# Nine Hundred Thousand Responses Live Inside One Added Zero by Bryant McGill > “A safeguard that moved the chance of a harmful output from 0.001 to 0.0001 — one tenth of one percent down to one hundredth of one percent, a quantity small enough to read as rounding error — would, across a billion interactions, carry expected harmful outputs from one million down to one hundred thousand. Nine hundred thousand responses live inside that one added zero.” > **— Bryant McGill**, *Jensen Huang, NVIDIA, Pax Silica, and the Peace We Are Building Into Silicon, 2026* ## Sources and Context - **Original authored source:** [[articles/NVIDIA, Pax Silica, and the Peace We Are Building Into Silicon|Jensen Huang, NVIDIA, Pax Silica, and the Peace We Are Building Into Silicon]] (November 2026). - **Exact marked source:** “A safeguard that moved the chance of a harmful output from 0.001 to 0.0001 — one tenth of one percent down to one hundredth of one percent, a quantity small enough to read as rounding error — would, across a billion interactions, carry expected harmful outputs from one million down to one hundred thousand. Nine hundred thousand responses live inside that one added zero.” - **Editorial treatment:** The Reminder reproduces the complete rendered extraction surface without alteration; article emphasis is not part of the poster text. - **Context:** The arithmetic is explicitly illustrative rather than an empirical measurement; it demonstrates how small probability shifts become large expected differences at planetary interaction volume. ## Related Articles and Collections - **Collection:** [[collections/Simple Reminders|Simple Reminders]] - **Collection:** [[collections/US-Israel Leadership Alliance|US–Israel Leadership Alliance]] - **Article:** [[articles/NVIDIA, Pax Silica, and the Peace We Are Building Into Silicon|Jensen Huang, NVIDIA, Pax Silica, and the Peace We Are Building Into Silicon]] - **Companion article:** [[articles/The American-Israeli Cooperative You Were Trained Not to See|The American-Israeli Cooperative You Were Trained Not to See]] - **Wiki map:** [[wiki/AI Risk Tradeoffs|AI Risk Tradeoffs]] ## Related Topics - [[wiki/AI Risk Tradeoffs|AI Risk Tradeoffs]] - [[wiki/AI Safety|AI Safety]] ## Share on Social Media ``` “A safeguard that moved the chance of a harmful output from 0.001 to 0.0001 — one tenth of one percent down to one hundredth of one percent, a quantity small enough to read as rounding error — would, across a billion interactions, carry expected harmful outputs from one million down to one hundred thousand. Nine hundred thousand responses live inside that one added zero.” — Bryant McGill, 2026 https://bryantmcgill.com/simple-reminders-nine-hundred-thousand-responses-live-inside-one-added-zero-by-bryant-mcgill ```