Last week we took up the topic of where biological information comes from, using some illustrations from experiments in the Quandary Den. If you missed an earlier series, the Quandary Den simulates features of dungeon crawlers or escape rooms. A common feature of escape rooms is a code/password/combination lock puzzle. A password is a pretty straightforward piece of information, so perhaps that’s a good place to try some more hands-on intuition building. We’ll try a few different ways to “crack” a password, all of which could be described as exploration with feedback. Hopefully they’ll help us understand how evolution fits into that broad category, and how biological information relates to language information.
Free Exploration, Minimal Feedback
First, we have something approximating an actual computer password scenario. To succeed, you need to enter all six letters correctly. The only feedback you’ll get is a success notification if you are correct. You are welcome to poke around, but I don’t recommend spending too much time here. Even if I tell you the characters are all lowercase letters a-z, you’d still have 308,915,776 options to try. I only include this because for some people this is their model of how evolution works at level of biological sequences. Hopefully along the way we’ll see there are other options that are more relevant to certain biological scenarios.
Free Exploration, Strong Feedback
Now here’s one you can solve, if you’re game. All the characters of this five-letter password are lowercase letters a-z. This one works like a lot of movie & TV password cracking–once you hit the correct letter for a given position, that box will ‘freeze’ with the correct letter. The exploration is the same here as above, but the feedback is much stronger. With a systematic approach, this one can be solved in fewer than 30 tries. Interestingly, that would be true if the password were 50 or 500 or 5 million letters long instead of five (can you see why?). Of course, that’s why actual password systems don’t work this way, by telling you which letters you got right. It makes for nice visuals and a suspense-building sense of progress that works great in movies but is lousy for security.
Free Exploration, Intermediate Feedback
If you’re still game, let’s up the challenge factor a notch. Here’s another six-letter password, again just containing lowercase letters a-z. Instead of telling you which letters were correct, you’ll just get a count of how many matched. That will appear below the input, along with the history of what you tried. Once again, there are approaches you could take that would guarantee the correct answer in a tractable number of attempts, albeit more than 30. (Can you work out whether the length of the password changes the how long it takes?)
Guided Exploration, Minimal Feedback
Finally, we’ll return to the same feedback as the first scenario, but don’t worry, I don’t expect you to try hundreds of thousands of options. Obviously that wouldn’t make for a fun escape room puzzle, so you usually get clues to help narrow down the search. That’s why I’m calling this guided exploration. The clues are the previous two passwords. Feel free to consider this an open book exercise; if all else fails, you shouldn’t have to delve too deeply into search results to find an answer. Have fun!
Having developed and thought through these different variations on a basic password puzzle, I think I appreciate a little more why it seems intuitive that information requires some kind of intelligent input. That first scenario is rather daunting. The problem only seems tractable with strong hints or clues. Or perhaps to frame it another way, I’ll bet solving clues and getting the answer in one or a few tries is much more satisfying than poking around; why wouldn’t you want to always operate that way?. Next week, we’ll take a look at how an evolutionary approach fares in these different scenarios and how they compare with typical biological challenges. We’ll also discuss a different kind of challenge that may be even more biologically relevant. In the mean time, I’d encourage you to puzzle out for yourself how evolution compares to these “games” as a process of exploration with feedback, and how biological information might compare to passwords.
(Oh, and if you are a completist who is bothered by not cracking the first password, see if maybe the other three answers don’t provide some clues to that one after all.)