peopleworks · peopleworks · Aug 5, 2026 · Aug 5, 2026
diff --git a/Docs/Blog/shorts/GUION-shorts.md b/Docs/Blog/shorts/GUION-shorts.md
@@ -16,11 +16,16 @@ Paleta SignsOfAI: `--bg:#0b1020` · léxico `#db2777` · retórico `#f59e0b` ·
 | 5 | Dos idiomas | ✅ | ✅ | ✅ 25 s · 26 s |
 | 6 | Reescritura | ✅ | ✅ | ✅ 25 s · 23 s |
 | 7 | Escritorio | ✅ | ✅ | ✅ 25 s · 23 s |
-| 8 | Artefactos | — | guion abajo | pendiente |
-| 9 | Bibliografía inventada | — | guion abajo | pendiente |
-| 10 | 61% / línea base | — | guion abajo | pendiente |
+| 8 | Artefactos | ✅ | ✅ | ✅ 34,5 s · 30,0 s |
+| 9 | Bibliografía inventada | ✅ | ✅ | ✅ 36,3 s · 35,5 s |
+| 10 | 61% / línea base | ✅ | ✅ | ✅ 40,2 s · 37,4 s |
+| 11 | El veredicto que nunca se dio | ✅ | ✅ | ✅ 35,5 s · 32,4 s |
 
-Los renders están en `out/<id>.mp4` + `.srt`. Los 8–10 necesitan HTML antes de narrar.
+Los renders están en `out/<id>.mp4` + `.srt`.
+
+**Los 8–11 duran más que los 1–7** (29–40 s frente a 23–30). Siguen dentro del límite de
+Shorts, pero el 10 es un tercio más largo que nada publicado antes: si la retención cae, el guion
+es lo que hay que recortar, no la animación.
 
 **Orden del pipeline:** narrar primero (`narrate-all.mjs`), sacar los tiempos de cue
 (`cue-times.mjs`) y **después** escribir el HTML alrededor de la duración medida, porque los
@@ -100,6 +105,23 @@ y una marca cayendo dentro, en verde.
 - **EN:** "Sixty-one percent. That is the share of essays by students writing in their second language that AI detectors flag as machine-written. Not of the cheats: of the essays. Because formal, careful writing looks like a machine, and that is how you write in a language you learned second. You do not fix that with a better detector. You fix it with a different question: not does this look like a machine, but does this look like the person who wrote the others. And if you write formally, your own baseline is already formal."
 
 
+
+## Short 11 — "Un `if` que siempre era falso" (el veredicto que nunca se dio)
+**Visual:** un documento arriba que se bifurca en dos tarjetas — igual que la portada del séptimo
+artículo, para que quien vio una reconozca la otra. Izquierda: el contador sube a `90/100` y aparece
+una barra ámbar sólida (el veredicto dicho). Derecha: el mismo `90/100` y una barra punteada
+**tachada** (el veredicto ausente). Entra el `≠` entre las dos barras, no entre los números —
+los números coinciden, ese es el chiste. Corte a la línea de código con
+`RecommendedThreshold is { } threshold`, y `null` cayendo sobre ella en rojo, dos veces: `en`, `es`.
+Cierre: `340 pruebas` y ninguna mirando ahí; luego la frase nueva, «Sin señales por encima del
+umbral medido», apareciendo bajo las dos tarjetas ya iguales.
+- **ES:** "Mi detector le puso noventa sobre cien a este texto y dijo: señales fuertes de escritura con inteligencia artificial. El informe que un profesor imprime y lleva a un comité, del mismo texto y en la misma ejecución, no dijo nada. Y no era un caso raro: no había emitido un veredicto nunca, en ningún idioma. Una condición pedía un umbral que ningún idioma tenía. Yo tenía trescientas cuarenta pruebas. Ninguna comparaba las dos caras entre sí."
