AI Compendium · Commission briefs — photos and illustrations
The only outstanding figures in the Compendium are external (not authorable in SVG): ~12 historical photos and ~7 pictorial scenes from Part III. This document delivers each one ready to commission — photos with source and license to verify; illustrations with a complete image-generator prompt. General catalog:
FIGURES.kmd.
Delivery spec (applies to all)
- Where to save:
03-linha-do-tempo/assets/F-III.<n>-<slug>.<ext>(e.g.,
F-III.1-turing.jpg,F-III.6-dartmouth.png). - Resolution: illustrations ≥ 1600 px on the longer side; photos ≥ 1200 px.
- How to embed: same pattern as the SVGs —
right after the brief in the.kmd. - Format: illustrations in SVG (if vector) or PNG; photos in JPG/PNG.
- Credits: record source + license in
../07-appendices/04-index-and-bibliography.kmd(Credits section) — mandatory for CC-BY/CC-BY-SA photos.
A — Historical photos (curated from free archives)
License rule: prefer public domain or Creative Commons (CC0 / CC-BY / CC-BY-SA). Verify the license of each file on Wikimedia Commons before publishing and attribute as required. For the public site (commercial + derivative use), confirm the license permits both.
Unifying treatment (house style): crop into a circular medallion, apply duotone (shadows in #1d3557, highlights in #f1faee), thin ring #a8dadc. This harmonizes photos from different sources into the Compendium's identity.
| Figure | Person / object | Suggested source (verify license) | Note |
|---|---|---|---|
| F-III.1 | Alan Turing (1912–1954) | Wikimedia Commons → "Alan Turing" (portrait ~1951) | PD in several jurisdictions; verify |
| F-III.2 | Claude Shannon (1916–2001) | Commons → "Claude Shannon" | verify PD/CC |
| F-III.4 | John McCarthy (1927–2011) | Commons → "John McCarthy (computer scientist)" | CC-BY-SA available |
| F-III.5 | Perceptron Mark I / Frank Rosenblatt | Commons / US Navy archive / Cornell | gov. equipment photo = likely PD |
| F-III.7 | Geoffrey Hinton | Commons → "Geoffrey Hinton" | existing CC-BY |
| F-III.8 | Yann LeCun | Commons → "Yann LeCun" | existing CC-BY-SA |
| F-III.10 | Vladimir Vapnik | Commons → "Vladimir Vapnik" | existing CC-BY-SA |
| F-III.11 | Fei-Fei Li | Commons → "Fei-Fei Li" | existing CC-BY |
| F-III.13 | Ilya Sutskever | Commons → "Ilya Sutskever" | existing CC-BY |
| F-III.14 | Demis Hassabis | Commons → "Demis Hassabis" | existing CC-BY-SA |
| F-III.16 | Transformer authors | — (no free group photo) | fallback: illustration of attention (see F-III.17) |
| F-III.18 | Anthropic team | — (no free photo) | fallback: illustration/brand, no faces |
Where no free photo exists (F-III.16, F-III.18), use an illustrated portrait in the Compendium's style or a conceptual scene — never a photo without a clear license.
B — Illustrations (image-generator prompts)
Style prefix (paste before each scene):
*Flat editorial vector illustration, clean modern infographic style. Limited palette: deep navy #1d3557, medium blue #457b9d, light blue #a8dadc, red accent
e63946, off-white background #f1faee. Soft geometric shapes, subtle depth, no
photorealism, generous negative space, didactic and elegant museum-poster aesthetic. 16:9 landscape.*
Negative prompt (default):
*photorealistic faces, watermark, brand logos, cluttered composition, harsh gradients, dark background, low contrast, distorted text.*
The prompts below are in English (the convention for image models); the final caption in the Compendium stays in en-US.
F-III.3 — From logic to machine (Era 1)
Prompt: <style> + "A conceptual left-to-right composition: on the left, Boolean algebra symbols and logic notation engraved on stone tablets and old paper; in the center, the paper tape of a Turing machine threading through a simple read-write head; on the right, a stylized McCulloch–Pitts neuron morphing into a clean circuit node. The visual reads as the passage from abstract thought to physical mechanism."
F-III.6 — The Dartmouth summer (Era 2)
Prompt: <style> + "A 1950s seminar room, mid-century modern, optimistic and retro. A chalkboard with abstract diagrams; researchers in period attire in animated discussion; to the side, the Perceptron Mark I machine as a tall panel of wires and dials, and the Shakey robot as a boxy wheeled robot with a sensor head." *(Override: the only word allowed on the board may be "AI"; otherwise no text.)*
F-III.9 — Ice and rules (Era 3)
Prompt: <style> + "A split composition. Left half: a frozen winter landscape with downward funding charts and a dusty, switched-off LISP machine (the AI winter). Right half: an expert-system flowchart of IF-THEN rule boxes and a doctor's coat (MYCIN). At the bottom center, a single small glowing ember surviving the frost, symbolizing backpropagation enduring."
F-III.12 — The tide of data (Era 4)
Prompt: <style> + "A large ocean wave formed from millions of tiny images and data records (the ImageNet dataset). On one side a chessboard (Deep Blue) sinking into the water; on the other side stacks of GPU boards rising from the foam. Conveys the silent accumulation of data and compute."
F-III.15 — The cat that changed everything (Era 5)
Prompt: <style> + "An iconic deep-learning scene: a convolutional neural network processing a cat photo through visible layers (edges, textures, parts, whole object) as stacked translucent planes; a glowing GPU board pulsing beside it; in the soft background, a Go board with a single highlighted stone (AlphaGo's move 37)."
F-III.17 — The attention that connects everything (Era 6)
Prompt: <style> + "An infographic of self-attention: a short sentence with weighted arrows linking each word to every other word, scaling outward into a vast connected graph (a large language model). Beside it, a folded protein ribbon (AlphaFold) and an image crystallizing out of visual noise (diffusion). Theme: attention plus scale unlocks everything." *(Can also be done in SVG if preferred — it's semi-schematic.)*
F-III.19 — The AI that converses, creates, and acts (Era 7)
Prompt: <style> + "A triptych panel. (1) A chat bubble representing conversation. (2) Screens simultaneously generating an image, a video frame, a music waveform, and lines of code (multimodal). (3) A stylized agent figure operating a computer with multiple tools — hands on a keyboard, API call icons. Theme: perception + generation + action, the present era."
Summary
| Class | Qty | Method | Blocker |
|---|---|---|---|
| Photos | ~10–12 | Wikimedia/PD curation + duotone | per-file license verification |
| Illustrations | 7 | image generator (prompts above) | access to an image generator |
With these briefs, any image generator or illustrator produces the assets directly in the Compendium's style; once they arrive, just save them in 03-linha-do-tempo/assets/ and embed them — the rest of the Compendium is already done.