handled it is the problem.

00:11:51: In many cases it became a topic of the HR agenda.

00:11:55: You know one these soft people topics.

00:11:58: But psychological safety isn't soft and it's not just the nice to have.

00:12:02: It is a key performance indicator, especially now because most consequential workplace decisions that most organizations are currently making is AI adoption.

00:12:13: And when they're making in climates of fear Adoption won't be the only problem.

00:12:17: Edmondson herself has been looking at this specifically for past years.

00:12:22: She named two dynamics that most leaders aren't yet recognizing.

00:12:26: Despite the productivity gains AI is supposed to deliver, overall team performance is declining in many organizations.

00:12:34: People are starting to second-guess themselves and trust is eroding in ways most companies don't realize.

00:12:41: The first dynamic she called Trust Ambiguity.

00:12:45: It's what happens when AI produces confident output.

00:12:48: that turns out to be wrong.

00:12:50: Team members lose confidence not just in technology but also their own judgment.

00:12:55: Research now shows that sustained AI use undermines people's confidence in their ability to challenge AI output, even when they have the knowledge to do so.

00:13:06: The expertise that took years to build... ...now gets quietly overridden by a tool that sounds too certain!

00:13:12: The second dynamic at Minson called Workslop – AI-generated output….

00:13:17: …that doesn't move work forward.

00:13:19: It is an output which creates works for colleagues who either need fix it Or worse, redo the work.

00:13:26: The productivity gains show up in dashboards but the cost of cleaning out Workslop sits with people doing it in silence They may not feel safe to address because It easily sounds like blaming new shiny tools.

00:13:39: This is what's happening inside Teams right now and its only one layer Because at same time People are watching others lose their jobs.

00:13:49: In twenty-twenty five companies in the US announced more than fifty thousand layoffs and named AI directly as a reason.

00:13:56: Last September, The CEO of Salesforce went on a podcast and said that company had cut nearly half its customer support staff.

00:14:05: His exact phrasing was I need less heads.

00:14:08: Heads headcount.

00:14:10: something about this.

00:14:11: wording has always been very telling.

00:14:13: He also said AI now handles roughly half of all Salesforce's customer conversations.

00:14:19: What is worth noting, just two months earlier he had publicly said that AI would augment people and not replace them.

00:14:27: Two months!

00:14:28: That how quickly the position changes when cost savings are most visible metric to top Amazon Microsoft Meta All cutting thousands jobs and naming AI as reason which brings me to the part that should concern any leader currently making AI adoption decisions.

00:14:46: The same year, that AI usage at work jumped by thirteen percent.

00:14:50: confidence in using those tools dropped by eighteen percent.

00:14:54: so people say they are using it but don't believe their doing.

00:14:58: well maybe you remember my episode.

00:15:01: the end of why curiosity is what drives learning and adaptability.

00:15:06: fear shuts both down And we can see what such fear produces already.

00:15:12: These are early warning signs that AI adoption needs to be handled with care.

00:15:17: Another research found this, eight in ten employees carry a strong worry about at least one thing.

00:15:24: Sixty-five percent worry about being replaced by someone who uses AI better.

00:15:29: Sixteyone percent worry using it makes others think they no longer bring anything unique.

00:15:36: Sixty percent worry they'll be judged as less competent for using it at all.

00:15:41: This is fear about the implications of human work at scale, happening in environments that are already highly ambiguous.

00:15:50: Now you might expect people with such fears to use AI less Turns out.

00:15:55: opposite is true.

00:15:56: Employees with highest anxiety reported using AI for sixty-five per cent their job.

00:16:02: The less anxious ones only use it for forty-two percent.

00:16:07: So fear drives usage up and buy in down at the same time, which means dashboards on AI adoption are probably lying.

00:16:15: High usage numbers look like adoption is working but underneath a lot of that usage is fearful people protecting themselves using the tool because they're scared not to while quietly withholding judgment questions.

00:16:32: But the behavior isn't just silent.

00:16:35: A twenty-twenty six enterprise AI survey found something even more uncomfortable.

00:16:41: Twenty nine percent of employees admit to actively sabotaging their company's AI strategy.

