The KPI of doing more

The pitch of AI has always been simple: make us more productive so we can do more with less effort. Work smarter, not harder. Reclaim your time. Focus on what matters. But here's what actually happened. The ceiling moved. Because AI expanded what's possible, it also expanded what's expected. The KPI didn't stay the same while the work got easier. The KPI went up because the capability went up. And now we're running just as hard as before, except the finish line is further away.

The productivity trap

A February 2026 study covered by Harvard Business Review put it bluntly: AI doesn't reduce work, it intensifies it. Researchers found that rather than lightening workloads, AI tools are increasing the speed, density, and complexity of work. The Wall Street Journal reported similar findings from a study of 164,000 workers: after adopting AI, time spent in email, messaging, and chat more than doubled, while time spent in focused, deep work fell by 9%. The pattern is straightforward. AI automates a task. The freed-up capacity doesn't become free time. It becomes room for more tasks. Organizations layer AI on top of existing processes, roles, and reporting structures, and the result is acceleration rather than simplification. BCG even coined a term for it: "AI brain fry." Their research found that workers dealing with high AI oversight expended 14% more mental effort, experienced 12% greater mental fatigue, and reported 19% greater information overload. Many described a "fog" or "buzzing" that forced them to physically step away from their screens. And here's the kicker. A study from METR, a nonprofit that evaluates frontier AI models, ran a randomized controlled trial with experienced open-source developers. When they used AI tools, completing tasks took them 19% longer than going without. But when asked, the developers reported that AI had sped them up by 20%. Feeling productive and being productive are two very different things.

The job application doom loop

Nowhere is this more visible than in the job market. Because AI makes it easier than ever to customize a resume and tailor a cover letter to each application, the expectation is that you should be applying to hundreds, even thousands of jobs. The KPI went up. The numbers are staggering. Job seekers now need somewhere between 400 and 750 applications to land a single offer. Success rates have collapsed from around 5% a decade ago to as low as 0.1% to 2% today. And it's not because people are less qualified. It's because everyone is submitting the same AI-polished, ChatGPT-optimized applications, and hiring managers can see right through them. A March 2026 Robert Half survey found that 67% of hiring managers say AI-generated applications are actually slowing down the hiring process, with about 20% reporting delays of around two weeks. The flood of near-identical, AI-crafted resumes has created so much noise that the signal gets lost entirely. So now you have this absurd arms race. Candidates use AI to generate more applications. Companies use AI to screen them out. The volume goes up on both sides, and the actual human connection, the thing that gets people hired, gets buried under algorithmic noise. Companies are increasingly falling back on networking and referrals because the application pipeline has become unusable. If you don't use AI, you fall behind. If you do use AI, you blend in with everyone else. There's no winning.

More output, same grind

This dynamic isn't limited to job hunting. It's everywhere. Thousands of CEOs surveyed in early 2026 admitted that AI had no measurable impact on employment or productivity at their companies, despite adoption rates climbing from 61% to 71% of firms. PwC found that only 30% of CEOs reported any revenue increase from AI, while 22% said their costs actually went up. The promise was that AI would free us to do higher-value work. The reality is that AI raised the baseline. What used to be impressive output is now just the minimum. A developer who ships twice as much code is expected to ship twice as much code. A marketer who produces content in half the time is expected to produce twice as much content. The goalposts moved, and they moved in one direction. This mirrors what happened with every previous productivity technology. Email was supposed to reduce the need for meetings. Instead, we got more meetings and more email. Smartphones were supposed to let us leave the office. Instead, the office followed us everywhere. AI is following the exact same pattern, just at a larger scale.

The real question

The uncomfortable truth is that productivity tools have never been about giving people their time back. They've always been about extracting more output. AI is just the latest, most powerful version of that cycle. So when someone tells you AI will make your life easier, ask: easier for whom? Because if your capability doubles and your KPI doubles to match, your workload hasn't changed. You're just running faster on the same treadmill. The real value of AI might not be in doing more. It might be in having the clarity to recognize when more isn't the point.

References