4 ms·
http://www.paulgraham.com/know.html http://www.paulgraham.com/know.html >I've read Villehardouin's chronicle of the Fourth Crusade at least two times, maybe th
by sbierwagen 6y ago
http://www.paulgraham.com/know.html http://www.paulgraham.com/know.html
>I've read Villehardouin's chronicle of the Fourth Crusade at least two times, maybe three. And yet if I had to write down everything I remember from it, I doubt it would amount to much more than a page. Multiply this times several hundred, and I get an uneasy feeling when I look at my bookshelves. What use is it to read all these books if I remember so little from them?
>[...]
>But how had I come to believe in this idea in the first place? A combination of my own experience and other things I'd read. None of which I could at that moment remember! And eventually I'd forget that Hilbert had confirmed it too. But my increased belief in the importance of this idea would remain something I'd learned from this book, even after I'd forgotten I'd learned it.
A somewhat more 2020 way of looking at it is to consider NN training. A image classifier might ingest tens of gigabytes of input to produce a model megabytes in size. Reading any given article, though it be thousands of words in length, containing much "information", will only nudge the strength of a couple of connections in your existing neural net. And if you're lucky, commit a single fact to long term memory.
- renewiltord 6y agoThat's a pretty good point. Thank you. Something I think about often is how wrong data still adjusts the weights and moves your position disproportionate to its true likelihood.