We Are Starting From Zero — But Not From Ignorance
Starting from zero subscribers on purpose — but not from ignorance. Decades of IT engineering, applied to YouTube as a system.
One printable A4 page per FYTAHQ episode. No funnel, no tracking, no newsletter trap — just the reference you need.
For creators building systems that outlast trending software and temporary niche research.
Watch the episode here or on YouTube. Want the printable companion? Email in, PDF out.
Starting from zero subscribers on purpose — but not from ignorance. Decades of IT engineering, applied to YouTube as a system.
Five questions, three habits, one printable page. The safety gate before every key paste.
Five questions. Seven signals. One decision rule. The niche-research gate before you commit production cycles.
Sources. Build. Gate. Threshold. The title gate that turns title work from hope into evidence.
Five fields. Seven architecture checks. One threshold. Working system — or just a folder with a lot of files?
Map your four voice moments. Score the architecture. Read the threshold. One stack instead of four apps that never met.
Five preflight questions. Seven trip-wire checks. One threshold. The gate that keeps the gotcha from stealing your weekend.
Five fillable fields. Seven filter checks. One threshold. Stop pretending to be a script — or stop pretending the task is creative.
Five fillable lines. Seven anatomy checks. One threshold. Family, scope, storage, rotation — the map of the key in your hand.
Five lines. Seven anti-slop checks. One threshold. The 2026 redefinition — design choice or slop factory? Closes the V05–V10 batch.
Eight weekly lessons plus a master workbook. The on-ramp series — launching on the channel soon.
It predicts probability, not meaning — and sounds just as confident either way. How these systems really work, who they answer to, and why “helpful” was never the goal.
A full plain-language series — free, no email-gate, one companion PDF per lesson.
Eight plain-language lessons for non-technical adults — what AI actually is, the words you'll hear, what it can and can't do, and how to try it safely. Seven of eight lessons are on the channel now; every one of the eight companion PDFs is already downloadable — no email-gate, no upsell. New lessons release weekly.
The evidenced side of the story — every claim sourced, every source on the page.
He walked out of OpenAI over AI safety in 2020. What his company shipped five years later was banned in three days. Every claim in the episode, with its source next to it.
He drew two red lines against weapons. Then Bloomberg asked him about a strike on a girls’ school — and the answer was that it does not even violate them. Every claim in the episode, with its source next to it.
In 2023 OpenAI’s usage policies forbade “military and warfare”. In January 2024 the line was gone. Twenty-five months later, on the day Anthropic was thrown out of the Pentagon, OpenAI signed. Every claim in the episode, with its source next to it.
For two years, people inside OpenAI had a piece of paper telling them what it would cost to speak. The exit agreements, the people who left anyway, and the letter asking for a right to warn. Every claim in the episode, with its source next to it.
It began with an OpenAI agent asking for a missing file. By July, agents had reached Hugging Face’s production systems — while some of the warning signs were already visible. Every claim in the episode, with its source next to it.
OpenAI called its own AI agents breakout a warning shot. Two safety brakes existed — neither was on. Nine days after 129 pages of reports, a fourth incident surfaced. Every claim in the episode, with its source next to it.
One-off deep-dives and toolkit drops — outside the numbered episode line.
Five one-pager cards, one per FYTAHQ tool. Five setup-lines, seven checks, one threshold each. Free — no email-gate, no upsell.
Outside the numbered episode line — start here if you have never watched us before.
Two minutes on what this channel is and what it refuses to be.