+- **EN:** "My detector scored this text ninety out of a hundred and called it strong signs of A I writing. The report a teacher prints and carries to a committee, same text, same run, said nothing at all. And that was not a rare case: it had never given a verdict, not once, in any language. One condition asked for a threshold that no language had. I had three hundred and forty tests. Not one of them compared the two faces."
+
+**Por qué este short y no otro:** es el único de la serie donde la herramienta queda mal, y por eso
+funciona. Los diez anteriores explican algo que hace bien; este cuenta que se contradecía a sí misma
+y que lo publicamos igual.
+
 ---
 
 ## Build (una vez existan los HTML de cada short)

diff --git a/Docs/Blog/shorts/en/short10-linea-base.html b/Docs/Blog/shorts/en/short10-linea-base.html
@@ -0,0 +1,65 @@
+<!DOCTYPE html><html lang="en"><head><meta charset="utf-8"><style>
+/* Timed to the measured narration (36.0s). Cues from cue-times.mjs:
+   8.1s "Not of the cheats" · 10.8s "Because formal" · 22.1s "You fix it with" · 33.7s "your own baseline". */
+html,body{margin:0;padding:0}
+body{width:1080px;height:1920px;overflow:hidden;position:relative;
+ background:radial-gradient(700px 700px at 78% 10%,rgba(240,85,111,.20),transparent 60%),
+   radial-gradient(700px 700px at 18% 66%,rgba(47,210,122,.14),transparent 60%),
+   linear-gradient(165deg,#0b1020,#0d1428 55%,#111a2e);
+ color:#eaf1fb;font-family:'Segoe UI',system-ui,-apple-system,sans-serif;text-align:center}
+@keyframes rv{from{opacity:0;transform:translateY(40px)}to{opacity:1;transform:none}}
+@keyframes pop{0%{opacity:0;transform:scale(.4)}60%{transform:scale(1.14)}100%{opacity:1;transform:scale(1)}}
+@keyframes cross{from{transform:scaleX(0)}to{transform:scaleX(1)}}
+@keyframes dim{to{opacity:.34}}
+@keyframes marker{from{opacity:0;left:8%}to{opacity:1;left:47%}}
+.brand{position:absolute;top:80px;width:100%;display:flex;align-items:center;justify-content:center;gap:20px;font-weight:800;font-size:44px}
+.bmark{width:60px;height:60px;border-radius:16px;background:conic-gradient(from 210deg,#db2777,#f59e0b,#22b8cf,#f0556f,#db2777);box-shadow:0 0 0 1px rgba(255,255,255,.15)}
+/* The number this opens with, and that almost nobody quotes. */
+.pct{position:absolute;top:206px;width:100%;font-size:230px;font-weight:900;color:#f0556f;line-height:1;opacity:0;animation:pop .8s .9s forwards}
+.cap{position:absolute;top:472px;left:90px;right:90px;font-size:46px;line-height:1.3;color:#eaf1fb;opacity:0;animation:rv .6s 3.2s forwards}
+.cap b{color:#f0556f}
+.notcheats{position:absolute;top:650px;width:100%;font-size:52px;font-weight:900;opacity:0;animation:rv .6s 8.1s forwards}
+.notcheats s{color:#7c8aa6;text-decoration-color:#f0556f}
+.why{position:absolute;top:754px;left:80px;right:80px;border-radius:18px;padding:26px 30px;font-size:42px;line-height:1.34;
+ background:rgba(10,18,35,.85);border:2px solid rgba(255,255,255,.12);color:#9aa8ff;opacity:0;animation:rv .6s 10.8s forwards}
