AI Misleading Scores LinkedIn Drop

Lachlan Reed

So, this developer, right? Best on the team. I- I mean, absolute rockstar. For six months, this person is pulling down perfect scores on the- the company's automated tracking dashboard. Commits are on time, ticket resolution is blazing fast, the AI tool that monitors their workflow is essentially throwing digital confetti. But here's the thing. They hadn't actually cared about the job since last winter. They were completely checked out. Like, mentally sitting on a beach in Bali. The body was at the desk, the metrics were green, but the actual human being? Long gone.

Dr. Zara Sterling PhD

That is a classic case of silent attrition, Lachlan. And what is fascinating from a psychological perspective is that the system was perfectly satisfied with a ghost. The algorithm was measuring output, but it was completely blind to the fact that the employee's discretionary effort, their- their innovation, their cognitive presence, had plummeted to zero.

Simon Carver

Wow, a ghost with a perfect performance review. That is wild. Welcome to The Human Workforce, everyone. I'm Simon Carver, and today we are looking at this bizarre disconnect. We've got Lachlan Reed in his shed in Sydney, and we are absolutely thrilled to have guest host Dr. Zara Sterling, PhD, an organizational psychologist who studies how we actually function inside these high-tech systems. Zara, this scenario... it feels like a giant contradiction. We're deploying these incredibly advanced AI tools to make everything better, but people are just... shutting down?

Dr. Zara Sterling PhD

Exactly, Simon. It is what we call the growth paradox. Right now, in 2026, organizations are scaling up their artificial intelligence capabilities at an exponential rate. But because they are focusing almost exclusively on systemic optimization, they are simultaneously deoptimizing the human beings within those systems. They are- they are treating human psychology as a static variable, and that is a massive strategic error.

Lachlan Reed

It's like... it's like buying a high-performance racing engine but forgetting to put oil in the- the gearbox. You're tracking how fast the pistons are moving, but the gears are just grinding themselves to dust. We're using AI for scheduling, task assignment, even deciding who gets a bonus, but we're totally hollowing out the actual experience of doing the work. You get managed by a machine, and pretty soon, you start acting like one. Or worse, you just stop caring.

Simon Carver

And that's the "silent" part of silent attrition, right? It's not like they're making a big scene or throwing a resignation letter on the desk. They just quietly withdraw their humanity.

Dr. Zara Sterling PhD

Yes, the psychological contract is broken. In organizational research, we see that when individuals feel monitored rather than mentored, they shift from intrinsic motivation... doing something because it is inherently satisfying... to a purely transactional mindset. They do exactly what the algorithm requires to avoid a red flag, and not a single drop more.

Dr. Zara Sterling PhD

To really understand how this happens, we have to look at the psychological mechanisms of what I call the hollowing-out scenario. There are five distinct drains, and the first is algorithmic dehumanization. This occurs when feedback is stripped of all human context. If a system flags you because your typing speed dropped by eight percent on a Tuesday morning, it doesn't know... and it cannot care... that you were up all night with a sick child, or that you spent three hours helping a junior colleague troubleshoot a massive problem that doesn't show up in your personal metrics. The data capture is clean, but the lived experience is incredibly messy.

Lachlan Reed

Ah, that's it. It's the absolute death of empathy. And then you get autonomy erosion, which is the second one. Look, I love working on my old motorbikes, right? I- I love it because I get to figure out the puzzle myself. If some automated voice was in my ear telling me exactly which wrench to turn, every ten seconds, based on some global efficiency database... I'd- I'd probably throw the wrench through the window. But that's what we're doing to workers! The algorithm decides the "optimal workflow," and if you deviate, you get flagged. You- you stop using your own brain because the machine has already decided for you.

Simon Carver

I would pay to see you throw that wrench, Lachlan. But seriously, that leads right into the third drain: transparency collapse, doesn't it? If the system is making decisions about your shifts, or your performance, but you have no idea *how* it's deciding, you start getting paranoid.

