● For Joe's Eyes Only · Complete Concept & Strategy Report
Proof of Concept · Complete Report
Wanderers of the Cosmos
the cosmos, one small wonder at a time
Prepared for
Joseph Sutliff
Report date
June 15, 2026
Status
Phase 0 — Foundation
End-goal standard
Kurzgesagt-grade
Budget envelope
$50–200 / month
Handle
@wanderersofthecosmos
1

The idea & thesis

An educational, animated, AI-driven science channel — the youth on-ramp to evidence-based wonder, built to run as a self-sustaining loop that grows into passive revenue.

Most faceless channels look identical: a robot voice over stock footage. They scale volume but not identity. The edge here is craft — a flat-vector universe with Studio Ghibli warmth, one warm narrator, original music, and visible rigor. The benchmark is Kurzgesagt; the lesson borrowed from them is that the business is sponsorships, products, members, and audience multiplication — ads are the floor, not the plan.

25–45s
Optimal short length
1 / day
Target cadence at scale
~$125/mo
Stack at full automation
Reach via cross-posting
The one-sentence strategy
Prove the format by hand, lock the brand into a reusable template, then automate the labor — never the judgment — so the channel runs as a loop that compounds.

The editorial north star

From Joe's blog Everyday Bullshit & SOL: "how to think, not what to think." Evidence-first, honest about uncertainty, wonder as the doorway. Each short delivers a fun payoff and quietly models how we know it's true — a curiosity-and-skepticism habit that leads young viewers toward SOL-grade thinking.

The honest gap
Kurzgesagt is a studio — a large team, months per video, hand-crafted frame by frame. A solo, AI-assisted operation won't match that artistry on day one. The realistic target: hit the Kurzgesagt-grade feel at a fraction of the cost using AI + templated motion graphics, and raise the bar steadily as the channel earns.
2

Watch the pilot

"How far is the nearest star?" — a ~23s sample episode. This concept animation shows timing, beats, captions, the scale-zoom, and brand feel; the final render adds full animation, the original score, and the Ray-Porter-register voice.

?
0 min
4.2
light-years
0 trillion km
Wanderers of the Cosmos
the cosmos, one small wonder at a time
Hook0–3s
0–3s"See that star? Of every star in the sky — that one is our closest neighbor."
3–6s"So… how far is 'next door' out here?"
6–10s"Light is the fastest thing there is. From the Sun, it reaches your eyes in eight minutes."
10–15s"But from Proxima Centauri, the same light takes four point two years."
15–19s"Over forty trillion kilometers… just to say hello."
19–23s"Wanderers of the Cosmos. Follow — next, we fall into a black hole."
3

The brand world

A lantern-lit Little Wanderer explores the cosmos from a quiet hill under the stars. Flat-vector shapes for consistency; Ghibli-night soul for warmth. The lantern's glow is understanding — it brightens as each idea clicks.

Night Field palette

#1C2C4C
#141F38
#5FA8A0
#2E4A3F
#F4C95D
#FFB23E
#FBF6EC
#9FB0C7

Type

  • Display: Quicksand (rounded, warm)
  • Captions: Inter; key word glows gold

Voice

  • Warm baritone, Ray Porter register
  • Original (not cloned); one voice forever

Music

  • Original Hisaishi-adjacent score (partner)
  • One sound across every language

Reach

  • Multi-language re-voice + captions, day one
  • One animation, many audiences
Why this wins
Science channels are either cold-techy or generic-faceless. A warm, Ghibli-hearted science world is emotionally inviting — and the counter/scale-zoom motif maps perfectly onto "how big / how far" science.
4

The mascot — the Little Wanderer

The channel's face and guide. A small, brave, curious child-explorer who never speaks — the narrator is the voice. He communicates through posture and the lantern's glow.

