# Tight feedback loops 7/14/26 I wasn't satisfied by my analysis in [[Dead players]]. It came from a real gut feeling, but I don't know that I pointed to the thing I intended to. I kept it up because it (inadvertently) started circulating more broadly; and I didn't want to create link-rot. But with the benefit of time + newer conversations, I'd probably channel that feeling into a different — if still related — conclusion. The perceived excesses of the 2010s — billions poured into crypto, juice machines, whatever — eventually triggered a corrective fetish: for real, tangible, *business*. I participated in this reversion myself, moving from [[Ramp#^6f2a45|crypto]] to _the_ now-exemplary [[Ramp#^5618f2|B2B SaaS]] of our time (though it wasn’t in fashion yet back then). The hallmarks are real & burning customer problems, with immediate time & money to be saved. The people I know who experienced success under this model, have since switched to _newer_ "real & burning" problems. Many of them don't even _leave_ their previous gig, before collecting letters of intent from business customers, booking some amount of ARR on paper, touring Sand Hill Road, and drafting term sheets. This is a de-risking, yes — risk having been a component of my original essay — but honestly, who *cares* if they de-risked it. That isn't the part I really regard with suspicion. **It's the veneration of tight feedback loops.** * Targeting markets that are demonstrably large, _now_ * ARR — as *fast* as possible * A sharp unwillingness to be early to anything, expressed as "timing is everything" * The idea that last-movers emerge from *deferring* something until the model is already proven; and not instead from making sure that what you're digging is a _moat_ from day one, or from doubling-down on contrarian — unfashionable — insights that other people would only follow once it's already too late. * Building low-stakes, scrappy products as "wedges," in hopes that low-stakes buy-in today will magically graduate into high-stakes buy-in "one day" * [Indeterminate optimism](https://blakemasters.tumblr.com/post/23435743973/peter-thiels-cs183-startup-class-13-notes/amp#:~:text=It%20is%20thus%20best%20to%20just%20figure%20out%20incremental%20things%20to%20do%20and%20wait%20for%20progress%20to%20come.), and a skepticism of longer-term planning Many of the "highest impact" people you and I personally know, were primed by 1-4 year vesting schedules. They were motivated by having the greatest impact possible, while knowing they only have 1 or 2, _maybe_ 3-4 years to _have_ it before hitting reset. They don't want to have the highest impact possible; they want to have the highest impact possible *within a short window of time*. They'd move on from an idea — their "company" only being a loose scaffold around it — if it couldn't book revenue within its first few months. Most long-term games have lost prestige in tech. AI research became high-prestige after it was proven and consensus. No one chooses to do, nor build, the thing that would take them months instead of weeks. No one wants to defer ARR, in favor of a product that redefines or monopolizes a category over years, decades, centuries. You look either naive or narcissistic for even trying to do this. What could you possibly know? Have some humility! Tech nowadays is less Kanye — extending projects by years, spending years on albums everyone *instantly* dislikes, betting on _yourself_. It's more 2016-era Young Thug — new songs all the time, oriented to the market, only a handful of them durable (and only so in piecemeal) over time. Few people want their own best verse to happen on a Kanye album; they'd rather it happen on their own, less-ambitious project. Time corrects all; returns to passing the marshmallow test will continue. If anything, they'll grow: *everyone*'s agents are fast nowadays. They're also good. But how many of them operate with genuinely long time-horizons? How many humans or agents are genuinely _selfish_ in what they value, to the point of being unique and irreplaceable? <div style="left: 0; width: 100%; height: 152px; position: relative;"><iframe src="https://open.spotify.com/embed/track/4NVNapccSX7E5JLiW0uQEy?utm_source=oembed" style="top: 0; left: 0; width: 100%; height: 100%; position: absolute; border: 0; border-radius: 12px;" allowfullscreen allow="clipboard-write *; encrypted-media *; fullscreen *; picture-in-picture *;"></iframe></div> > "_Only real music's gonna last_. *All that other bullshit, is here today and gone tomorrow.*" Prestige-hunters — in tech, and everywhere else — are partial to tight feedback loops. They need proof, fast, or they'll move on. They're the "loosely held" in strong convictions. Almost by definition, then, all of the alpha is moving into things that you *can*'t see with your eyes when you first start preparing for them. Into operating _without_ definitive proof or certainty in the near term. Into a hint of *selfishness,* rather than so-called *humility* in the face of the customer and competition. Most people don't believe in their own ability to envision the currently-invisible. They'd rather start from the visible, book the quickest wins possible, and convince themselves that they'll *eventually* pivot to a visionary approach: once they've had their fill of positive feedback. A tight feedback loop is ultimately a tight *pull* originating from outside of you — from your customers, from your competition, whoever. I'd claim that over time, it *precludes* your capacity for sovereign, creative thought, and atrophies your ability to apply contrarian — offensive — insight. Sometimes it's a way to cope with your inability to see further into the future than the present-day; to zoom out of Twitter, or from the job you've held since before Claude Code and the culture you internalized in that time. You just haven't realized it yet. And upon reflection: this "you" was me too, for the longest time. Might still be!  *Featured in [TL;DR Tech](https://tldr.tech/tech/2026-07-15#:~:text=Prestige%2Dhunters%20are%20partial%20to%20tight%20feedback%20loops%20as%20they%20need%20proof%2C%20fast%2C%20or%20they%27ll%20move%20on.)* %% ![[Pasted image 20260714121229.png]] %%