+.why b{color:#eaf1fb;font-weight:900}
+/* The two questions. The first is struck out; the second is the useful one. */
+.q{position:absolute;left:80px;right:80px;border-radius:18px;padding:28px 30px;font-size:44px;font-weight:800;opacity:0}
+.q1{top:1000px;background:rgba(240,85,111,.08);border:2px solid rgba(240,85,111,.34);color:#ffb9c6;
+ animation:rv .5s 22.1s forwards, dim .6s 23.3s forwards}
+.q1::after{content:"";position:absolute;left:6%;right:6%;top:50%;height:5px;border-radius:3px;background:#f0556f;
+ transform:scaleX(0);transform-origin:left;animation:cross .5s 23.3s forwards}
+.q2{top:1154px;background:rgba(47,210,122,.10);border:2px solid rgba(47,210,122,.45);color:#eaf1fb;
+ animation:rv .5s 24.3s forwards}
+.q2 b{color:#2fd27a}
+/* Their own range, not ours: the mark falls inside it. */
+.scale{position:absolute;top:1370px;left:90px;right:90px;height:96px}
+.track{position:absolute;top:36px;left:0;right:0;height:22px;border-radius:11px;background:rgba(124,138,166,.22)}
+.band{position:absolute;top:36px;left:26%;width:48%;height:22px;border-radius:11px;background:rgba(47,210,122,.42)}
+.mark{position:absolute;top:20px;width:22px;height:54px;border-radius:11px;background:#2fd27a;opacity:0;
+ box-shadow:0 0 22px rgba(47,210,122,.7);animation:marker .9s 29.2s forwards}
+.slabel{position:absolute;top:1482px;width:100%;font-size:38px;color:#7c8aa6;opacity:0;animation:rv .5s 30.2s forwards}
+.foot{position:absolute;bottom:150px;width:100%;opacity:0;animation:rv .8s 33.7s forwards}
+.foot .big{font-size:52px;font-weight:900;line-height:1.24}
+.foot .big b{background:linear-gradient(90deg,#2fd27a,#4aa8ff);-webkit-background-clip:text;background-clip:text;color:transparent}
+</style></head><body>
+<div class="brand"><span class="bmark"></span> SignsOfAI</div>
+
+<div class="pct">61%</div>
+<div class="cap">of essays by people writing in <b>their second language</b>, flagged as AI</div>
+
+<div class="notcheats"><s>Not of the cheats.</s> Of the essays.</div>
+<div class="why">Formal, careful writing <b>looks like a machine</b>, and that is how you write in a language you learned second.</div>
+
+<div class="q q1">Does this look like a machine?</div>
+<div class="q q2">Does it look like <b>whoever wrote the others</b>?</div>
+
+<div class="scale">
+ <div class="track"></div>
+ <div class="band"></div>
+ <div class="mark"></div>
+</div>
+<div class="slabel">their own range, measured on their own work</div>
+
+<div class="foot"><div class="big">If you write formally,<br><b>your baseline is already formal.</b></div></div>
+</body></html>
diff --git a/Docs/Blog/shorts/en/short11-veredicto.html b/Docs/Blog/shorts/en/short11-veredicto.html
@@ -0,0 +1,74 @@
+<!DOCTYPE html><html lang="en"><head><meta charset="utf-8"><style>
+/* Timed to the measured narration (31.0s). Cues from cue-times.mjs:
+   7.4s "The report" · 17.1s "it had never" · 21.3s "One condition" · 26.0s "three hundred".
