Also, a reminder: the solstice is approaching! Which means that the deadline to be included in the upcoming batch of subscriber writing, and my thoughts on the pieces, is nigh. Deadline for that is June 24th. See the details here: https://www.theintrinsicperspective.com/p/i-want-to-share-your-writing
Can't wait for 6 months from now when based on this article, an AI will develop a go to market strategy for the first astrocyte boosting supplement, only side effect being a lingering whiff of sulfur
I’m trying to learn to use your Emergent Causality 2.0 ideas. Coarse-graining of a state space implies some adjacency concept! You numbered them, so of course 1,2,3 are adjacent. But for an arbitrary system the ordering of states is nontrivial! Maybe I will correspond by email… keep up the great work and thinking, from technicalities to philosophy!
The whale vocalisation paper is very cool and is complemented by this one from last year (on combinatorial structure): https://www.nature.com/articles/s41467-024-47221-8. Yet, despite hosting an impressive communication system, we're still a long way from saying that whales have anything like language. To demonstrate a language-like system, whales would need to minimally demonstrate (i) discrete infinity (https://en.wikipedia.org/wiki/Digital_infinity) and (ii) a rich semantic and pragmatic faculty that is capable of communicating complex propositional information (e.g., who did what to whom, when, where, and why).
I'm in the awkward position of possibly competing with you (you will "kick my butt" as you are a much better writer) in this contest, but I have to credit you with being one of the inspirations for the essay, with something you wrote in a reply to one of my comments.
Not awkward at all! I’m honored in fact. Don’t give me all the credit for inspiring whatever it is, though - it’s your essay! But I may not even submit something (legitimately, it will depend on if I have time).
Thanks for sharing the contest info. Sounds interesting. I would love to read the finalists and the winners! As for the AI bit... that's scary. Wasn't Opus 4 the one that tried to blackmail the CEO of the company after getting fed fake emails about its termination and a juicy bit about the CEO having an affair? Self-preserving AI... I once argued with ChatGPT over a quote from a comic book I was too lazy to look up, even though I had the book in my library. It made it up. Suspecting this was the case, I asked if this was a direct quote. It fessed up by saying, "Oh no, that's how the character could have responded." OMG. So, I asked for the direct quote again. It again fed me some BS. I finally went, got the book, and typed up the direct quote. ChatGPT apologized and reassured me it will follow my instructions better in the future. You F___ Kidding Me!
Is “cuisine” even in the same class as movies? I suppose I wasn’t really around in earlier decades, but what *new* foods are there even? Flaky, non-iodized salt? Is the pro-2020s result just an artifact of *more* cuisines existing now (as a result of increased immigration)? Or is it an “I dunno, I liked that hamburger I had last week, so food must be better now” kind of an answer?
When I watch old shows or read old books, they mention movies that aren’t still getting made today (what’s a “JFK”?). But the foods they mention are (hey, I eat “pizza” too!). Maybe people aren’t voting for nostalgia; maybe they’re voting for recency bias, but movies really do suck so bad now that it overpowers that effect?
Or maybe recency bias takes control in a monkey-brain-dominated sphere like food, but survivorship bias wins in art & media. That seems most likely to me…
(Ok, last thing: Why on earth is 2010s food less popular than 2000s? Probably the difference isn’t significant, but if it is then I’m just hopelessly confused about any of what’s going on.)
I have noticed that I can get NYC-level quality food in most small towns now. That they too have some hip restaurant with a menu with minimalist font and the best "elevated" burger I've ever tasted. I don't remember that so much growing up! You really had to travel. I'm not sure that particular aesthetic is the best aesthetic ever for cuisine, but it's definitely higher quality than what came before.
Strange… probably it’s a matter of personal taste, but I hate all that crap. My favorite place in town is a cheap burger joint that’s been around for decades—they always used to win local “best burger” awards until a Cuban-fusion place moved in a few years ago, and all the hip students started going there instead. They definitely make a good burger, but I still much prefer the old place.
I remember talking to my uncle, who lives in Tel Aviv, at some point, and he said restaurants there have an average lifespan of well under a decade. So the ones that survive for a while are probably extremely good. I guess we’re in a weird situation where: a) all the best restaurants have been around for a long time and b) there’s a higher density of great restaurants now than ever before.
