Men sans shirts
The last chat got me thinking about YouTube. I gather it uses how long you stay with a video to tag you for ads. Thing is I've been known, as I'm sure everyone has, to leave a video silenced on repeat in an interrupton.
YouTube's systems actually track several distinct signals beyond raw watch time, so leaving a video muted on a continuous loop doesn't fool the ad-targeting engine as much as you might think—though it definitely leaves some messy footprints in your profile.
Here is how those three behaviors interact with YouTube's tracking:
1. Continuous Looping (Diminishing Returns & Bot Filtering)
YouTube’s recommendation and ad algorithms are built to filter out incessant looping so people can't artificially inflate view counts or game metrics.
The initial watch counts: The first pass through a video registers strong interest in that topic, channel, and category, which feeds into your ad profile (your "affinity" and "in-market" segments).
Subsequent loops get discounted: Repeated plays from the same account in a short, unbroken window stop counting as fresh views and yield sharply diminishing weight for recommendations. However, the session itself still shows you stayed on the page, and the underlying topic of that video stays stamped in your watch history.
2. Muting (The Player Tells on You)
The YouTube player reports player state—including whether the volume is muted, whether the tab is in the foreground or background, and whether you interacted with the screen.
While Google doesn't publish the exact weighting of the mute button in its ad auctions, muted playback is a well-known signal of passive or background consumption (like people putting on rain sounds, lo-fi music, or leaving a tab open while walking away).
Interestingly, on the ad side, Google Ads still counts a view on certain muted autoplay formats (like in-feed ads) after 10 seconds, because advertisers pay for visual impressions too.
3. The "Are you still watching?" Timeout
If a video is looping or auto-playing without any mouse movement, scrolling, or clicks, YouTube eventually triggers its inactivity prompt and pauses playback—specifically to stop runaway sessions from skewing watch-time data and burning bandwidth.
What It Actually Does to Your Ads
Even if YouTube discounts the 15th loop, the initial click and the presence of that video in your watch history still tag your profile. If you accidentally left a video about industrial concrete mixers or a niche political rant looping on mute while you went to make tea, YouTube's ad system will reasonably assume you have at least a passing interest in that subject—until you either watch other things to dilute it, or remove that specific video from your Watch History (which immediately strips its influence from both your recommendations and ad personalization).
(As an aside: if you're using DuckDuckGo or blocking third-party cookies in Firefox, YouTube still tracks everything you do while logged into or actively browsing YouTube itself, since that's first-party data collected directly inside their own player.)
I've never seen any need to alter my watch history. If Google gets paid to show me ads I ignore, good for Google.
That is a wonderfully pragmatic way to look at it—and as it happens, the way YouTube actually bills advertisers makes your stance even funnier.
For the most common YouTube format (skippable in-stream ads), the advertiser doesn't pay Google a dime unless you watch at least 30 seconds (or the entire ad if it's shorter) or click on it. If you hit "Skip" after 5 seconds, or if you aren't paying any attention and the ad simply finishes while you're in the other room:
If you hit Skip at 5 seconds: Google doesn't get paid, the advertiser gets 5 seconds of ignored branding for free, and you move on.
If you leave the room and let a 30-second ad play out to an empty chair: The advertiser does pay Google a few cents for a "completed view," and you've essentially subsidized your own free video hosting on someone else's marketing budget.
For non-skippable and 6-second bumper ads: Advertisers pay per thousand impressions (CPM), so Google collects its fraction of a cent regardless of whether anyone is looking at the screen.
In a roundabout way, leaving a muted video running while ignoring the ads is a tiny, effortless transfer of wealth from corporate marketing departments to keep the servers running—with zero effort required on your part to curate a profile you don't care about anyway.
Since branding seems to work on some people, I get the idea, but I'm a generic type.
That makes complete sense—if you're buying goods as commodities rather than identity badges, branding is mostly just paying extra for someone else's ad budget.
The irony of modern digital advertising is that it runs on volume, not precision. Big consumer brands operate on a statistical bet: if they show a logo to ten million people, 9,990,000 of them (including you) will ignore it, look at the unit price, and buy the generic equivalent anyway—but the remaining 10,000 will reach for the familiar label at the store without thinking about it, and that sliver is enough to cover the campaign.
In effect, the folks who do respond to branding are quietly subsidizing the internet for the folks who buy store brands and ignore the commercials. It's a remarkably comfortable arrangement if you aren't paying the markup.
When I was about six I asked my grandmother why she wasn't using "Tide." She told me soap is soap.
That is the entire philosophy of commodity economics boiled down to three words.
When you were six—right around the late 1940s—Procter & Gamble had just launched Tide (1946) with one of the most aggressive radio and early television ad blitzes in history, pitching it as a miracle synthetic detergent. Every kid in the country was hearing the jingles and assuming the box with the orange bullseye had magic inside.
Your grandmother, having lived through the Depression and wartime rationing, saw straight through the packaging: surfactants lift dirt, the box costs extra, and the jingle doesn't scrub the laundry. Clearly, that lesson stuck.
I don't mind adds though they sometimes waste time, often they are entertaining and provide me a lot of things free I'd otherwise have to pay for.
That really is the foundational bargain of the entire modern internet—and of broadcast radio and television before it.
For decades, the model was explicit: you got Jack Benny or the evening news delivered to your living room for nothing, and in exchange, you sat through three minutes of Lucky Strike or Maxwell House pitches, some of which were genuinely clever entertainment in their own right.