Idle Lantern raised — the "aha" Tiny — scale-zoom Lantern states dim · question steady · explore bright · it clicks cloak hood satchel lantern

Character rules

  • Silhouette-first; no detailed face at distance
  • Chibi proportions (head ≈ 45%)
  • Warm, calm, curious — wonder, not a toy

The lantern = emotion meter

  • Dim = an open question
  • Steady = exploring
  • Brightest = the payoff lands
Paste-ready image prompt
Flat vector illustration, Studio Ghibli-inspired warmth, small hooded child explorer holding a glowing lantern, full body. Deep night-blue cloak (#26405E), darker hood, slate-blue satchel, brass lantern casting warm amber glow (#FFB23E). Rounded chibi proportions, no detailed face, strong silhouette, soft inner gradients, subtle grain, starlit indigo background.  Avoid: realistic faces, harsh outlines, neon, text, watermark.
5

The strategy loop

The whole business is one loop that runs over and over. Each pass gets cheaper as manual steps become automated — but a human gate stays before publish.

Ideate
Script
Voice
Visuals
Assemble
◆ Gate
Publish
Measure

What performs becomes the next ideas — that feedback is how a passive channel compounds instead of plateauing.

Quality held

  • Brand locked in a template
  • Every claim fact-checked
  • Human gate before publish

Effort reduced

  • Labor automated stage by stage
  • One asset → three platforms
  • Gate moves to spot-checks over time

Cadence sustained

  • Backlog always 10+ deep
  • Consistency beats volume
  • Winners refill the pipeline
6

Tool stack & costs

A cheap manual stack to prove the format, and an automated stack where the loop runs behind a human gate. Prefer tools with real APIs so automating later doesn't mean re-learning.

StageManual (Phase 1)Automated (Phase 3)
Ideation + scriptClaude / CoworkClaude via n8n
VoiceElevenLabs (~$22)ElevenLabs API
VisualsLeonardo/Flux + Veo heroSame via API
AssembleCapCut (free)Creatomate (~$41)
Orchestration— (you)n8n (free/~$20)
PublishManualYouTube API + cross-post
~$34/mo
Manual proof stack
~$125–150/mo
Full automation
$50–200/mo
Budget envelope
The "animated" decision
Hybrid: templated motion-graphics (Creatomate) as the consistent, automatable backbone; AI image-to-motion for custom scenes; one Veo hero clip where it earns attention. Keeps you out of the generic-faceless look while staying automatable.
7

The automation loop, step by step

The n8n node map. Build it manually first, then replace each step with a node — keeping the gate.

#NodeDoes
1TriggerSchedule or manual run
2IdeateClaude reads pillars + last week's winners → topic
3ScriptClaude writes from voice rules + template
4Fact-checkSecond Claude verifies each claim; flags shaky → human
5VoiceElevenLabs renders the signature voice
6VisualsImage gen per beat + optional Veo hero clip
7AssembleCreatomate renders brand template + captions
8Human gatePreview to Joe — approve / revise / reject manual
9PublishYouTube API + cross-post TikTok & Reels
10MeasureAnalytics → top performers feed step 2
Why the gate stays
One inaccurate or off-brand video can dent a channel's standing. The gate costs ~60 seconds and protects the asset. As trust builds, move it from every video to spot-checking 1 in 5.
8

Shorts + long-form engine

Shorts win attention; 1–5 minute videos win money. The short is the trailer, the explainer is the film — and it earns standard watch-page ad revenue, far above the Shorts pool.

$5–15
Education long-form RPM
$0.03–0.10
Shorts pool, per 1,000 views
5–20×
Long-form advantage
2 paths
Watch-hours + Shorts views to YPP
Format1M monthly views earns…Best at
Shorts only~$30–100Reach, discovery, subs
1–5 min long-form~$5,000–15,000Ad revenue, sponsors, depth
Sequencing
Prove shorts first (Phases 1–2). From Phase 3, add one long-form per week by expanding your best short — it reuses work already done and turns the top idea into your highest-earning asset.
9

Revenue projections & risk map

Honest ranges, not promises. Shorts ad money is low and slow; affiliate, products, and sponsorships are the real levers. This is a compounding play.