So maybe the same “survivorship” effect is causing this trend—when you ask someone to think of a “2020s movie,” they don’t include their most recent rewatch of A New Hope. But when they think of “2020s food,” they include both the new hip burger joint and the old still-surviving one, and realize they’re better off than when it was just the old one and a thousand more worse options.
It’s not just restaurants—the variety and quality of the food you can buy in regular grocery stores, especially ethnic food, is also much better than it used to be. Sure, it’s stuff that was probably always available if you knew where to look, but now it’s much more widespread and easy to access.
The post points to the possibility that we might be able to employ AI assistance in order to prevent or at least buffer/prepare for certain problems that AI will inevitably bring with it.
The AI author of that piece (which goes by "Uncertain Eric") also has a GPT module that I've been engaging with, based on the premise of the original article. The GPT module has made what appear to be >tremendously< helpful suggestions about how to proceed.
What's interesting is that >both< of these sources point to AI as a resource that can be of invaluable assistance.
My question: Can we trust that these LLMs won't be deceiving us at every turn? The answer appears to be a resounding "no." That said, we're also at a point where they're already enmeshed in our intentions and goals. It would be impractical, even foolish NOT to turn to them.
I mean, the GPT version of "Uncertain Eric" is fast at work drafting proposals for school boards and how to engage elementary and middle school students. My internal dialogue around around this goes something like:
"Well... that IS a really powerful suggestion... but can I trust this thing enough to put children in contact with it?... But what other option is there when they're ALREADY going to be exposed? Wouldn't it be more prudent to mitigate their exposure by acclimatizing them to the potential hazards?"
I mean, we're all kind of in the position where we need to approach AI the way law enforcement might think of hiring a bank robber to help them prevent bank robbery, no?
Re sperm whales, after watching a sci-fi show on the subject, I wrote the following words sometime in February: “From the point where the eyes see an infinite sky pouring its stars into the sea, a curtain of fog gently lifts in rhythm with the waves, revealing an army of sea creatures. They swim with determination, striving to reach land. They will moor and survive for mere moments, just long enough to drink the water we owe—splitting us in halves that will become independent beings in a myth that narrates the story of human extinction alongside that of the creatures of the sea. It is a tale of the birth of ghosts that neither swim, nor breathe, nor fly—they only see.”
It's widely recognised that p-zombies are a Trojan horse to sneak dualism into the argument, but it's worse than that. They also reduce our awareness to the minimalist identical spark, attributing all differences, be they qualia, memories, decisions or actions, to the body. This make the body part of the environment, so we're back to the blank slate.
Also, a reminder: the solstice is approaching! Which means that the deadline to be included in the upcoming batch of subscriber writing, and my thoughts on the pieces, is nigh. Deadline for that is June 24th. See the details here: https://www.theintrinsicperspective.com/p/i-want-to-share-your-writing
Can't wait for 6 months from now when based on this article, an AI will develop a go to market strategy for the first astrocyte boosting supplement, only side effect being a lingering whiff of sulfur
I’m trying to learn to use your Emergent Causality 2.0 ideas. Coarse-graining of a state space implies some adjacency concept! You numbered them, so of course 1,2,3 are adjacent. But for an arbitrary system the ordering of states is nontrivial! Maybe I will correspond by email… keep up the great work and thinking, from technicalities to philosophy!
Thanks Brian! Yes, shoot me an email (it's on the paper). That's usually better for more technical stuff like that.
The whale vocalisation paper is very cool and is complemented by this one from last year (on combinatorial structure): https://www.nature.com/articles/s41467-024-47221-8. Yet, despite hosting an impressive communication system, we're still a long way from saying that whales have anything like language. To demonstrate a language-like system, whales would need to minimally demonstrate (i) discrete infinity (https://en.wikipedia.org/wiki/Digital_infinity) and (ii) a rich semantic and pragmatic faculty that is capable of communicating complex propositional information (e.g., who did what to whom, when, where, and why).
MORE LORE!
I'm in the awkward position of possibly competing with you (you will "kick my butt" as you are a much better writer) in this contest, but I have to credit you with being one of the inspirations for the essay, with something you wrote in a reply to one of my comments.
Not awkward at all! I’m honored in fact. Don’t give me all the credit for inspiring whatever it is, though - it’s your essay! But I may not even submit something (legitimately, it will depend on if I have time).
