The shirt prints four words in white capitals: I HAD POTENTIAL. Beneath them, a white curved ramp descends from an upper-left ledge to a lower-right flat ledge. A dashed empty circle sits at the top of the ramp. A solid white ball rests at the bottom. That is the whole design. No arrows, no labels, no equations. The joke depends on the reader supplying the missing physics.

I want to argue that the drawing is doing more work than the caption, and that the caption is doing something different from what it first appears to do. The phrase I had potential is a confession of wasted capacity. The diagram, read literally, is a picture of capacity being converted. Those two readings do not quite agree, and the gap between them is the interesting part.

What the drawing actually shows

Start with the geometry, because the geometry is precise. The ramp is curved, not a straight diagonal. That matters. A straight incline with a ball at the bottom would suggest sliding friction and a final resting place. A curved ramp suggests a track, a path with a shape, something a ball follows rather than something it merely sits on. The dashed circle at the upper left is the ball’s starting position, drawn as an outline because it is no longer there. The solid ball at the lower right is where it ended up.

Read as a physics diagram, this is a potential energy sketch. An object at the top of a ramp has gravitational potential energy relative to the bottom. Released, it converts that potential into kinetic energy, rolls down, and comes to rest on the flat. The dashed circle is the before; the solid ball is the after. The design is not showing a ball that failed to move. It is showing a ball that moved and stopped.

That is a different claim from the caption. I had potential implies the potential was never realized. The diagram implies the potential was realized and then spent. The ball did not stay at the top. It went down the ramp. The flat ledge at the bottom is not a failure state; it is the end state of a system that worked exactly as physics intended.

The double meaning the design is trading on

In physics, potential is not a compliment. It is a quantity: the energy an object has by virtue of its position in a field. A ball at the top of a ramp has more potential energy than a ball at the bottom. That is all the word means in this context. It says nothing about whether the ball is talented, promising, or likely to succeed. It says something about where the ball is sitting.

In ordinary speech, potential is a compliment and a burden. To say someone has potential is to say they are not yet what they could be. It is a statement about the future, delivered in the present, usually by someone with a stake in the outcome. The shirt takes that word and puts it next to a diagram where potential is a measurable quantity that gets converted and then gone. The joke is that the physics reading is more honest than the motivational one. The ball had potential. The ball used it. Now the ball is on the flat, and the flat is fine.

I do not think the design is claiming that rest is failure. If anything, the solid ball on the lower ledge looks settled. The dashed circle at the top looks like a position that has been vacated. The visual argument is that potential is a starting condition, not a permanent identity. You can have it, spend it, and end up somewhere lower and flatter, and the diagram will still be accurate.

Why the minimalist execution matters

The shirt is white-on-white in the artwork description: white capitals, white ramp, white ball, presumably on a darker garment ground. That is a deliberate constraint. A physics diagram with axis labels, arrows, and a legend would be a different object. It would explain itself. This one does not. It gives you the ramp and the ball and lets you supply the rest. The dashed circle is the only concession to process; everything else is a still frame.

That restraint is what makes the shirt wearable as a joke rather than a lecture. The caption is four words. The drawing is two shapes and a curve. The reader who knows what potential energy is gets the double reading immediately. The reader who does not still gets a legible image: something rolled down a hill and stopped. Both readings work, and neither requires a caption explaining the caption.

This is the difference between a slogan and a usable promise. A slogan tells you what to feel. A diagram shows you what happened. The shirt is closer to the second thing, which is why it does not read as motivational. It reads as a post-mortem.

The work-culture reading, held lightly

I am not going to claim this shirt is about performance reviews or quarterly planning, because the artwork does not say that. But I will note that the structure of the joke is familiar to anyone who has sat in a meeting where potential was used as a forward-looking word and results was used as a backward-looking one. The diagram collapses that distinction. It shows potential as a quantity that gets spent, not a quality that gets rewarded. The ball does not get credit for having been at the top. It gets a solid outline at the bottom.

That is a useful mental model for anyone who works inside systems that measure outcomes rather than inputs. The ramp is the input. The flat is the output. The dashed circle is the version of you that had not yet been measured. The shirt is not cynical about this. It is just literal. Potential is a position, and positions change.