00:16:47: among Gen Z workers, The number rises to forty four percent.

00:16:52: so fear doesn't Just stay silent.

00:16:54: it eventually shows up when people push back.

00:16:59: And it tells us the problem isn't AI adoption itself.

00:17:02: It's how leadership is handling it.

00:17:05: We've watched this with every technology wave before.

00:17:08: Every major shift in our work gets done has produced the same pattern Fear of being replaced Hidden use of new tools Workarounds and resistance.

00:17:18: What is different now, Is speed.

00:17:21: Previous transitions happened over decades.

00:17:24: There was time to study effects to evaluate what the technology meant for people using it.

00:17:29: And most importantly, to upskill people on how to use effectively.

00:17:34: But the AI race is collapsing their timeline into months.

00:17:38: The window-to-handle transition well... ...is so short that even best leadership gets challenged by it.

00:17:45: Research of what makes teams effective has been clear over twenty five years.

00:17:50: Yet many companies are trying to navigate the biggest disruption of human work with management practices from the past, ignoring how little they helped to produce better results.

00:18:01: So the way we handle AI adoption is no longer a management but leadership decision.

00:18:07: The company succeeding in this will show what difference between management and leadership And I believe AI will eventually turn great human leadership into one of the biggest assets that many companies have yet to recognize.

00:18:22: But before this happens, we'll probably hear more examples like the one from Klarna.

00:18:27: The Swedish fintech company became one of earliest and loudest voices for all-AI workplace.

00:18:35: In twenty-twenty three.

00:18:37: their CEO said publicly That AI could already do every human job.

00:18:42: The company froze hiring, replaced seven hundred customer service workers with AI chatbots and cut the overall workforce by around forty percent.

00:18:51: This savings were celebrated!

00:18:54: The strategy was held up as proof that aggressive AI adoption could replace human work at scale.

00:19:00: Then in May twenty-twenty five they reversed course And the CEO admitted That the Company had focused too much on efficiency & cost.

00:19:10: The result in his own words was lower quality and not sustainable.

00:19:14: Customer satisfaction had dropped, complaints had risen so they began rehiring human customer service workers.

00:19:22: the question that clearly wasn't answered before They made the decision where did the value of their customer service really come from?

00:19:30: Performance in metric obsession has left many managers blind to whether value really sits.

00:19:36: It has turned management into a mechanistic, number-obsessed practice far from what we should call leadership.

00:19:43: But even Klarna's walk back didn't slow the broader pattern at all!

00:19:48: So the fear across the workforce isn't paranoia but irrational response – it is an answer to movement that continues regardless of already proven to be backfire In particular, the decision-making on AI technology being mostly in hands of people furthest away from actual work.

00:20:07: And if those organizations run on authority and fear –the way Nokia did– no one will share information that matters.

00:20:15: Information about where AI adoption really extends capabilities... ...and where it undermines them… …and produces anxiety instead!

00:20:25: So companies that fail to build psychological safety in the past won't just struggle with AI adoption.

00:20:31: They will fail harder and faster, keeping human workforce capabilities they actually still need because how companies treat AI adoption not only exposes leadership weaknesses it amplifies them but there are also companies approaching this differently.

00:20:50: In the same Harvard Business Review article that named Trust, Ambiguity and Workslop, Adminson's co-author described how her team at three M is working with AI.

00:21:01: When they rolled out a new AI tool... ...they didn't mandate adoption.

00:21:05: They invited volunteers.

00:21:07: The signal was clear This is learning And not full deployment mode.

00:21:12: Volunteers in the wider organization knew they were helping this team to understand what works.

00:21:18: They highlighted improvements based on user feedback.

00:21:22: Instead of leading with time saved or tasks automated, they relied on testimonials to build confidence.

00:21:28: among others This team also openly posted about new learnings unexpected insights and AI limitations.

00:21:37: Identified problems were shared widely along what changes had been made And rewarded the right behaviors celebrating people who caught AI errors instead of blindly accepting outputs.

00:21:50: So questioning AI became a sign of good judgment and not one of resistance to technology, And the move that I think mattered most — they modeled fallibility.

