Join 📚 Kevin's Highlights
A batch of the best highlights from what Kevin's read, .
If you’re mathematically inclined,
then you could use the [pigeonhole principle](https://en.wikipedia.org/wiki/Pigeonhole_principle)
to describe hash collisions more formally:
> Given *m* items and *n* containers,
> if *m* > *n*,
> then there’s at least one container
> with more than one item.
In this context,
items are a potentially infinite number of values
that you feed into the hash function,
while containers are their hash values
assigned from a finite pool.
Build a Hash Table in Python With TDD
Bartosz Zaczyński
As shocking as that might initially seem, we should not to be totally surprised.
Seven years of AI has taught us that deep learning is unpredictable, and not always human like.
“Adversarial attacks” like these have shown remarkable weakness, time and again, repeatedly establishing that what deep learning systems do just isn’t the same as what people do:
[](https://substackcdn.com/image/fetch/f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F8f1f1fed-7908-4cc0-a29b-d88005ebaa10_2000x794.png)
David Beats Go-Liath
Gary Marcus
A [*New York Times* book review](https://en.wikipedia.org/wiki/The_New_York_Times_Book_Review) on Brian Hall's 2008 biography *Fall of Frost* states:
"Whichever way they go,
they're sure to miss something good on the other path."
The Road Not Taken
wikipedia.org
...catch up on these, and many more highlights
Readwise