More on this book
Community
Kindle Notes & Highlights
the study of placebo was really the study of context and expectation and faith
as the algorithm has no ability to understand the language that is written on a website or parse a website’s arguments or logic. The best it can do is simply take every individual word it finds on a website and assign that word a point in theoretical space and then analyze how those points cluster together, assigning high search-result values to those words that often appear near each other, which is why two very different sentences— The government is poisoning us with fluoride and The government is not poisoning us with fluoride
are, for the algorithm, basically identical. They both have most of the same words clustered in most of the same ways.
This is why websites that attempt to debunk conspiracy theories end up paradoxically boosting those very conspiracies, by making specific combinatio...
This highlight has been truncated due to consecutive passage length restrictions.
It’s another one of those loopholes the math m...
This highlight has been truncated due to consecutive passage length restrictions.
It’s why these websites all seem to reach the same conclusions and describe those conclusions in much the same language: it’s just a reflection of the underlying algorithm’s mathematical bias toward linguistic clusters and patter...
This highlight has been truncated due to consecutive passage length restrictions.
what the algorithm promotes and rewards is a kind of cognitive quicksand. It is predisposed to deliver stories to get trapped in.
Thus in this specific circumstance, what the algorithm amplifies is not actually authority but instead intensity. Obsession. Zeal. Pique.
Which makes for a kind of airtight, unfalsifiable theory whereby evidence against the conspiracy becomes, weirdly, evidence for the conspiracy.
semantic paradoxes (a sad lie if true!!!
when there’s an informational vacuum, the mind will naturally rush in to fill it.