The object itself

The shirt is sold as the I Had Potential Funny Physics Pun Sarcastic Minimalist Quote under the Quirky Cotton Collective brand. The design is the ramp-and-ball diagram described above, printed in white on the garment. I have not worn it, tested it, or seen it on a body; my reading is based on the retail artwork as observed. The design is the argument, and the argument holds up on inspection: a curved ramp, a vacated starting position, a settled ball, and four words that mean something different in physics than they do in a performance review.

Disclosure: This publication and Quirky Cotton Collective share an owner. This is not an independent review.

The native garment preview shows a dark shirt with very small, centered lowercase meh. lettering. No animal illustration is present. That is the entire design: four letters and a period, set at a scale that almost dares you to lean in. The shirt is listed as Meh Funny Sarcastic Introvert Minimalist Saying by Quirky Cotton Collective.

Editorial note: This publication and Quirky Cotton Collective share an owner.

Most slogan shirts fail because they overcommit. They shout a punchline across the chest, and the wearer becomes a billboard for a joke that was funny for about three seconds. This one does the opposite. The scale is the argument. At that size, the word is not a broadcast; it is a private signal for anyone close enough to read it. The period is what makes it work. Without punctuation, meh is a sound—a shrug rendered as text. With the period, it becomes a complete sentence, a verdict delivered and closed. The period is the difference between a noise and a decision.

That distinction matters more than it might seem for a shirt marketed with words like sarcastic and introvert. Those labels describe a posture, not a personality. The design itself is closer to a boundary marker than a joke. It says: I have registered the situation, I have no further comment, and I am not going to pretend otherwise. The small type enforces the same rule. You have to be paying attention to receive the message, which is exactly the kind of attention the wearer is signaling they are not offering in return.

There is a useful contrast here with the other end of the slogan-shirt spectrum. A shirt that says Let’s Root For Each Other makes a promise and quietly demands reciprocity. A shirt that says meh. makes no promise at all. It is a closed loop. The period is the closure. That is why the design reads as minimalist rather than merely small: every element is doing work, and the work is refusal. The word declines to escalate. The period declines to continue. The scale declines to explain itself.

For technical readers, the appeal is probably not the word itself but the discipline of the form. It is a terse message with no ambiguity about its scope. It does not claim to be a system, a diagnosis, or a solution. It is a status code rendered in fabric: acknowledged, no action taken. The shirt does not cure anything, improve anything, or demonstrate anything about the wearer’s profession. It is a piece of clothing with a small, complete sentence on it. That is enough.

The design also avoids the trap of the animal variant or the pride variant—the visual pun that turns a mood into a mascot. There is no illustration to soften the word or give it a second meaning. The absence is deliberate. A drawing would make meh. cute. The period would not survive the cuteness. What remains is a flat, lowercase verdict on a dark ground, and the only thing it asks of you is that you read it at the distance it was designed for.

That is the argument of the shirt, and it is a coherent one: understatement is not the same as indecision. A very small word with a period is still a complete thought. The wearer is not shrugging. The wearer has shrugged, past tense, and moved on.

The shirt reads NOT FRIENDLY across the top, DO NOT TOUCH beneath it, and between the two lines sits a yellow caution triangle containing a crossed-out hand. The lettering is set in the visual register of industrial signage: condensed capitals, high contrast, the implicit promise that what follows is a hazard classification rather than a preference. The triangle is the part worth staring at. A caution triangle is a standardized container for a hazard—it tells a reader that a specific danger has been identified, located, and bounded. The crossed-out hand inside it is doing something else entirely. It is not naming a hazard. It is issuing a prohibition. The sign has borrowed the authority of a safety placard to deliver a personal boundary, and the seam between those two functions is where the design gets interesting.

Standard warning architecture separates two jobs. A hazard sign identifies a condition: hot surface, high voltage, slippery floor. A prohibition sign forbids an action: do not enter, no smoking, do not touch. The two are related but not identical, and the difference matters because they assign responsibility differently. A hazard sign says the environment is dangerous; act accordingly. A prohibition sign says your behavior is the problem; stop it. The shirt collapses both into one graphic. The triangle frames a hazard, but the hazard it frames is the reader’s hand. The wearer is not warning you about the floor. The wearer is warning you about you.

