Estimate: crudely describing human biology needs 1000x more data than humanity stores
IgorCarron · x · 2026-09-28
A Slavov Lab blog post estimates the data needed to describe human biology: 4×10¹³ cells × 10⁶ molecule types means one snapshot holds 4×10¹⁹ values (200 EB). Accounting for dynamics via hourly sampling yields 5,000 EB per day and 10²⁶ bytes over a lifetime — roughly 1000× all digital data humanity currently stores (10²³ bytes).
This is still crude: localization, molecular interactions, and genotype/lifestyle variability would add many more orders of magnitude. Today's large single-cell studies quantify 10¹⁰ values — about 9 orders of magnitude short of a single snapshot. A raw database-style description of human biology is physically out of reach, so scale alone won't close the gap; commenter Igor Carron notes sparsity must be involved.
More from Infra
- Spectral deflation framework improves Muon: consistent validation loss gains in GPT-2 pretraining — hankyang94 · 2026-09-28
- First Audited Look at Inference-Economics: MiniMax Hit 24.6% Margin, Peer Lost 75% of OpenRouter Volume — AccBalanced · 2026-09-28
- Yunnan Germanium report: indium for InP is tight in China, export controls aren't the bottleneck — pstAsiatech · 2026-09-28
- Apple's free on-device fm paired with decision model Jev beats big-model routing in tests — jasonkneen · 2026-09-28
- Cooling setup runs 6x RTX 6000 at full 325W for 3 months, GPUs at just 41°C — TheZachMueller · 2026-09-28
- MiniMax H3 on RTX 3090: 6,294 s/step, 7-Hour Video Run Lost to Audio Error — play150 · 2026-09-28