When systems lose their soul
In the institutional graveyard, some of the most painful deaths aren’t closures—they’re transformations that leave original participants wondering what happened to the mission they once believed in. These systems didn’t die; they drifted, abandoning their core purpose for compliance metrics, bureaucratic safety, or performative excellence, leaving behind people who remember when the work actually mattered.
The pattern is remarkably consistent: a simple, focused mission captures passionate early participants, achieves meaningful impact, then systematically dismantles everything that made it special in pursuit of risk mitigation or measurable outcomes. The most devastating examples share a common thread—they forgot that their earliest believers weren’t just participants, but the reason the system had value at all.
Medical education became checkbox theater
Medical education started with a beautiful premise: experienced doctors sharing what actually worked with newer doctors. The Flexner Report of 1910 established that physicians should learn from both science and clinical experience. For decades, continuing medical education (CME) meant attending grand rounds where senior doctors presented challenging cases, reading journals about new treatments, and learning procedures from colleagues who’d perfected them.
Then came the compliance avalanche. By 2024, doctors spend 40-60 hours annually clicking through mandatory modules. A 2023 survey of 2,000 physicians found that 73% of their required CME covered compliance topics: HIPAA violations (12 hours), billing codes (8 hours), sexual harassment training (4 hours), fire safety (2 hours), password security (2 hours). One emergency physician told me: “I spent 6 hours learning about elevator safety but zero hours on the new sepsis protocols that could actually save lives.”
The drift accelerated with digital platforms. Companies like Medscape and CMEList turned education into engagement metrics. Modules are designed to take exactly the minimum required time—if the state mandates 2 hours on opioid prescribing, the module has enough slides to fill 120 minutes, regardless of whether it takes 20 minutes to teach the actual content. Doctors report opening modules in background tabs while seeing patients, clicking “next” every few minutes to maintain the illusion of engagement.
The numbers tell the story: 3 billion clicking simulator.”
Productivity tools became procrastination systems
Notion launched in 2016 promising to be “the all-in-one workspace.” Users could write, plan, collaborate, and organize. The early adopters were individuals seeking simple, flexible note-taking. The tool was fast, minimal, and focused on content rather than formatting.
By 2024, Notion became something else entirely. Browse YouTube and you’ll find 10,000+ videos on “Notion setup,” with popular ones reaching 2 million views. There’s a $50 million market for Notion templates—people selling elaborate dashboards that take days to configure. Users report spending 15-20 hours setting up their “perfect productivity system” before writing a single note.
The pattern repeats everywhere. Roam Research users spend weeks learning “bi-directional linking” before taking notes. Obsidian users install 50+ plugins creating what they call “second brains” that are too complex to actually use. One user confessed: “I’ve spent 200 hours building my Zettelkasten system and written 12 actual notes.”
SuperMemo, which I mentioned getting obsessed with, represents peak productivity dysfunction. The app uses complex spaced repetition algorithms requiring users to review items at mathematically optimal intervals. But users spend more time optimizing the algorithm parameters than learning. The subreddit has 1,000+ posts debating interval modifiers but only 50 posts about what people actually learned.
Social networks optimized for engagement over connection
Twitter’s 2006 prompt was “What are you doing?” Users shared mundane life updates: “eating a sandwich,” “waiting for the bus,” “can’t sleep.” It was intimate, human, boring in the best way. The 140-character limit forced conciseness but not performance.
The prompt changed to “What’s happening?” in 2009, but the real shift came with algorithmic timeline in 2016. Engagement became the metric. Tweets that triggered responses—especially angry ones—got promoted. By 2024, Twitter (now X) optimizes for “dwell time” and “interaction rate.” The result: false information spreads 6x faster than truth, quote-tweet dunking generates 3x more visibility than original thoughts, and rage-bait accounts with 100K followers earn $10K monthly from engagement payouts.
LinkedIn underwent similar drift. Launched in 2003 for professional networking, early posts were job changes and industry articles. By 2024, the feed is what users call “LinkedInfluencers” sharing fabricated inspiration stories. Analysis of 10,000 top posts found: 34% were humble-brags disguised as lessons, 28% were corporate virtue signaling, 23% were “I gave a homeless person a job” fiction, and 11% were actual professional content. The platform’s algorithm specifically promotes “dwell time content”—posts that keep users scrolling rather than connecting.
The mechanics of institutional drift
These drifts follow predictable patterns:
1. Metrics replace missions. Medical education measures hours not outcomes. Productivity tools measure setup time not output. Social networks measure engagement not connection. When you optimize for what’s measurable, you get what’s measurable—not what matters.
2. Safety beats utility. Every failure adds restrictions. No one removes old rules. Medical education adds compliance modules after every lawsuit. Platforms add content policies after every scandal. Eventually, preventing failure becomes more important than enabling success.
3. There’s a certain nostalgia that can’t be recovered. Early Facebook was just your friends from school posting party photos. Twitter was people sharing what they had for lunch. These moments are gone, and no amount of “fixing” these platforms will bring them back.
4. Intermediaries extract value. Certification bodies, compliance consultants, and platform companies insert themselves between practitioners and practice. They don’t add value—they add process, taking money while making the actual work harder.
5. Original believers leave or burn out. The people who understood why something existed get replaced by people who only understand how it operates. The mission becomes the machine.
What I’m doing about it
I can’t fix these systems—they’re too big, too entrenched, too invested in their current form. But I can make the drift visible.
My medical education project maps what doctors actually need to know versus what they’re required to learn. Not to reform CME, but to help doctors navigate around it. If you have to spend 40 hours on compliance, at least know which 4 hours contain actual medical knowledge.
The tools I build acknowledge that drift is inevitable. Instead of pretending systems work as intended, they create navigation layers—ways to find the 10% of value buried in the 90% of bureaucracy.
Because you can’t stop institutional drift. But you can help people find what they came for, even after the system forgot why it exists.