That inversion is the joke, and it is also the argument. The phrase NOT FRIENDLY is not a standard sign legend. You will not find it in ISO 7010 or ANSI Z535, the two standards that govern the shape, color, and symbol vocabulary of safety signage. Those standards specify warning, caution, danger, and a fixed set of prohibition symbols, but they do not include a category for the person wearing this garment is disinclined to chat. The shirt is therefore not a real warning sign. It is a sign-shaped object that has borrowed the semiotics of one—the triangle, the capitals, the imperative mood—to make a claim that no safety standard would certify. The borrowed form is the whole point. A plain sentence reading I am not friendly, please do not touch me would be a request. The sign format converts it into something closer to a regulation.

This is where the design earns its keep as an object of study rather than a novelty. The caution triangle is a container. It tells the reader that whatever is inside has been classified, that someone has done the work of identifying the hazard and deciding how to represent it. By placing a crossed-out hand inside that container, the shirt implies that the boundary has already been assessed and codified. The wearer is not improvising a preference in the moment. The wearer is citing a rule. That is a different social move. A preference can be negotiated; a posted rule is harder to argue with, because arguing with it means arguing with the sign itself, and the sign is wearing a triangle.

The crossed-out hand is the most literal element and the one that does the most work. A hand is the universal symbol of approach—the reaching gesture, the tap on the shoulder, the handshake, the friendly touch on the arm. Crossing it out does not say do not touch this object in the way a museum sign says it about a painting. It says do not touch this person, and it says it in the grammar of a facility rule. The hand is also the reader’s hand, not the wearer’s. The sign is addressed outward. It is a boundary that has been pre-printed, which means the wearer does not have to say it out loud, and the reader does not have to be told in conversation. The shirt absorbs the awkwardness of the exchange and relocates it into a graphic.

There is a real design tension in the phrase NOT FRIENDLY. It is a negative definition—it tells you what the wearer is not, rather than what they are. Negative definitions are cheap to print and hard to falsify. Friendly is a social performance that varies by context, energy, and audience; not friendly is a stable claim that requires no follow-through. The shirt is therefore making a promise it can keep. That is a different kind of reliability than a slogan like be kind or good vibes only, which commits the wearer to a behavior they may not sustain. The warning sign commits the wearer to nothing except the boundary it prints. It is, in that narrow sense, an honest garment.

The introvert framing in the listing title is doing separate work from the artwork. The graphic itself does not say introvert; it says not friendly, do not touch. The introvert label is an interpretation layered on top, and it softens the sign considerably. Not friendly is a blunt refusal; introvert is a personality category with a sympathetic gloss. The same graphic reads differently depending on which frame the reader brings. On a stranger in a crowd, the sign is a wall. On a colleague at a team offsite, the same sign is a joke about social battery. The artwork is fixed; the meaning is not. That is not a flaw in the design. It is the reason the design travels.

For technical readers, the interesting parallel is not to security or to software but to the general problem of writing a boundary that is legible without being hostile. A warning sign works because it is standardized: the reader already knows what a yellow triangle means before they read the words. The shirt borrows that prior knowledge and redirects it. The result is a message that is terse, unambiguous, and slightly absurd, because the form is doing something the form was not designed to do. That is the same move as a well-placed comment in a codebase or a terse error message: it uses a familiar container to deliver a specific instruction, and it trusts the reader to recognize the container before they parse the content. The shirt is a small exercise in that skill, printed on cotton.

The Not Friendly Do Not Touch shirt is a warning sign that has been repurposed into a personal boundary, and the repurposing is visible in every element: the triangle that frames a hazard where no hazard exists, the crossed-out hand that forbids an action rather than naming a condition, the capitals that borrow authority from a standard the shirt does not follow. The design works because it does not resolve that tension. It lets the sign be a sign and the joke be a joke at the same time, and it leaves the reader to decide which frame applies. That is a more useful object than a slogan, because a slogan tells you what to feel and a warning sign tells you what to do. This one tells you to keep your hands to yourself, in the most official-looking way a t-shirt can manage.

Editorial note: This publication and Quirky Cotton Collective share an owner.

Where the TCF Consent String Actually Lives in a Bid Request

A bid request arrives with regs.gdpr set to 1. The bidder reads the consent string, finds nothing, and bids without personalization — or doesn’t bid at all. Nothing in the payload is malformed. The JSON validates. The auction runs.

That is the failure mode worth understanding. The Transparency and Consent Framework (TCF) consent string is not hard to parse once you have it. The hard part is that the same logical signal occupies two different JSON locations depending on which OpenRTB version produced the request, and a bidder wired to only one path will silently miss consent on the other version’s traffic. This article traces the field from the CMP that writes it to the bidder that decodes it, and treats the 2.5/2.6 path split as the operational hazard it is.

