Please, Thank You, and Other Expensive Habits
By now, I’m sure we’ve all heard that politeness is apparently costing OpenAI tens of millions in compute. A fascinating little data point, right? Almost begs for a shrug and a move on.
Before we go around teaching preschoolers that the magic word is no longer “please” and they should just be tapping an enter key, let’s take a second to think about the implication of Sam Altman’s revelation.
It’s more than just tech trivia; it’s a symptom.
The simple act of saying "thank you" to a machine that feels nothing has been quantified, metered, and priced like electricity. A courtesy becomes a cost center. An interaction becomes a transaction analyzed for efficiency.
Makes sense though if you really think about it, in a bleakly logical way. The quantification of everything continues its relentless march.
Commodity futures have long assigned numeric value to real world phenomena, trying to wrestle messy realities of supply, demand, and weather onto a predictable grid.
Now prediction markets let you hedge on whether JD Vance will be the next pope or if aliens will land tomorrow. Why not price politeness per token?
It seems we've reached a point where our most innate social impulses, the very things that grease the wheels of human connection, register as computational drag.
This isn't accidental; it's the logical endpoint of a specific worldview dominating most of Silicon Valley. This belief system whispers that the messy, unpredictable heart of human interaction and creativity is a bug, not a feature. It’s an obsessive desire to measure, predict, and optimize everything. To reduce friction. To make everything data-driven. To find one truth.
Let's call it Quantiphilia.
The true believers, the Quantiphiles, seem convinced that anything truly valuable can, and should, be expressed as a number.
You know these guys. The ones who believe (like truly) that SAT scores reflect innate intelligence. Mostly because they themselves got perfect scores despite not being able to unpack their own relationship with correlation vs. causation.
This pervasive drive to quantify inevitably bumps up against the messy, unquantifiable stuff that makes life interesting (and functional).
Which brings us back to that OpenAI anecdote and the question it sparks.
What does humanity devoid of any and all friction look like?
It’s the hidden tax in feeds designed purely to confirm existing biases, minimizing the friction of encountering different ideas.
It’s reaching for the same algorithmically curated playlists, avoiding the 'inefficiency' of discovering something unexpected.
It’s sanding the edges off posts hoping to appease keyword scanners, sacrificing nuance for optimized delivery.
We streamline ourselves into comfortable grooves, sometimes mistaking convenience for actual progress, frictionlessness for freedom.
In VC world, this pursuit of predictability can reach levels of high absurdist theater. Pitch decks become standardized templates, minimizing the friction of diverse presentation styles.
Growth curves are demanded with Newtonian certainty, avoiding the messy reality of startup trajectories. VCs, sometimes the most ardent believers in quantifiable truths, preach disruption while demanding spreadsheets that look like rigid train schedules.
The mantra shifted from "break things" (inherently friction-filled) to "predict everything" (friction-averse). We crave certainty so badly we risk optimizing away the very chaos, the productive friction, that births breakthroughs.
The irony? Almost physically palpable.
The folks LinkedIn agreed about the absurdity of predictability.
The relentless quest for predictable outcomes strangles the unpredictable insights. Prediction isn't creation. Optimization isn't discovery. It’s just… tidying up the messiness that might actually matter.
This isn't just about pitch decks; it’s a creeping issue visible everywhere this drive for quantification takes hold.
That pursuit of frictionless perfection inadvertently bleeds the vitality out of experience. Customer service becomes automated loops engineered for deflection, eliminating the friction (and potential resolution) of genuine human interaction.
Education risks becoming feedback systems optimizing for metrics while potentially dulling the inefficient, friction-filled process of real curiosity.
We strip away the waiting, the ambiguity, the potential for surprise—all forms of friction—and what’s left often feels sterile, predictable, deadened. It's the uncanny valley of interaction.
When nothing unexpected, nothing inefficient can happen, the spark vanishes. It's a world without grit, without texture. It's a flatline.
It makes sense, really. Computers are good at many things. Except one.
Randomness.
Cloudflare’s solution to computers’ little randomness problem.
Human Randomness: The Value of Inefficiency and Friction
So if removing friction leads to a flatline, what's the valuable friction we're losing? What does the messy, unpredictable, inefficient stuff that Quantiphilia tries to eliminate actually look like?
It's not just chaos; it's the source of valuable, unexpected outcomes. Life often teaches us that the best results arise from the places we least expect.
In venture capital, we supposedly chase this same idea – pursuing outliers according to the power law principle, backing the renegades who will unlock massive value.
Paradoxically, however, we then build mountains of analysis, ultimately trying to fit these very ideas into neat boxes using predictable heuristics.