Lachlan Reed

Oh, absolutely! It's like... remember those old red-light cameras? If you don't know where they are, or why they're flashing, you're not driving better... you're just staring at the speedometer the whole time, terrified of getting a ticket. You start performing for the camera instead of focusing on the road. People end up gaming the system. They learn what behaviors make the algorithm happy, even if it's completely useless work in the real world.

Dr. Zara Sterling PhD

That is a well-documented phenomenon. We call it "gaming the metric." When transparency collapses, trust is replaced by surveillance anxiety. But the damage goes even deeper. The fourth mechanism is skill atrophy. This is a quiet, slow-burning crisis. When we outsource critical cognitive tasks to AI... things like complex problem-solving, resolving difficult customer complaints, or giving constructive feedback... we stop using those neural pathways. It is the "use it or lose it" principle of neurobiology. If a manager relies on an AI to write all their team feedback reports, that manager is not learning how to navigate difficult human emotions. Five years down the road, you have a leadership team that is completely incapable of holding an authentic, difficult conversation.

Simon Carver

Man, that's terrifying. You literally forget how to be a leader. And then, I guess, the final blow is purpose displacement. We are meaning-seeking creatures, right? We need to feel like our work actually matters. But an algorithm can't measure "meaning." It measures quotas, response times, lines of code. If you reduce the entire workday to a series of numbers to be optimized, you extract the soul of the work.

Dr. Zara Sterling PhD

Yes, purpose cannot be quantified. When optimization replaces purpose as the primary organizing principle, the workplace becomes psychologically barren. And that barrenness is where silent attrition thrives. It is the ultimate cost of a hollowing-out strategy.

Lachlan Reed

But it- it doesn't have to be this depressing, does it? I mean, we're not saying AI is the bogeyman here. If we get the design right, it can actually be... well, liberating. Zara, you talk about this "liberation scenario." What does that look like on a normal Tuesday?

Dr. Zara Sterling PhD

It looks like a massive reduction in what we call "work about work." Think about the sheer volume of administrative overhead that clogs up the average knowledge worker's day. Organizing meetings, writing status updates, retrieving buried files, filling out compliance forms... these are cognitive taxes that drain our energy. When AI is deployed to absorb that repetitive, low-value friction, it returns something incredibly precious: time and cognitive bandwidth. It frees us to do the things that actually require human intelligence... strategic thinking, mentoring, creative problem-solving, and deep relationship building.

Simon Carver

Right, so instead of using AI to micromanage the human, we use AI to manage the- the digital clutter *for* the human. It's a complete shift in perspective. But to make that transition, you need some solid rules, right? You mentioned some design principles.

Dr. Zara Sterling PhD

Yes, we need a rigorous framework. The first principle is transparent purpose alignment. If you deploy an AI system to assist with management or tracking, the workforce must understand exactly why it is there, what data it is collecting, what it *cannot* see, and how human judgment fits into the equation. Transparency doesn't make monitoring invisible, but it removes the toxic anxiety of the unknown.

Lachlan Reed

It's like telling your crew, "Hey, we're putting GPS on the trucks, not to spy on your lunch breaks, but to help optimize the routes so you're not stuck in gridlock for two hours." Once they know the "why," they're not trying to cover up the sensor with a- a meat pie wrapper, you know?

Dr. Zara Sterling PhD

Exactly, Lachlan. And that leads directly to the second principle, which is human primacy in high-stakes decisions. This is non-negotiable. While an algorithm is excellent for optimizing a complex schedule, it should never have the final say on decisions that alter a person's career trajectory... things like performance evaluations, promotions, compensation, or termination. A human being must be meaningfully involved. In organizational psychology, we refer to this as procedural justice. People need to feel that their complete, nuanced context was evaluated by someone who can actually comprehend human experience, not just parse a data set.

Simon Carver

Procedural justice. I love that term. It's the difference between being judged by a person who can look you in the eye, versus being processed by a cold machine. It changes everything about how we respect the organization.

Simon Carver

Before we dive into the rest of these design principles, I want to take a quick second to say a huge thank you to Dr. Zara Sterling for joining us today. This conversation is so vital right now. If you're listening and finding value in this, please hit that subscribe button, share this episode with a colleague or a manager who needs to hear it, and leave us a review. It really helps us keep bringing these deep dives to you. Alright, let's get back to it. Zara, we've got three more principles to cover for this liberation scenario.

