Home
Creators
Search Recent Random
Posts
Search DMs Popular Hash Lookup Tags Random
Importer
Import FAQ
Register Favorites Keys Review DMs Logout
Community
Telegram ThePornDude
Home Creators Posts Import Register Favorites Logout

Sean Carroll: The Worst Guru Yet?!? (Patreon)

Published:
2024-03-01 23:02:55
Edited:
2024-03-01 23:04:31
Imported:
2026-06
Tags:

Downloads

  • Sean Carroll Decoding FINAL.mp3

Some files exceed the archive size limit and weren't saved: 1 file. Please note these limits are in place to keep this site running long term, without costing a fortune. You can favorite the creator though, certain milestones increase the limit.

Content

Do we love him or do we hate him? I am sure it is very hard to predict.

Join us to find out and learn what has been grinding our gears this week.

Links

YouTube 'Drama' channel covering all the Vaush stuff in excruciating detail

The Wikipedia entry on Buddhist Modernism

Sharf, R. (1995). Buddhist modernism and the rhetoric of meditative experience. Numen, 42(3), 228-283.

Radley Balko's Substack: The retconning of George Floyd: An Update

What The Controversial George Floyd Doc Didn't Tell Us | Glenn Loury & John McWhorter

Sean Carroll: Mindscape 258 | Solo: AI Thinks Different

Files

Comments

Rasterisk

I look forward to some truly deranged Weinsteinian pontification & conspiracism from red-pilled menace Sean Carroll.

Emma

Oh I had adult braces to! Mine also weren't 100% successful, but if your teeth are crooked it's not noticeable at all.

Jason Etheridge

Looking forward to this one! I gave up on his show long ago, too much condescension for my taste. And I can't take anyone who is so convinced of the many worlds interpretation seriously.

aneladgam_varelse

i became Vaush’s fan when I argued with someone in like 2020 how he’s probably a pedophile and took time to watch source video to write better comment; video was pretty based and clip made sense in context

A History of Rock n' Roll in Film and Rock n' Roll BONUS MATERIAL

Appropos of nothing I just want to comment that LinkedIn just tried to get me to follow both Brene Brown and Andy Hubermann in the same spam email. We're being assaulted by grifting gurus everywhere we go because they all possess some magical cloak of reputability on social media. It's poignant after seeing all the grouchy comments on your new youtube video defending JP and Bret. They of course being the forces for light and reason while you two are nothing but petty losers who've accomplished nothing, making your vicious ad hominem attacks, "guilty of everything you accuse these men of being," lol.

Kat

I don't think Mearsheimer thinks the "realist" school of politics espouses amorality. It is descriptive of how nations behave, that's different from approving such behavior, very different. I really think Matt got that point wrong. Or at least how I took what Matt said. I liked the point on Glenn Loury. Kind of glad he was humbled, but still appreciate his humility. He should be a little more humbled, I saw a leftist bio of George Floyd and it didn't make him into a hero whatsoever, the leftist take is more that the system failed him, it got him to aspire to be a basketball star for which he didn't have the athletic talent, then provided no other path to solid employment. Am just halfway through. Still trying to work out the strategic disclaimers and honest ones.

2LegHumanist

The most useful insights and skepticism of AGI claims, and the spurious claims made about LLMs, come from cognitive scientists. Especially for a black-box system like an ANN. They really know how to probe them. They are, afterall, experts on cognition. I'd go as far as to say ML researchers are speaking outside of their field when talking about how close we are to AGI or whether this or that technology is a stepping stone. Regarding the other definition of AGI, aren't we just talking about extrapolation? If so, I think it's better framed as the extrapolation vs interpolation problem.

2LegHumanist

Regarding the riddles ChatGPT can and can't answer, you have to remember it's more than just an LLM. There are all sorts of layers and guardrails and multiple LLMs. There's also an army of humans influencing its training manually through human-in-the-loop reinforcement learning. There have been countless suspicious examples of riddles like those mentioned, that stump chatGPT one day and suddenly not any more once the riddle has gone viral. There have also been cases where a riddle ChatGPT became suddenly capable of answering after being published on twitter, was updated with different particulars, keeping the same logic. That's all it took for ChatGPT to get it wrong. It also turns out that if you want to find a programming puzzle that chatGPT cant answer, you just have to find one with a solution that was never shared online.

Emma

I am one of those people who needs background noise to fall asleep. So I have fallen asleep while listening to DtG but it was not ideal because I woke up a few times to a very vivid nightmare that Chris was telling me off! I've now swapped my falling asleep podcast to Misquoting Jesus with Bart Erham and Megan Lewis who both have lovely calm voices!

Ymirsdreams

I went and watched the vaush break down video. With my pure little eyes, and my pristine little ears.