00:22:01: Leaders openly admitted when they didn't understand AI outputs... ...and when output seemed questionable They shared their own mistakes about using it….

00:22:11: …and what they learned.

00:22:12: That is the practice of confident humility I talked about in my episode, The Confidence Trap.

00:22:18: Being open about mistakes asking questions staying open to not knowing without any fear.

00:22:25: this Is the environment we all need to grow and develop a leader who models.

00:22:29: This is paving the way for others To do the same.

00:22:32: this builds psychological safety.

00:22:35: an SAI handles more Of the routine work.

00:22:38: the human Work that remains becomes even More complex more uncertain and more dependent on people thinking together.

00:22:46: This is exactly the kind of work psychological safety is meant to protect, which means leadership now holds even more responsibility to create worksplaces where psychological safety isn't a claim presented in a slide deck but daily leadership practice –a practice incorporated by the entire leadership of a company–which brings us your choice!

00:23:10: the choice to build a safe environment where people can speak up, share their concerns address what is working and what isn't.

00:23:18: A workplace where they bring in ideas on how leverage the use of AI to extend capabilities without fear being replaced next month And when leaders clearly see that value and effectiveness for human work aren't just numbers game.

00:23:34: So if you are making AI adoption decisions In your organization right now Watch how your team responds.

00:23:40: Are you seeing the patterns this episode has named?

00:23:43: People hiding, How and when they use AI Others using it without questioning output Wrong Output leading to doubts about human capabilities Overly ambitious productivity goals from top Driving Workslop in layers below.

00:23:58: The question is not whether climate around us perfect.

00:24:02: It rarely Is!

00:24:05: What I'm talking about is the consistent practice of building and maintaining an environment where people closest to work can safely say what they see.

00:24:16: That's under every leader's control, And it's a leadership responsibility.

00:24:20: Google has shown that psychological safety Is foundation for team effectiveness.

00:24:26: It was long celebrated For its employee-centric culture.

00:24:30: The founders view on sustainable Long term performance.

00:24:34: But whether this culture has survived changing leadership, shifting market forces and new goals is a question worth asking.

00:24:42: And one I will come back to in the reflection episode on Thursday because the changes that happened at Google tell us something important about how easily the climate we know we need can be eroded under pressures an uncertainty every organization faces now.

00:25:00: Just think about how much we've been bombarded with predictions about people losing their jobs to AI in the last twelve months.

00:25:08: Recently, Sam Altman walked back his own predictions speaking at a conference in Sydney.

00:25:14: he said I'm delighted to be wrong about this.

00:25:18: He admitted that open AI had been pretty wrong on the social and economic implications of AI.

00:25:24: And he's not the only one.

00:25:25: Dario Amadei has also been walking back his earlier predictions.

00:25:31: His warning last year was that AI could eliminate up to half of entry-level white collar jobs and push unemployment into double digits, now he frames AI as a productivity multiplier rather than job destroyer.

00:25:46: So were the early predictions just overconfident claims?

00:25:55: What we do know is that both companies are heading into enormous IPOs.

00:26:01: These offerings need investors and share buyers to believe in a future of prosperity, the job apocalypse doesn't quite fit that picture.

00:26:10: so if thats the environment were dealing with what kind leadership you think can hold up against it?

00:26:17: I hope this episode has shown that challenging times like these require leaders who can navigate uncertainty and ambiguity well.

00:26:26: The cost-and-headcutting game is not sustainable, Not even for shareholder profits.

00:26:32: Two of the key figures in the AI race just showed us.

00:26:36: Building trust with people around you And establishing psychological safety are among most important moves to make right now.

00:26:44: In reflection on Thursday I will give you the key practices that help you do this well.

00:26:49: For today, i just want to leave you with one question.

00:26:52: If someone on your team knew a way to improve The entire teams performance by ten percent using AI technology Would they tell you?

00:27:03: That's it for Today.

00:27:04: Connect With Me and Tell me how AI Technology is impacting Your work.

00:27:09: Please share feedback And stay curious.

00:27:12: Stay tuned.

00:27:13: Thanks for tuning into Newark Playbook where work is better because people matter.