Back to Lenski
Chapter 7 described a real experiment in which bacteria acquired a new ability: Richard Lenski's flasks of E. coli, one of which, after about fifteen years, learned to feed on citrate. Set that experiment beside the Weasel program, and three differences stand out.
The first is time. In the one flask where it happened, about a trillion and a half bacterial divisions, and some twenty thousand real generations, went by before even the first, easier change was in place. The second change took nearly another trillion divisions. You may have noticed that this is rather more than the forty-odd generations the Weasel program needed.
The second difference is survival. A mutation isn't the end of the story; the organism carrying it still has to survive. In one sense, the flasks were a comfortable place for a mutation to happen — a warm bath of glucose, served fresh every day, no predators, no competitors from outside. In another sense, they were a genuinely dangerous place to be. Every day, ninety-nine cells in a hundred were thrown out. A new mutation starts out in a single cell, and whether it survived that daily cull depended heavily on timing. A mutation that occurred late in the day left little time to multiply before the cull arrived, and a lone cell, or a small handful of them, faced real odds of being wiped out, taking the mutation with them. A cell that mutated near the start of the day's growth would typically have around a hundred descendants by the time the cull came, which made it reasonably likely that at least one of them would be among the one percent carried forward. The first change seems also to have had a modest edge of its own, helping its carriers grow and hold their ground day after day. It may well have arisen earlier, more than once, and simply been lost. So may the second. There's no way to know.
A mutation occurring, in other words, is not the same thing as a mutation taking hold. It still has to survive, and then it still has to spread, and neither is guaranteed in the real world. None of this appears anywhere in the Weasel demonstration. In that game, the best copy each round always survives, always prospers, and never has to compete for anything — not for survival, not for the chance to reproduce. By leaving all of this out, and never once mentioning that "forty generations" was only ever a figure of speech, the program misleads without ever having to say a false word out loud.
The third difference is order. In Lenski's flask, the second change worked only because the first was already in place; put the second into a cell without the first, and it produces only a feeble trace of citrate use. The order mattered. In the Weasel world, order never matters at all. Any change that happens to match a target letter is kept the moment it occurs — first, middle, or last in the sequence, it makes no difference whatsoever.
Last revised 2026-11-19