Dr. Zara Sterling PhD

Yes, the third principle is capability amplification rather than capability replacement. When an organization decides to implement an AI tool, the leadership must ask a critical question: Will this tool make our people more skilled and adaptable over time, or will it make them dependent and deskilled? If the tool simply replaces human cognitive processes without providing a path for the worker to grow into higher-value tasks, you are creating a fragile, brittle workforce. We must design systems that actively teach and elevate the user.

Lachlan Reed

Yeah, it's like... if you use a calculator to do basic math, that's fine, it saves time. But if you forget how to do addition entirely, you're in trouble when the battery dies. We need tools that help us see patterns we missed, not tools that just tell us "click this button and don't worry about why." And that ties into the fourth principle: meaningful feedback architecture.

Dr. Zara Sterling PhD

Correct, Lachlan. AI shouldn't be the one delivering the feedback to the employee. Instead, the AI should process the complex data to give the *manager* deeper, richer insights, which then fuel a high-quality, human-to-human coaching conversation. The technology should enhance the relationship, not replace it. The manager still has to do the hard work of connecting, but now they have better data to support that connection.

Simon Carver

That makes so much sense. It turns the manager back into a coach, rather than just a metrics-enforcer. And what about the fifth principle, autonomy by design?

Dr. Zara Sterling PhD

Autonomy by design means building pathways within the system that allow employees to contest algorithmic recommendations, to provide qualitative context, and to exercise their own professional judgment. When a system allows for human agency, it fosters a sense of ownership. And psychological ownership is the literal bedrock of innovation and resilience. If you strip away agency, you strip away the very capacity to adapt when things inevitably go wrong.

Lachlan Reed

So, if you're a leader sitting in a boardroom right now, how do you actually know if you're on the- the bad path or the good one? Is there a way to check under the hood?

Dr. Zara Sterling PhD

There is. We can use four simple diagnostic questions as an organizational "check engine light." First: Do your people actually understand how the AI systems affecting their work make decisions? Second: Over the last year, have your employees had *more* high-quality, growth-oriented conversations with their managers, or fewer? Third: Have your people become more capable, independent, and creative, or are they increasingly dependent on systems they don't comprehend? And fourth: When your employees describe their work, do they talk about meaningful challenges and relationships, or do they talk about avoiding system flags and hitting algorithmic metrics?

Simon Carver

Man, those are some incredibly convicting questions. If you're honest, a lot of companies are going to look at that and realize their check engine light is blinking red. So, what's the move? How do we re-engineer leadership for this new era?

Lachlan Reed

Well, for starters, we've got to stop measuring ROI purely on "efficiency." If you only measure how fast the conveyor belt is moving, you're going to optimize for speed and completely miss the fact that your workers are burning out and ready to bolt. You have to measure engagement, retention of your top performers, and how fast you're actually innovating.

Dr. Zara Sterling PhD

Precisely. And we must hold managers accountable for human development outcomes, not just output targets. If a manager is evaluated solely on their team's production numbers, they will inevitably default to using AI as a whip. But if they are also evaluated on the growth, psychological safety, and retention of their people, they will use AI as a tool to free up time for real leadership.

Simon Carver

It's about resisting that false choice between efficiency and humanity. The most successful organizations in the next decade won't be the ones with the absolute fastest algorithms... they'll be the ones that figure out how to make their technology and their people genuinely complement each other. It's a design challenge, but it's also a deeply human one.

Lachlan Reed

Well said, mate. It's about remembering that the human being at the desk isn't just a collection of metrics. If we want them to bring their best ideas, we've got to treat them like humans, not components in a machine. Alright, I think that's our time for today. Zara, thank you so much for bringing your brilliance to the shed.

Dr. Zara Sterling PhD

It was an absolute pleasure, Lachlan. Thank you, Simon.

Simon Carver

Thanks, Zara. And thank you all for listening. Take those diagnostic questions back to your teams this week. Let's build a workforce that actually thrives. We'll catch you in the next one.