+   The two cards deliberately repeat the seventh cover, so whoever saw one recognises the other. */
+html,body{margin:0;padding:0}
+body{width:1080px;height:1920px;overflow:hidden;position:relative;
+ background:radial-gradient(700px 700px at 82% 8%,rgba(245,158,11,.16),transparent 60%),
+   radial-gradient(700px 700px at 14% 62%,rgba(109,124,255,.16),transparent 60%),
+   linear-gradient(165deg,#0b1020,#0d1428 55%,#111a2e);
+ color:#eaf1fb;font-family:'Segoe UI',system-ui,-apple-system,sans-serif;text-align:center}
+@keyframes rv{from{opacity:0;transform:translateY(40px)}to{opacity:1;transform:none}}
+@keyframes pop{0%{opacity:0;transform:scale(.5)}60%{transform:scale(1.12)}100%{opacity:1;transform:scale(1)}}
+@keyframes strike{from{transform:scaleX(0)}to{transform:scaleX(1)}}
+@keyframes drop{from{opacity:0;transform:translateY(-26px)}to{opacity:1;transform:none}}
+.brand{position:absolute;top:80px;width:100%;display:flex;align-items:center;justify-content:center;gap:20px;font-weight:800;font-size:44px}
+.bmark{width:60px;height:60px;border-radius:16px;background:conic-gradient(from 210deg,#db2777,#f59e0b,#22b8cf,#f0556f,#db2777);box-shadow:0 0 0 1px rgba(255,255,255,.15)}
+.title{position:absolute;top:186px;width:100%;font-size:74px;font-weight:900;line-height:1.14;opacity:0;animation:rv .7s .3s forwards}
+.title b{color:#f59e0b}
+.card{position:absolute;left:80px;right:80px;border-radius:24px;padding:30px 40px 38px;opacity:0}
+.tag{font-size:34px;font-weight:800;text-transform:uppercase;letter-spacing:.06em;color:#7c8aa6}
+.score{font-size:150px;font-weight:900;line-height:1;margin-top:4px;font-variant-numeric:tabular-nums}
+.score span{font-size:48px;font-weight:700;opacity:.5}
+.vbar{height:26px;border-radius:13px;margin:26px auto 0;width:70%}
+/* Top: the command line. It states a verdict. */
+.a{top:396px;background:rgba(245,158,11,.08);border:3px solid rgba(245,158,11,.42);animation:rv .7s 2.4s forwards}
+.a .score{color:#f59e0b}
+.a .vbar{background:#f59e0b}
+/* Below: the report that goes to a committee. Same text, and it says nothing. */
+.b{top:1004px;background:rgba(124,138,166,.06);border:3px solid rgba(124,138,166,.24);animation:rv .7s 7.4s forwards}
+.b .score{color:#7c8aa6;opacity:.8}
+.b .vbar{background:none;border:3px dashed rgba(124,138,166,.5);position:relative}
+.b .vbar::after{content:"";position:absolute;left:6%;right:6%;top:50%;height:4px;border-radius:2px;
+ background:rgba(124,138,166,.75);transform:scaleX(0);transform-origin:left;animation:strike .5s 9.4s forwards}
+/* The ≠ sits between the bars, not the numbers: the numbers agree. */
+.neq{position:absolute;top:872px;left:50%;transform:translateX(-50%);width:132px;height:132px;border-radius:50%;
+ background:#0b1020;border:5px solid #f0556f;color:#f0556f;font-size:82px;font-weight:900;line-height:126px;opacity:0;animation:pop .5s 17.1s forwards}
+.code{position:absolute;top:1330px;left:80px;right:80px;border-radius:18px;padding:26px 30px;text-align:left;
+ background:rgba(10,18,35,.9);border:2px solid rgba(255,255,255,.12);font-family:Consolas,'DejaVu Sans Mono',monospace;
+ font-size:33px;line-height:1.5;color:#9aa8ff;opacity:0;animation:rv .6s 21.3s forwards}
+.code em{color:#7c8aa6;font-style:normal}
+.nulls{position:absolute;top:1508px;width:100%;display:flex;gap:26px;justify-content:center}
