Human-Centred AI: Turning performance data into dialogue, HR News, ETHRWorld

Time of India

a day ago

Advt
Advt
By & ,
ETHRWorld Contributor
Join the community of 2M+ industry professionals. Subscribe to Newsletter to get latest insights & analysis in your inbox.
All about ETHRWorld industry right on your smartphone! Download the ETHRWorld App and get the Realtime updates and Save your favourite articles.
Once upon a time in the kingdom of Pratham Garh, the wise but slightly anxious king, Raja Dutta Dev, grew tired of hearing mixed reviews about his courtiers. The commander said the soldiers were lazy. The poets claimed they were overworked. The minister of productivity insisted that the royal astrologer had been idle since Mercury went retrograde. Raja Dutta Dev sighed and declared, 'I want objectivity! No more opinions. Let's bring in… Artificial Intelligence !'He summoned the kingdom's smartest mind, Acharya Algorithm Anand, who built the first-ever performance system: RajBot 1.0—a magical machine that tracked every scroll sent, every sword swung, and every tea break taken. Soon, chaos followed. The court jester got flagged for 'non-serious behaviour during work hours'—which, frankly, was his actual job. The head chef was accused of 'excessive stirring without measurable output.' The treasurer was put on watch for 'counting coins with no documented outcomes.' Worst of all, General Veer Pratap, who had just returned victorious from war, was given a low score. Why? Because he'd missed 30 royal Zoom calls while saving the kingdom.Raja Dutta Dev called Acharya Algorithm Anand and said, 'I wanted insight, not insult. Let's build a system that helps my people—not humiliates them.' The Acharya bowed, and together they rewrote RajBot's code with four royal principles: Clarity, Context, Compassion, and Consent.These days, AI tools are doing more than filtering cat photos or recommending your next binge-watch—they're shaping careers. In the modern workplace, algorithms now assess productivity, predict potential, and even try to read your emotional state through sentiment analysis. Promotions, bonuses, and hiring decisions are increasingly data-driven—which sounds efficient, until you realize no one quite knows how the data works. I've experienced it first-hand: asynchronous interviews scored by AI, video assessments with no human follow-up, and automated tests that somehow decide if you're 'collaborative' based on word choice and webcam lighting. Sure, it's scalable, but it's also disorienting. You're left wondering whether you were evaluated on your answers or your apartment's background lighting. These systems promise objectivity, but without transparency or empathy, they risk reducing people to data points. The fix isn't to scrap AI altogether—but to design it better. We need systems that combine analytical power with human dignity, fairness, and trust.Organizations are rolling out AI performance tools faster than HR can say 'algorithmic accountability.' From tracking keystrokes to decoding Slack emojis, companies are using tech to find out who's productive and who's just good at pretending during meetings.Take Amazon, for instance, where warehouse employees are timed between scans like they're on a game show—with the prize being not getting penalized for a bathroom break. Or Meta, where engineers are reportedly rated by lines of code—because apparently, in 2025, size does matter. Meanwhile, sentiment analysis tools try to detect how happy (or sarcastically dead inside) you sound in emails.Sentiment analysis, while sophisticated, can misinterpret harmless frustration or workplace humour as negativity, affecting career advancement or compensation decisions unfairly. Algorithms trained on historical data can unintentionally replicate biases related to gender, age, or cultural differences, disproportionately disadvantaging certain employee groups. Bias? Oh yes. Historical data might think women don't lead teams or that older employees aren't tech-savvy—forgetting that half your IT department is run by people who still remember floppy disks.These hyper-monitoring systems can erode trust, stifle creativity, and make everyone feel like they're living in a Black Mirror episode—except with worse lighting and no plot resolution. A culture of surveillance, however well-intentioned, undermines genuine engagement and innovation. Ironically it reduces the very productivity these analytics are intended to enhance.Leaders must rethink not only what they measure but how and why they measure it.Designing AI that's ethical doesn't mean sending it to therapy. It means embedding four essential principles that remind the system—and its creators—that people are more than the sum of their Slack messages.1.: Employees should know what's being tracked—because nobody wants to find out their Zoom background plants were analysed for stress levels. Spell it out clearly: What's being measured? Why? And is there a secret leaderboard?2.: Don't measure your graphic designer like you do your logistics manager. One creates magic with pixels, the other makes sure your coffee arrives on time. Different strokes for different spreadsheets.3.: Don't just tell employees they 'scored low on collaboration.' That's vague enough to sound like a buzzword salad. Offer them real, actionable steps—like 'schedule one brainstorming session this week' or 'talk to Dave more nicely.'4.: Data should start conversations, not end them. Let people explain themselves. Maybe the coder didn't log time because they were mentoring a junior. Maybe someone's 'negativity score' came from shouting at the printer, not their boss.When analytics adhere to these principles, companies do more than collect data. They build trust, motivation, and authentic engagement.Picture a product management team at a fast-paced SaaS company—where people type faster than they talk, and meetings are scheduled like cricket matches: endlessly.Now, instead of dumping yet another cookie-cutter dashboard onto already dashboard-fatigued employees, the company rolls out a performance analytics tool with an unexpected twist—it's actually ethical and human-centred. Yes, really.First, the tool is transparent. Not 'terms and conditions nobody reads' transparent—but genuinely clear. Everyone knows exactly what's being tracked—project timelines, collaboration frequency, stakeholder feedback—and not whether they stretched awkwardly during a Zoom call. Even better, the system respects context. It doesn't compare the UX designer who lives in Figma to the product manager who lives in meetings. No more apples-to-orange comparisons—or worse, apples-to-orange-juice. And here's the best part: instead of just throwing raw data at you like a robotic boss yelling 'Do better,' the tool gives useful, personalized nudges. Like 'maybe don't schedule seven meetings on a Monday' or 'this Slack thread could've been an email.' Employees can opt in for deeper insights—or not. It's analytics with consent, not analytics with handcuffs.The result? People stop feeling like they're being watched by HAL 9000 and start feeling like they're being coached by someone who's... well, not emotionally dead inside. Productivity goes up, not because Big Brother is watching, but because the team finally has a system that supports growth without breathing down their necks.Even small changes can start the shift from creepy to constructive:1.Get HR, data science, ethics nerds, and maybe a philosopher into one room. Trust me, it's more fun than it sounds.2.Regularly check if the algorithm is showing favouritism. Bias doesn't wear a name tag.3.If a metric can't be explained in plain English (or memes), don't use it. Prioritize simple visualizations, clear definitions, and intuitive interactions.4.Ask your people: 'Do these tools help, confuse, or mildly terrify you?' Then, listen. Regular dialogue ensures analytics evolve with human needs, not against them.5.Let employees choose what's shared. No one wants their mood swings graphed without permission. Whenever possible, give employees control over what data they share and how it's used. Offering meaningful consent options helps build trust and reduces anxiety.In the age defined by data and algorithms, measuring performance isn't the problem. Measuring it like you're building a robot army is. We need systems that understand context, invite conversation, and actually help people grow—not panic every time they step away from their keyboard. Let's build analytics that see people as people—not Wi-Fi-enabled productivity bots. Because in the end, the best workplaces aren't powered by data.They're powered by trust, coffee, and that one perfectly timed cat meme that makes everyone believe again—in humanity and Mondays.Even in heaven, if your AI can't tell the difference between divine work and 'non-compliance,' it needs a rewire.