What the string is, and who is allowed to write it

The TC String is defined by IAB Europe’s TCF specification, with IAB Tech Lab stewarding the technical documents. Its purpose is narrow and specific: it encapsulates what was disclosed to a user and what that user chose, encoded as a compact HTTP-transferable string. As the TC String format specification puts it, the string “enables communication of transparency and consent information to entities, or ‘vendors’, that process a user’s personal data.”

The authorship rule matters for anyone debugging a missing or stale string. The same specification states:

A Transparency & Consent String may only be created by an IAB Europe TCF registered CMP using its assigned CMP ID number in accordance with the Policies. Vendors or any other third-party service providers must neither create nor alter TC Strings.

So when a consent string is absent from a bid request, the question is not “can we synthesize one?” — the answer is no. The question is where in the chain between the CMP and the bidder’s decoder the value was dropped, or whether it was ever generated for that user.

The specification also describes what the string is for at the receiving end: “Vendors decode a TC String to determine whether they have the necessary legal bases to process a user’s personal data for their purposes.” That is the whole economic function of the field. It is a legal-basis lookup key, not a targeting signal in its own right.

The two paths: user.ext.consent and user.consent

GDPR support arrived in OpenRTB as a pair of extension fields defined by the IAB GDPR advisory: regs.ext.gdpr and user.ext.consent. OpenRTB 2.6 promoted both to first-class attributes, and the change log records each as “previously ext,” according to the privacy-signal reference at rtblint.org.

In 2.6, then:

• regs.gdpr is an integer: 0 means the request is not subject to GDPR, 1 means it is, and omission means unknown.
• user.consent is a string holding the TCF consent string when GDPR applies.

Two details in that layout trip people up. First, the consent string is a string on user — not an object, and not under regs. The regs object says whether GDPR applies; user carries what the user chose. Second, the pre-2.6 locations are still alive in production traffic. Old SDKs and hand-rolled adapters keep emitting regs.ext.gdpr and user.ext.consent, and a bidder reading only the first-class fields sees no signal at all on those requests.

The rtblint reference is explicit about the consequence of a half-populated request: “A request with regs.gdpr: 1 and no user.consent tells bidders consent is unresolved; most will bid without personalization or not at all.” That sentence is hedged — “most” — and should be read that way. The point is not a precise no-bid rate. The point is that the request has communicated an unresolved state, and the bidder’s response is a business decision made under uncertainty.

The path split is handled in code, not by convention

If you want evidence that this is a real engineering problem rather than a documentation nicety, look at how an open-source exchange handles it. Prebid Server’s ConvertDownTo25 function in openrtb_ext/convert_down.go exists to translate a 2.6 request into a 2.5-shaped one before passing it to bidders that expect the older schema. The consent handling is a dedicated function:

// moveConsentFrom26To25 modifies the request to move the OpenRTB 2.6 GDPR consent
// field (req.user.consent) to the OpenRTB 2.5 location (req.user.ext.consent). If
// the OpenRTB 2.5 location is already present it may be overwritten.
// The OpenRTB 2.5 location is expected to be empty.
func moveConsentFrom26To25(r *RequestWrapper) error {
	if r.User == nil || len(r.User.Consent) == 0 {
		return nil
	}
	consent26 := r.User.Consent
	r.User.Consent = ""
	userExt, err := r.GetUserExt()
	if err != nil {
		return err
	}
	userExt.SetConsent(&consent26)
	return nil
}

Three things are worth pulling out of that snippet. It reads the 2.6 location, clears it, and writes the value to the 2.5 extension location. The comment notes that an existing 2.5 value “may be overwritten” and that the 2.5 location “is expected to be empty” — an assumption, not an enforced invariant. And the function returns early when r.User.Consent is empty, which means a request that arrived with consent only at user.ext.consent passes through untouched. The down-conversion is one-directional.

The same file’s clear26Fields function zeroes user.Consent as part of stripping 2.6-only fields, consistent with the comment that “the previous OpenRTB 2.5 spec did not specify that bidders must tolerate new or unexpected fields.” That is the design rationale for the whole conversion layer: a 2.5-era bidder may reject or ignore fields it does not recognize, so the adapter must move the signal to where that bidder looks.