+.n{padding:12px 30px;border-radius:999px;font-size:38px;font-weight:800;font-family:Consolas,monospace;
+ background:rgba(240,85,111,.14);color:#f0556f;border:2px solid rgba(240,85,111,.45);opacity:0}
+.n1{animation:drop .4s 23.1s forwards}.n2{animation:drop .4s 23.7s forwards}
+.tests{position:absolute;top:1606px;width:100%;font-size:44px;font-weight:800;color:#7c8aa6;opacity:0;animation:rv .5s 26.0s forwards}
+.tests b{color:#eaf1fb;font-size:52px}
+.foot{position:absolute;bottom:96px;width:100%;opacity:0;animation:rv .8s 28.6s forwards}
+.foot .big{font-size:46px;font-weight:900;line-height:1.26}
+.foot .big b{background:linear-gradient(90deg,#f59e0b,#f0556f);-webkit-background-clip:text;background-clip:text;color:transparent}
+</style></head><body>
+<div class="brand"><span class="bmark"></span> SignsOfAI</div>
+<div class="title">An <b>if</b> that was<br>always false</div>
+
+<div class="card a">
+ <div class="tag">command line</div>
+ <div class="score">90<span>/100</span></div>
+ <div class="vbar"></div>
+</div>
+
+<div class="neq">≠</div>
+
+<div class="card b">
+ <div class="tag">the report that goes to a committee</div>
+ <div class="score">90<span>/100</span></div>
+ <div class="vbar"></div>
+</div>
+
+<div class="code">RecommendedThreshold is { } threshold<br><em>// the threshold measured for that language</em></div>
+<div class="nulls"><span class="n n1">en → null</span><span class="n n2">es → null</span></div>
+
+<div class="tests"><b>340</b> tests · not one compared the two</div>
+
+<div class="foot"><div class="big">Now it says what it measured:<br><b>&ldquo;No signs above the measured boundary&rdquo;</b></div></div>
+</body></html>
diff --git a/Docs/Blog/shorts/en/short8-artefactos.html b/Docs/Blog/shorts/en/short8-artefactos.html
@@ -0,0 +1,65 @@
+<!DOCTYPE html><html lang="en"><head><meta charset="utf-8"><style>
+/* Timed to the measured narration (28.6s). Cues from cue-times.mjs:
+   5.0s "Swap the Latin" · 13.1s "Seven detectors" · 17.4s "Now SignsOfAI" · 22.8s "A percentage". */
+html,body{margin:0;padding:0}
+body{width:1080px;height:1920px;overflow:hidden;position:relative;
+ background:radial-gradient(700px 700px at 80% 10%,rgba(240,85,111,.18),transparent 60%),
+   radial-gradient(700px 700px at 16% 60%,rgba(109,124,255,.16),transparent 60%),
+   linear-gradient(165deg,#0b1020,#0d1428 55%,#111a2e);
+ color:#eaf1fb;font-family:'Segoe UI',system-ui,-apple-system,sans-serif;text-align:center}
+@keyframes rv{from{opacity:0;transform:translateY(40px)}to{opacity:1;transform:none}}
+@keyframes pop{0%{opacity:0;transform:scale(.5)}60%{transform:scale(1.12)}100%{opacity:1;transform:scale(1)}}
+@keyframes flip{0%{transform:rotateY(0);color:#eaf1fb}49%{transform:rotateY(90deg)}50%{color:#f0556f}100%{transform:rotateY(360deg);color:#f0556f}}
+@keyframes fade{to{opacity:0}}
+.brand{position:absolute;top:80px;width:100%;display:flex;align-items:center;justify-content:center;gap:20px;font-weight:800;font-size:44px}
+.bmark{width:60px;height:60px;border-radius:16px;background:conic-gradient(from 210deg,#db2777,#f59e0b,#22b8cf,#f0556f,#db2777);box-shadow:0 0 0 1px rgba(255,255,255,.15)}
+.title{position:absolute;top:184px;width:100%;font-size:70px;font-weight:900;line-height:1.14;opacity:0;animation:rv .7s .3s forwards}
+.title b{color:#f0556f}
