February Subscriber Post

Bayesian conspiracy, Kabbalistic apocalypse, and Australian prehistory

MidJourney 6.0

It’s been a while since the last subscriber post, figure it’s time to check in. I’m sitting on a couple of longer posts, including EToC v3; keep an eye out for them. In the meantime, there will be more podcast episodes. The plan is for the podcast to take a backseat to writing, which I prefer. However, the audio format is great for collaborative or experimental ideas. Even if you’re here for the snake stuff, I encourage you to delve into some of the other content. The interview with David Stillwell on the state of psychology and the Cambridge Analytica scandal is a rare view into the field.

Links

The Bayesian Conspiracy

Eneasz Brodski

and

Steven

at the Bayesian Conspiracy pod cover the Snake Cult of Consciousness. Throughout the episode, they say, “Andrew would probably say…” and they’re always right. They have an accurate mental model of what I meant to present and add some wonderful commentary. As part of their prep, they had a British AI read the snake cult piece, which you can listen to here:

The Snake Cult of Consciousness - By Andrew Cutler

Unsong

I recently finished Scott Alexander’s sci-fi novel Unsong. As he describes it:

“Kabbalah is true, all patterns are meaningful, and the world runs on a combination of strained analogy and wordplay. Big Silicon Valley corporations copyright the Names of God and make a killing. International diplomats transform the ancient conflict between Heaven and Hell into a US-Soviet proxy war. An autistic archangel and his eight-year old apprentice laboriously debug the laws of physics. A group of billionaires hire a magical ship to go find God and tell Him what He’s doing wrong. Cells of militant Unitarians harbor dangerous placebo terrorists. And amateur kabbalist Aaron Smith-Teller, distant relative of nuclear physicist Edward Teller (“Not ushering in the apocalypse is not really a family strong point”) discovers a legendary Name of God and hatches a plan to usher in the Messianic age from his home computer, which goes exactly as well as you would expect.”

It’s laugh-out-loud funny, and in the end, he does give an answer to the problem of evil (not really a spoiler). Readers of this blog are primed for thought experiments about esoteric Judaism being true, so I think they’ll enjoy it. Unsong had me wondering why Scott knew so much about Kabbalah, and I ran into this piece that answers that and much more:

The Mysticism of Scott Alexander

The problems of Australia’s deep past

I’m often critical of anthropologists in an offhand way. To be legible, my main criticism is quite boring: anthropologists have given up grand theories. Broadly, science has been bureaucratized, and funding is usually awarded to projects that show marginal gains on a narrow topic—for example, characterizing the sound profile of bullroarers. This is interesting work (which I cite), but the elephant in the room is ignored: why are bullroarers used in similar ritual settings worldwide? You need a grand theory to answer this, and those have fallen out of favor.

In the case of bullroarers, the simplest explanation—diffusion—has also fallen out of favor, and not just for being a grand theory. In chemistry, molecules diffuse from higher to lower concentrations. Likewise, ideas such as a ritual instrument diffuse from more to less complex cultures. Identifying a culture as less complex is extremely unpopular in anthropology. Which brings me to my second criticism of the field: that it tends to be swept up by ideologies that warp the facts on the ground beyond recognition.

This is a problem for any scientific endeavor. Is computer science ideological? Sure. It’s a value judgment on whether to prefer an algorithm based on theoretical proofs or empirical results. Or take the more obvious example of the disparate impact of algorithmic decisions. For fundamental reasons, it’s harder for a computer vision system to detect white people in the snow and black people at night. Given that, how should we design self-driving cars? The answer requires values that come from ideology.

Anthropologists give popular ideologies too much leeway to distort the truth. This is hard for an outsider to understand because the debates are often arcane. It’s not usually worth the work to grasp the facts and the ideological battle lines that determine how they are reported and interpreted. Occasionally, anthropologists will try to bend reality in obviously wrong ways. For example, when anthropologist Kathleen Lowrey¹ tried to give a talk titled “Let’s Talk About Sex Baby: Why Biological Sex Remains a Necessary Analytic Category in Anthropology,” it was canceled with an aggressive denunciation. Almost everyone believes men and women are different. It’s a radical position to say they are not. And much more radical to police the speech of colleagues who want to argue sex exists as an analytic category. But this is a dominant position in anthropology.

It’s good to remember that this level of distortion is regularly applied to other areas of anthropology, particularly when it touches on matters of identity. It’s just harder to see because we have no real-world experience with bullroarers or the peopling of Australia. Growing up, I was taught that Australia was settled once, and the Aboriginals have had essentially the same lifestyle since they arrived. It’s a nice, clean story, but there are many problems.

Stone Age Herbalist

read over 100 papers to piece together a more complete vision. I highly recommend his three-part series. In the third, he quotes a book that nicely sums up the situation:

“On an anecdotal level, Australian archaeologists regularly complain amongst themselves about the dismissive or cursory way Australian archaeology is dealt with in global surveys of human prehistory. Yet for the most part, Australianists largely ignore the outside world, positing that the ancestors of today’s Aboriginal Australians came, saw and conquered about 65,000 years ago, and then, with the exception perhaps some 3000–5000 years ago of adopting the dingo, a canine that as a non-marsupial had to come from elsewhere, remained cut off from the outside world until seasonally visiting Macassan sea-slug gatherers from Indonesia began exploiting the northern coastline just before Europeans appeared on the scene. Extraordinarily, most general surveys of Australian prehistory do not even acknowledge more than in passing (if at all) that the continent was joined to New Guinea by dry land from the time of initial colonisation until the early Holocene” ~Lapita: The Australian connection, Ian Lilley

The Problems of Australia's Deep Past - Part One

1. I interview Lowrey on the next podcast episode; stay tuned