This is one implementation, and I am not claiming it describes every exchange. What it does demonstrate is that the 2.5/2.6 consent path difference is significant enough that a major open-source exchange wrote explicit translation code for it. If your integration assumes a single canonical path, you are assuming something the code does not.

What the bidder does once it has the string

Locating the string is step one. Decoding it correctly is step two, and the TCF Implementation Guidelines are specific about evaluation order. For each purpose, vendors must “Evaluate publisher restrictions prior to any other signal.” Only after publisher restrictions are resolved does the vendor apply its own default legal basis or the publisher-defined one.

That ordering has a practical implication for debugging. A bidder that suppresses an impression may be doing so because of a publisher restriction encoded in the string, not because the user withheld consent for that vendor. If your logs record only “consent present / absent,” you cannot distinguish the two, and you will misattribute the loss.

The guidelines also address where to read the string on the server side: “If an impression is received server side (through openRTB for example), you should read the information from the TC data payload.” For client-side paths — redirects, Prebid in the browser — the CMP API is the source, and the guidelines recommend the event listener rather than a one-shot read so you get the most current TC data object.

What I am not claiming

Several things commonly asserted about consent strings are not supported by the sources I checked, and I am leaving them out rather than dressing them up.

I am not quantifying the revenue or bid-rate impact of a missing consent string. No retrieved source provides a figure, and inventing one would be worse than saying so.

I am not describing the bit-level layout of the TCF v2.2 core string. The format specification was retrieved only as a partial excerpt covering scope, authorship, and purpose; it does not support field-level encoding claims here. If you need the bit layout, read the specification directly.

I am not asserting how any named exchange or ad server handles user.ext.consent in production. The relevant vendor documentation did not retrieve, and the Prebid Server code is evidence about Prebid Server, not about the rest of the market.

An operational checklist

The argument of this article is that consent-string handling is a transport problem before it is a legal-logic problem, and the two have different owners and different fixes. A transport bug — the value sitting at a path nobody reads — is fixed in an adapter. A legal-basis bug — the value read correctly but evaluated in the wrong order — is fixed in bidding logic. Conflating them wastes time in both directions.

What follows is my recommendation for instrumenting the transport layer, not a finding from any source:

1. Log which consent path was populated. Record, per request, whether user.consent, user.ext.consent, both, or neither carried a value. This single field in your logs surfaces adapter drift before it reaches bidding logic.
2. Flag the unresolved state explicitly. A request with regs.gdpr equal to 1 and no consent string at either path is a distinct condition. Count it separately from “GDPR not applicable” and from “consent present.”
3. Audit gating logic, not just schema validity. A structurally valid request can still carry a consent string that does not authorize the vendor or purpose in question. Validation of JSON shape will not catch that; only decoding and evaluating will.
4. Watch for the overwrite assumption. If your pipeline converts between versions, check what happens when both locations are populated. The Prebid Server comment says the 2.5 location “is expected to be empty” — verify that expectation holds in your traffic rather than trusting it.

The consent string is a small field with a large blast radius. It is worth knowing exactly where it sits in every version of the payload you accept, and worth logging when it isn’t there.

Questions readers actually ask

Is user.ext.consent still valid in OpenRTB 2.6?

The 2.6 specification promotes the field to user.consent. Whether an old-path value is honored depends on the receiving bidder’s implementation. The rtblint reference flags sending the pre-2.6 paths in a 2.6 payload as a common mistake, because “bidders reading only the first-class fields see no signal at all.”

Does regs.gdpr: 1 mean consent was given?

No. regs.gdpr indicates whether the request is subject to GDPR. The user’s choices live in the consent string on user. A request can be in scope for GDPR and carry no consent string at all, which is the unresolved state described above.

Can a DSP generate a consent string if one is missing?

Per the TCF format specification, only a registered CMP may create a TC String, and vendors “must neither create nor alter TC Strings.” A missing string is a signal to be handled, not a gap to be filled.

Where does the consent string go in a pre-2.6 request?

In the extension field user.ext.consent, paired with regs.ext.gdpr. OpenRTB 2.6 promoted both to first-class attributes.

What should a bidder do with a malformed string?

The retrieved sources do not prescribe a single handling rule for malformed strings, and I am not going to invent one. What the Implementation Guidelines do establish is the evaluation order for a valid string: publisher restrictions first, then purpose and vendor legal-basis signals. A string that cannot be decoded cannot be evaluated in that order, which is a condition your logging should distinguish from a valid string that simply withholds consent.

