Great stuff, thanks so much. Let me add to your collection: (1) The Pardoner's Tale in Chaucer's "Canterbury Tales" shows A+B>C & C>A+B. So this is "Death by Symmetry". (2) One of the stories in "Queen's Bureau of Investigation" by Ellery Queen involves 16 people being sequentially given 3 stock tips (buy or sell). At the end of this process, there should be 2 people who are convinced of the tipster's infallibility. So the point of the story is: why are there 3? I think it's the story "Double Your Money" but I haven't got a copy to hand.
Thanks, this was an interesting watch. I just wanted to give my thoughts on Queneau's formula for "lipogrammatic difficulty", which doesn't feel mathematically well-motivated. It's more than twice as hard to write a text excluding 50% of letters (by frequency) than one twice the length excluding 25%. So we should probably look at the total frequency of letters *not* excluded, and take (the absolute value of) the logarithm of that. This won't change things much for lower frequencies of omitted letters - La Disparition is still slightly more impressive than Les Revenentes - but will for higher frequencies. Also, we should arguably be counting frequencies not of excluded letters, but of words or even sentences. The numbers are still easy to estimate, at least with a computer: just look through a given text corpus and see what proportion of words / sentences lack the letters we're interested in. Exactly the same method works for texts excluding letter clusters.
This ties into the use of the AI GPT-3. I would like to propose a quote, "Any sufficiently complex system is equivalent to intelligence." You can witness this on your cell phone when it suggests words for you which mathemagically make sense to you and you alone.
Great stuff, thanks so much. Let me add to your collection:
(1) The Pardoner's Tale in Chaucer's "Canterbury Tales" shows A+B>C & C>A+B. So this is "Death by Symmetry".
(2) One of the stories in "Queen's Bureau of Investigation" by Ellery Queen involves 16 people being sequentially given 3 stock tips (buy or sell). At the end of this process, there should be 2 people who are convinced of the tipster's infallibility. So the point of the story is: why are there 3? I think it's the story "Double Your Money" but I haven't got a copy to hand.
Thanks, I shall look into this straightaway! Glad you enjoyed the lecture.
Thanks, this was an interesting watch. I just wanted to give my thoughts on Queneau's formula for "lipogrammatic difficulty", which doesn't feel mathematically well-motivated.
It's more than twice as hard to write a text excluding 50% of letters (by frequency) than one twice the length excluding 25%. So we should probably look at the total frequency of letters *not* excluded, and take (the absolute value of) the logarithm of that. This won't change things much for lower frequencies of omitted letters - La Disparition is still slightly more impressive than Les Revenentes - but will for higher frequencies. Also, we should arguably be counting frequencies not of excluded letters, but of words or even sentences. The numbers are still easy to estimate, at least with a computer: just look through a given text corpus and see what proportion of words / sentences lack the letters we're interested in. Exactly the same method works for texts excluding letter clusters.
Why is the lecturer's name not given in either the title or description?
30:00 telltale games approves this message
This ties into the use of the AI GPT-3. I would like to propose a quote, "Any sufficiently complex system is equivalent to intelligence." You can witness this on your cell phone when it suggests words for you which mathemagically make sense to you and you alone.
Hiii
Hello! Ty
Hey I am an American and having hard time understanding your speaking.
I'm an American too and I have no problem understanding her. 🤔
Are the French words throwing you off? Well too bad!
Hahahahaha. 😆👍❤