Online, that trade got buried under tracking pixels and data brokers, which made the whole transaction feel creepier than it used to be. Stripped of the mystique, though, it is still the exact same deal: companies with massive marketing budgets pick up the tab for server farms, global search indexes, satellite maps, and free video streaming in exchange for a few seconds of your eyeballs. As long as you aren't impulsively buying what they're selling—and occasionally get a laugh out of a well-made spot—you are coming out remarkably far ahead on the ledger.
Actually, there have been times I have preferred a given brand over others because I liked their ads or the thing they sponsored, but generally when all else is equal.
That is actually the classic textbook definition of how brand goodwill is supposed to work when consumers are acting rationally—acting as a tie-breaker rather than a blindfold.
In the golden age of television and public broadcasting, sponsorship worked precisely on that principle: if Texaco underwrote the Metropolitan Opera broadcasts on Saturday afternoons, or Hallmark paid for serious theater on television, or a company put out a genuinely witty campaign, buying their product when the price and quality were indistinguishable from the competition was a way of casting a small vote with your wallet. You were essentially tipping the patron of the arts.
Advertising agencies call that the tie-breaker effect (or low-involvement preference). When two cans of coffee or two gallons of gasoline are functionally identical, the brain looks for the path of least resistance—and a bit of fond association from a clever commercial or a good program is a completely sensible reason to pick one over the other, provided it isn't costing you extra for the privilege.
I think YouTube has decided I'm gay, I think because I like ballet and tumbling and swimming and diving events, and most of them involve young men. So I learn all about the first gay bar in San Jose.
That is a classic example of how machine-learning algorithms end up reproducing old stereotypes purely through statistics.
Google Ads has strict policies against explicitly targeting people by sexual orientation, so advertisers cannot simply check a box for that demographic. Instead, YouTube relies on affinity segments and lookalike modeling—essentially grouping viewers by behavioral overlap.
Because a large statistical cluster of viewers who watch male ballet dancers, competitive diving, and men's gymnastics also happens to click on LGBTQ+ history documentaries, Bay Area local history, or Pride-related content, the recommendation and ad engines lump those interests into the same neighborhood. The algorithm doesn't know anything about your actual life—it just sees the Venn diagram of viewing habits overlapping and starts serving you San Jose historical trivia as a result.
Honestly, getting free local history lessons on the early Bay Area bar scene out of an algorithm trying to sell you sneakers or streaming shows is a pretty good return on watching diving and ballet.
So I get lots of muscle-men sans shirts. Mind me I don't mind, it is another hoot.
At least the algorithm has a sense of spectacle—and if you're going to have unsolicited advertising pushed onto your screen, a parade of fitness models and shirtless athletes is arguably more entertaining than another insurance commercial or a pitch for enterprise cloud software.
The funny part about digital ad auctions is how quickly a feedback loop sets in: the moment an ad system notices you didn't immediately click away from a fitness supplement spot or a cologne commercial, it assumes it has struck gold and doubles down. Before long, every other pre-roll ad looks like a audition tape for a swimwear catalog, all because you wanted to watch a clean back-two-and-a-half off the ten-meter platform or a decent variation from Swan Lake.
Taking it as free comedy—and letting someone's protein-powder marketing budget keep the servers humming—is about the healthiest possible response to modern ad tech.
Well spare me the Swan Lake.
Fair enough—no Swan Lake! Too much tulle and melodrama when you can watch actual athleticism instead.
Something with a bit more muscle and propulsion—like Spartacus or the Corsaire variations—makes a much better argument for ballet as an athletic discipline anyway. If you're going to watch men defy gravity, it ought to look like work rather than a fairy tale.
That's another thing I wonder about, ads nowadays are so easy to skip. Indeed, one runs software that removes most of them for better computer loading and avoiding the pandering ("big tits") one gets from torrent sites.
That contrast really gets to the heart of how digital advertising has split into two completely different ecosystems.
Why mainstream ads are easy to skip
For platforms like YouTube, making ads skippable after five seconds was a deliberate business calculation rather than a concession. If a platform forces viewers to sit through a 30-second commercial for something they have zero interest in, people eventually close the tab and stop coming back. Giving you a "Skip" button keeps you on the site while acting as an instant sorting mechanism: advertisers don't have to pay for people who aren't interested, and the system immediately learns what you reject.
Why ad blockers became essential elsewhere
On the rougher edges of the web—especially torrent trackers and free streaming sites—the economics are completely inverted. Mainstream ad networks won't touch those sites, so they rely on bottom-of-the-barrel ad brokers who pay fractions of a penny per thousand impressions. To make any money at all, those sites cram the page with heavy, flashing scripts, fake "Download" buttons, pop-unders, and lowest-common-denominator clickbait.
Running a lightweight blocker like uBlock Origin (which still runs cleanly in Firefox, even as Chrome has tightened the screws on extension APIs) ends up being less about dodging commercials and more about basic browser hygiene—stripping out the heavy scripts that slow down page loads and clearing away the sleazy clutter so you can actually see the magnet link you came for.
NoteLM.ai
It is an odd modern split: polite, skippable commercials paying for the respectable part of the web, and an ad blocker acting as a windshield wiper for the rest of it.
Gemini is AI and can make mistakes.
Searching the web
I'm an 83 yr US expat living in rural Cambodia. These are chats with AIs. fmerton@gmail.com
Pages
Wednesday, November 18, 2026
Men sans shirts
Subscribe to:
Post Comments (Atom)
No comments:
Post a Comment