HorizonSubsMonthly viewsEst. monthly total
Months 1–30 → 1kbuilding$0 – $75
Months 4–61k → 10k0.5M – 2M$100 – $640
Months 7–1210k → 50k2M – 8M$370 – $2,760
Year 250k+8M – 25M+$1,300 – $7,000+

Bottlenecks & the redirections planned for them

RiskMitigation
"Made for Kids" trap overlooked — COPPA label kills personalized ads (RPM ~$1–3 vs $5–15)Build general-audience / all-ages; bring youth in via accessibility, not a kids designation
You become the bottleneck overlookedBatch approvals; move gate to 1-in-5 spot-checks; templated quality = less to review
Algorithm cold startCadence + patience; seed on TikTok; judge the template, not early views
AI visual consistencyTemplated motion-graphics backbone carries the brand; Veo for hero beats only
Platform/tool dependency overlookedModular, API-swappable stack; own the assets and brand template
Revenue concentrationLayer affiliate + products + sponsorships + long-form from the start
The honest bottom line
The first 90 days likely pay little; the value is the asset and the loop you're building. Each predictable failure point has a pre-decided redirect — so a setback becomes a turn, not a stop.
10

Scaling the portfolio

One theme house, many doors. The same standard pointed at different audiences. Each new channel is a clone reusing the pipeline, template, and fact-check standard, so cost and time-to-monetize fall with each launch.

Axis 1 · New channels

  • New audiences/verticals, same theme
  • Cross-promotion lifts the whole house

Axis 2 · Multi-language Kurzgesagt move

  • Re-voice the same animation (ES, FR, PT, HI…)
  • AI dubbing → near-free reach multiplier
HorizonChannelsBlended monthly viewsEst. blended monthly
Months 1–610.5M – 2M$100 – $640
Months 7–121 → 22M – 9M$370 – $3,200
Year 23 – 49M – 30M$1,500 – $8,000+
Year 35 – 6 + long-form30M – 80M+$4,000 – $20,000+
The guardrail
Don't fork until the prior channel runs itself — automated, monetized, and needing little of you. Grow the system, not your workload.
11

Monetization & compliance

The Kurzgesagt lesson: ads are the smallest layer. Build for sponsorships, products, and members from the start.

LayerWhatUnlocks
1 · AffiliateScience kits, books, gearDay one — before YPP
2 · Ad revenueShorts pool + long-form RPM1,000 subs + 10M Shorts views / 4k hrs
3 · Products / merchPrints, pins, figuresFirst loyal audience
4 · Sponsorships highest valueOne integrated read per video~10k–50k subs
5 · MembershipsRecurring community fundingLoyal core forms
6 · Multi-language + Channel #2Re-voice + clone the scaffoldOnce automated

AI disclosure

  • Synthetic voice on original writing = no disclosure needed
  • Never clone a real person's voice

Anti-spam

  • Original value every time
  • Under ~5 uploads/day

Accuracy & rights

  • Every claim fact-checked
  • Music + visuals licensed or generated
12

Lean expertise architecture

Keep the right knowledge available without bloating every session — a small always-on core that routes to deeper knowledge only when needed.

LayerWhat's in itWhen it loads
Always-on coreCLAUDE.md — goal, loop, rules, indexEvery session
Context modulesBrand, voice, strategy, pipeline docsOn demand by task
Role subagentsScriptwriter, Fact-Checker, Designer, Engineer, AnalystPer stage, own context
Skills + memoryProcedures + durable factsBy trigger / relevance
The routing rule
Core is always on; everything else is pulled on demand. Durable fact → memory. Repeatable procedure → skill. Stage expertise → subagent.
13

The phased roadmap

Prove cheaply before paying to automate. Each phase has one outcome and a clear trigger to advance.

This week
Phase 0 — Foundation
Brand & rails set. ✓ Scaffold, brand, mascot, pilot done.
Week 1–2
Phase 1 — Manual proof
3 finished shorts that look unmistakably yours.
Week 3–5
Phase 2 — Semi-automation
Idea→script→voice→visuals automated; you assemble + approve. 3/week.
Week 6–10
Phase 3 — Orchestrated auto
Loop runs daily; you approve a preview. Long-form companion begins.
Month 3+
Phase 4 — Monetize & scale
YPP → ads; affiliate + products + sponsors; clone for channel #2.
14

Status & decisions

Done

  • Strategy + risk + scaling report
  • Brand identity + mascot reference
  • Scripted, fact-checked, animated pilot

Your move

  • Claim @wanderersofthecosmos (free on YouTube)
  • Confirm TikTok / Instagram handles
  • Deploy this report to a .pages.dev URL

Next build

  • Audition the signature voice
  • Generate illustrated frames from the mascot prompt
  • Produce S008 as a real render

Supporting documents (in the project folder)