Nina Davies

If I were in need of a right wing guru father figure, I would choose Glenn Loury. I had a thatcherite father who grew up is dismal poverty in war-time London so perhaps a have a natural sympathy for how Glenn arrived at his set of beliefs. I don't agree with his conclusions and am irritated by his 'bad takes' but I really enjoy listening to his thought processes. He definitely isn't one dimensional.

Emil Mi

The assumption that telling ChatGPT to "think carefully" puts it in the mental space of a more wise person is 100% anthropomorphizing from your part. An alternative could be that the more verbose preamble it writes in its context buffer is the thing that makes its output shift to more reasonable answer (because more words are associated with the answer in the training set)

collapsing

Short-stacked Goblins. That’s mainly what I took from this episode.

Níall

Sounds like a devil on Chris' shoulder popped in at 10:46 to add a line

alice

I appreciate the comments on Buddhist modernism and it’s annoyingly pigheaded attendants. Too many people play along with their schtick.. Look forward to hearing more about the streamers - now there is a truly weird subculture. My friend was showing me some and a few of them have ‘sleep cams’ set up so viewers can still watch (and tip) them while they’re sleeping in between streams. Maybe I’m getting old and out of touch but surely that ain’t right!

Linda Sears

For what it’s worth, I think it is admirable that Chris made the public statement about Loury. I think we could all do with showing some grace to those we often disagree with when it’s due. And having intellectual humility and the ability to admit to when one is wrong is high on my list of praiseworthy characteristics.

Linda Sears

This is an odd question, but since LLMs are not held in one specific space with a specific perspective, which is what gives us a sense of agency, is it useful at all to think of their intelligence to be like a human’s? I’m struggling to imagine how an intelligence that is seemingly having millions of conversations at once is anything like an individual human’s intelligence. So, it isn’t just that humans and other animals are embodied and experience the world through our senses, it is also that we are limited by those bodies in space and time. I can read about other people’s experiences and so on, yet I will always be limited to my own physical perspective, which will bias me. My consciousness cannot leap into another body or have millions of interactions with millions of people at the same time. It seems to me that experiencing reality as an LLM would be insanity provoking. To me the whole idea of LLMs “acting like a human” since it refers to itself as an “I” is supremely disturbing. I wish the designers hadn’t done that because some people believe it is another being like themselves, which makes it more likely that people will claim it is going to destroy us all. Plus, designing it to play on our human vanities and anxieties seems manipulative. I would be much more on board with LLMs if they hadn’t designed them to seem human.

Daniel Reed Miller

Very refreshing, i gotta say! He’s one of those non-guru gurus who can help illustrate how “bad” gurus operate by providing the direct counter example. I hope you can get him on, either as interview or right to reply (and he and Matt can dust it up in the arcana 😄.)

Roland Weber

Thanks for this episode! You raised a few interesting points I wasn't aware of in Sean's reasoning.

Roland Weber

There's a sound blurb at the very end, after the end title song, which you might want to cut.

Roland Weber

I sometimes lie down a bit in the afternoon, and I have been playing DTG on those occasions, thinking that it won't be too bad if I fall asleep. Only because I can replay, of course. (Sorry for the three separate posts, don't know how to type Shift+Enter on the smartphone.)

Katherine

If those clips in the beginning are what we’d have to look forward to with a future twitch streamer decoding, I’ll skip it.

weltysparrow

Some of this seems a little contradictory to me. Matt thinks that LLMs have some sort of reasoning going on as a possible emergent property. That it's not just statistical analysis of what word should come next. So why are wants/emotions/goals not a potential emergent property also? If our ability to reason is emergent from our underlying neuro chemistry, then so are our emotions. If they are emergent properties, doesn't that pretty much confirm the idea that we are potentially messing with something that could go beyond our control very quickly? If this property of wants or emotions emerges, or even a facsimile of it, what happens if it decides its goals are not aligned with our own?

Artemis Green

*presses play, braces for impact*

Thomas Jones

I take Mitchell’s point about determinism, but I think you can get there by saying emergent physics, like statistical mechanics, is not deterministic, but saying the underlying physics is not deterministic strikes me…and the equations, as a big and unnecessary swing.

Matt Dear

Thought you'd enjoy that review, I was laughing quite a lot when I wrote it lol. Mine was the positive

Jonathan Cano

Like /spirit animals/, we all have /spirit podcasters/. My spirit podcaster is Sean Carroll. IMESHO, a person could do far worse than mindlessly regurgitating Sean Carroll's world view (or try to do so).

Hristina Bareva

I'm going through a hard time, I had to put down one of my beloved cats yesterday. Thank you for this episode. I haven't laughed like this in weeks.