The shirt reads I WEAR THIS SHIRT PERIODICALLY, and beneath that sentence sit four tiles in the visual grammar of the periodic table: 53 I, 74 W, 22 Ti, 16 S. Read the symbols in order and you get I W Ti S. Read the atomic numbers in order and you get 53, 74, 22, 16. Neither sequence spells the sentence above it. That gap is the joke, and it is also the most interesting thing about the design.

The pun works because periodically has two jobs. In ordinary English it means from time to time. In chemistry it means according to the periodic table. The shirt claims the first meaning and then illustrates the second, and the illustration does not quite cooperate. That is not a flaw in the artwork. It is the mechanism.

What the tiles actually say

Each tile carries an atomic number and a one- or two-letter symbol. The numbers are real: 53 is iodine, 74 is tungsten, 22 is titanium, 16 is sulfur. The symbols are real too: I, W, Ti, S. What the tiles do not do is concatenate into the word periodically. They concatenate into IW Ti S, which is not a word, not an acronym, and not a chemical formula. It is a near-miss.

That near-miss is the whole design. A literal periodic-table spelling of periodically would require element symbols that do not exist in that sequence. The designer chose four tiles that gesture at the sentence without reproducing it, and the result is funnier than a clean match would be. The shirt is not saying here is the word in elements. It is saying here is what happens when you try.

I read the tiles as a visual pun on the limits of the periodic table as a writing system, not as a claim about chemistry. The atomic numbers and symbols are correct for the elements they name. The sentence above them is correct English. The mismatch between the two is the point.

Why the mismatch is the joke

Most element-spelling jokes work by finding a word that can be spelled with valid symbols. BaCoN, CaNdY, SPaM. The pleasure is in the fit. This shirt inverts that pleasure. It sets up the expectation of a fit and then delivers a near-fit, and the near-fit is more interesting because it forces you to notice the rules.

To read the shirt correctly you have to do three things at once: parse the sentence, parse the tiles, and compare them. The comparison fails, and the failure is the punchline. That is a more demanding joke than a clean spelling, and it rewards the reader who actually looks at the tiles instead of glancing at the sentence and moving on.

There is also a quieter joke in the word periodically itself. The sentence says the wearer puts the shirt on from time to time. The tiles say the shirt is organized by the periodic table. The two meanings sit on top of each other, and neither cancels the other. The shirt is periodic in both senses, and the tiles are the evidence for the second sense even though they fail to spell the first.

The difference between a slogan and a usable promise

Technical readers spend a lot of time with messages that are supposed to do work: error strings, status codes, boundary warnings, configuration comments. A good one tells you what happened and what to do next. A bad one tells you something true but useless. The shirt is not a status message, but it is a useful example of the difference.

I wear this shirt periodically is a slogan. It is true, it is mildly funny, and it does not ask you to do anything. The tiles are the supporting graphic, and they do not resolve into a clean instruction either. That is fine for a shirt. It would be a problem in a log line. The design works precisely because it is allowed to be a near-miss, and near-misses are usually not allowed in the systems these readers operate.

I find that contrast useful. The shirt is a reminder that ambiguity can be a feature when the stakes are low and a bug when the stakes are high. The tiles are ambiguous on purpose. A status code that behaved the same way would be a incident.

What the shirt is and is not

The shirt is a standard t-shirt sold under the Quirky Cotton Collective brand, listed as I Wear This Shirt Periodically Funny Chemistry Science Pun. The design is the sentence plus the four tiles described above. That is the object. It is not a chemistry lesson, a software tool, or a security control, and nothing about the artwork demonstrates a technical result.

The phrase has measurable search interest: Google Keyword Planner reports a monthly range of 1,000 to 10,000 for the exact query i wear this shirt periodically in the US over the period September 2025 through August 2026. That number describes search volume, not sales, and it does not tell you whether the shirt is good. It tells you people look for the phrase. The design is what they find.

Editorial note: this publication and Quirky Cotton Collective share an owner, so this is not an independent review.

The takeaway

Read the tiles before you read the sentence. The tiles are where the design does its work. They are correct as elements and wrong as spelling, and that combination is the joke. A shirt that spelled periodically cleanly in element symbols would be clever. This one is cleverer because it shows you the attempt and lets you see where it stops.