Speed
“Speed” – A cartoon that illustrates how business execution speed is no longer an exclusive advantage of startups.
Speed has always been an advantage. Humans survived because we could outrun predators. Napoleon Bonaparte didn’t win with size, he won by moving faster than larger armies. During World War II, the U.S. Liberty ship program brought build times down to ~40 days on average, with a record of 4.5 days, a >10x compression.
That same compression is now happening in AI across all stages of companies … there is no lazy incumbent. In the last 12–18 months, release cycles have collapsed from quarterly to near-weekly. Anthropic pushed ~8 product updates in 8 weeks. xAI is building data centers in ~6–9 months versus 18–24 months historically. Google has already released multiple Gemini model iterations this year.
This is cascading into startup expectations. Seed-stage companies are now expected to hit what used to be Series A metrics, live product, paying users, meaningful traction, within 6–12 months. In Silicon Valley, 996 (9am–9pm, 6 days a week, ~72 hours) is becoming the norm because output expectations have effectively doubled while timelines have halved. At the same time, AI agents are increasing individual productivity by ~2–5x across coding, research, and operations. But the real shift isn’t just output, it’s expectation. Fluency with these tools is becoming table stakes, similar to how basic typing speed stopped being optional a generation ago.