So, where does the truly disruptive value, the kind venture capital claims to seek, actually originate if not purely from structured analysis? It arises precisely from those elements that resist easy quantification and prediction.
The paradox highlights that while we strive for predictive certainty, the real breakthroughs often stem from the unpredictable nature of human thought and interaction itself.
This human capacity for randomness, for generating productive friction, is the core differentiator in an age drowning in cheap prediction. Consider the spectrum of valuable inefficiencies:
Intuition: That illogical leap connecting unrelated fields; friction against linear thought.
Serendipity: Stumbling onto the cure while looking for glue; the friction of unexpected discovery.
Creativity: Genuine novelty, not just remixing hits; friction against predictable patterns.
Social Nuance: Reading the room, navigating ambiguity; the vital friction of complex human interaction.
Curiosity: Pulling threads that seem pointless; the friction of exploring outside optimized paths.
These aren't "soft skills" to be smoothed over. As predictable efficiency becomes a commodity, these sources of beneficial friction, of valuable inefficiency, are the hard currency.
This is the signal worth tuning into amidst the frictionless noise.
Why Machines Can't Replicate Beneficial Friction
AI finds the most probable, often the most frictionless, path based on past data. It operates on probability, smoothing out anomalies. Humans? We operate on possibility.
We can introduce productive friction. We jump the tracks. We generate true novelty precisely because we aren't perfectly optimized.
That OpenAI politeness tax? A perfect illustration. Quantiphilia sees politeness as costly friction.
But that "inefficient" dance of human connection, the complex 'why' behind a simple 'thank you,' is the valuable friction that builds trust and understanding
Something AI can only mimic at computational cost. AI is engineered to minimize deviations, to reduce friction.
But human creativity, connection, and breakthrough often live precisely in those deviations, in that necessary grit. Machines are optimized to crave smoothness; humans can find value, even beauty, in the friction.
So let them have the predictable, frictionless parts. Let the bots optimize. Fantastic.
It clarifies what they can't do.
They can't replicate our capacity for beneficial randomness, for introducing the grit and inefficiency that leads to genuine breakthroughs.
They can't manufacture the spark that arises from that friction.
Investing in the Human Algorithm: Re-Wilding Through Friction
Okay, so if unpredictability, inefficiency, and friction are actually our premium, how do we cultivate them in a world obsessed with eliminating them?
We've spent decades trying to streamline everything out, thanks to the pressures of Quantiphilia. Reversing course takes deliberate effort.
Organizations: Need to build "friction tolerance"
Embrace creative conflict by deliberately seeking out differing opinions and perspectives (good friction!)
Mine the overlooked margins by venturing into overlooked spaces to find the brilliant minds with unconventional backgrounds others dismiss (hard!)
Individuals: Need to consciously reclaim our own inefficiency
Wander intellectually (inefficient!)
Learn to trust your gut (often messy!)
The winning strategy isn't making humans into frictionless cogs. It's creating conditions where our unique, messy, gloriously unpredictable, friction-filled human algorithm thrives alongside machine efficiency.
Embrace the Chaos (and the Friction). It's Where Value Lives Now.
So, AI masters prediction and smooth efficiency.
Great.
Let it have the certainty game.
Its success only throws into sharper relief where the real, durable value lies: in the messy, unpredictable, inefficient, friction-filled, glorious human spark.
The sudden idea born from chaos, the counterintuitive insight sparked by friction, the serendipitous connection found off the optimized path, the nuanced empathy that defies quantification.
These aren't things you optimize into existence by eliminating friction, no matter how much the devotees of Quantiphilia might wish it were so.
The purely quantitative view offers cold comfort but ultimately leads toward stagnation, toward that frictionless flatline. But the good news?
The future belongs to those ready to dance with ambiguity, harness serendipity, embrace necessary friction, and generate novelty from the beautiful chaos of being human.
As predictable efficiency becomes table stakes, the premium on the innately human knack of generating valuable surprises through inefficiency and friction is set to skyrocket.
So don't just tolerate your inner weirdness, your inefficiencies, your friction against the smooth path; cultivate them.
Trust the illogical leap.
Follow the inefficient curiosity. Because in a world increasingly awash in algorithmic certainty and driven by Quantiphilia, the most disruptive, valuable, and ultimately human force?
It's that gloriously unpredictable spark, born from the grit and chaos.
And that, thankfully, will always remain beautifully, stubbornly, hopefully out of scope.
Joe Alalou is a co-founder and General Partner at Daring Ventures, a pre-seed fund investing in software that amplifies uniquely human skills in complex, relationship-driven fields.
If you enjoyed this, please share or subscribe!