+/* The word. One letter turns and stops being Latin. */
+.word{position:absolute;top:404px;width:100%;font-size:190px;font-weight:900;letter-spacing:.02em;font-family:Consolas,'DejaVu Sans Mono',monospace;opacity:0;animation:rv .6s 1.8s forwards}
+.word i{display:inline-block;font-style:normal;animation:flip .9s 5.0s forwards}
+.same{position:absolute;top:620px;width:100%;font-size:40px;color:#7c8aa6;opacity:0;animation:rv .5s 6.4s forwards}
+.cp{position:absolute;top:684px;width:100%;font-size:44px;font-weight:800;font-family:Consolas,monospace;color:#f0556f;opacity:0;animation:pop .5s 6.0s forwards}
+/* The signal count, collapsing while nothing on screen changes. */
+.count{position:absolute;top:800px;width:100%;display:flex;align-items:center;justify-content:center;gap:40px;opacity:0;animation:rv .6s 8.6s forwards}
+.num{font-size:132px;font-weight:900;font-variant-numeric:tabular-nums}
+.was{color:#f59e0b}.now{color:#2fd27a}
+.arrow{font-size:76px;color:#7c8aa6}
+.clabel{position:absolute;top:952px;width:100%;font-size:38px;color:#7c8aa6;opacity:0;animation:rv .5s 9.2s forwards}
+/* Seven detectors, below chance. */
+.paper{position:absolute;top:1050px;left:80px;right:80px;border-radius:20px;padding:26px 32px;
+ background:rgba(240,85,111,.10);border:2px solid rgba(240,85,111,.36);opacity:0;animation:rv .6s 13.1s forwards}
+.paper .k{font-size:40px;font-weight:800;color:#ffb9c6}
+.paper .v{font-size:58px;font-weight:900;margin-top:8px;font-variant-numeric:tabular-nums}
+.paper .v s{color:#7c8aa6;text-decoration-color:#f0556f}
+/* What SignsOfAI returns now: not an opinion, a coordinate. */
+.rep{position:absolute;top:1268px;left:80px;right:80px;border-radius:18px;padding:24px 30px;text-align:left;
+ background:rgba(10,18,35,.9);border:2px solid rgba(255,255,255,.12);font-family:Consolas,'DejaVu Sans Mono',monospace;
+ font-size:34px;line-height:1.55;color:#9aa8ff;opacity:0;animation:rv .6s 17.4s forwards}
+.rep b{color:#f0556f;font-weight:800}
+.rep em{color:#7c8aa6;font-style:normal}
+.zero{position:absolute;top:1494px;width:100%;font-size:42px;font-weight:800;color:#2fd27a;opacity:0;animation:rv .5s 19.6s forwards}
+.foot{position:absolute;bottom:150px;width:100%;opacity:0;animation:rv .8s 22.8s forwards}
+.foot .big{font-size:58px;font-weight:900;line-height:1.22}
+.foot .big b{background:linear-gradient(90deg,#f59e0b,#f0556f);-webkit-background-clip:text;background-clip:text;color:transparent}
+</style></head><body>
+<div class="brand"><span class="bmark"></span> SignsOfAI</div>
+<div class="title">Find and replace<br><b>switches off a detector</b></div>
+
+<div class="word">d<i>e</i>lve</div>
+<div class="cp">U+0435 · Cyrillic &ldquo;е&rdquo;</div>
+<div class="same">Nothing changes on screen</div>
+
+<div class="count"><span class="num was">17</span><span class="arrow">→</span><span class="num now">6</span></div>
+<div class="clabel">signals found, same paragraph</div>
+
+<div class="paper">
+ <div class="k">7 detectors, COLING 2025</div>
+ <div class="v"><s>MCC 0.64</s> → −0.01</div>
+</div>
+
+<div class="rep">character <b>U+0435</b> CYRILLIC SMALL LETTER IE<br><em>line 37, column 64</em></div>
+<div class="zero">contributes 0 to the score</div>
+
+<div class="foot"><div class="big">A score is arguable.<br><b>A character is there or it is not.</b></div></div>
+</body></html>