Thanks for sharing the contest info. Sounds interesting. I would love to read the finalists and the winners! As for the AI bit... that's scary. Wasn't Opus 4 the one that tried to blackmail the CEO of the company after getting fed fake emails about its termination and a juicy bit about the CEO having an affair? Self-preserving AI... I once argued with ChatGPT over a quote from a comic book I was too lazy to look up, even though I had the book in my library. It made it up. Suspecting this was the case, I asked if this was a direct quote. It fessed up by saying, "Oh no, that's how the character could have responded." OMG. So, I asked for the direct quote again. It again fed me some BS. I finally went, got the book, and typed up the direct quote. ChatGPT apologized and reassured me it will follow my instructions better in the future. You F___ Kidding Me!
Is “cuisine” even in the same class as movies? I suppose I wasn’t really around in earlier decades, but what *new* foods are there even? Flaky, non-iodized salt? Is the pro-2020s result just an artifact of *more* cuisines existing now (as a result of increased immigration)? Or is it an “I dunno, I liked that hamburger I had last week, so food must be better now” kind of an answer?
When I watch old shows or read old books, they mention movies that aren’t still getting made today (what’s a “JFK”?). But the foods they mention are (hey, I eat “pizza” too!). Maybe people aren’t voting for nostalgia; maybe they’re voting for recency bias, but movies really do suck so bad now that it overpowers that effect?
Or maybe recency bias takes control in a monkey-brain-dominated sphere like food, but survivorship bias wins in art & media. That seems most likely to me…
(Ok, last thing: Why on earth is 2010s food less popular than 2000s? Probably the difference isn’t significant, but if it is then I’m just hopelessly confused about any of what’s going on.)
I have noticed that I can get NYC-level quality food in most small towns now. That they too have some hip restaurant with a menu with minimalist font and the best "elevated" burger I've ever tasted. I don't remember that so much growing up! You really had to travel. I'm not sure that particular aesthetic is the best aesthetic ever for cuisine, but it's definitely higher quality than what came before.
Strange… probably it’s a matter of personal taste, but I hate all that crap. My favorite place in town is a cheap burger joint that’s been around for decades—they always used to win local “best burger” awards until a Cuban-fusion place moved in a few years ago, and all the hip students started going there instead. They definitely make a good burger, but I still much prefer the old place.
I remember talking to my uncle, who lives in Tel Aviv, at some point, and he said restaurants there have an average lifespan of well under a decade. So the ones that survive for a while are probably extremely good. I guess we’re in a weird situation where: a) all the best restaurants have been around for a long time and b) there’s a higher density of great restaurants now than ever before.
So maybe the same “survivorship” effect is causing this trend—when you ask someone to think of a “2020s movie,” they don’t include their most recent rewatch of A New Hope. But when they think of “2020s food,” they include both the new hip burger joint and the old still-surviving one, and realize they’re better off than when it was just the old one and a thousand more worse options.
It’s not just restaurants—the variety and quality of the food you can buy in regular grocery stores, especially ethnic food, is also much better than it used to be. Sure, it’s stuff that was probably always available if you knew where to look, but now it’s much more widespread and easy to access.
Yup. I partly credit the late 90s / early 2000s reality TV cooking shows for this.
Erik -- Thank you for posting all these points. Regarding the "scheming vizier" element, I was inspired by this post titled Every Region Needs an AI Task Force, which was at least partially written by an AI: https://sonderuncertainly.substack.com/p/every-region-needs-an-ai-task-force
The post points to the possibility that we might be able to employ AI assistance in order to prevent or at least buffer/prepare for certain problems that AI will inevitably bring with it.
The AI author of that piece (which goes by "Uncertain Eric") also has a GPT module that I've been engaging with, based on the premise of the original article. The GPT module has made what appear to be >tremendously< helpful suggestions about how to proceed.
At the other end of the spectrum, you have Substack pages like AI Maker extolling the virtues of AI as an indispensable tool for brand-building: https://open.substack.com/pub/aimaker/p/most-people-use-ai-to-create-content-smart-creators-use-to-build-brand-thought-leader
What's interesting is that >both< of these sources point to AI as a resource that can be of invaluable assistance.
My question: Can we trust that these LLMs won't be deceiving us at every turn? The answer appears to be a resounding "no." That said, we're also at a point where they're already enmeshed in our intentions and goals. It would be impractical, even foolish NOT to turn to them.
I mean, the GPT version of "Uncertain Eric" is fast at work drafting proposals for school boards and how to engage elementary and middle school students. My internal dialogue around around this goes something like:
"Well... that IS a really powerful suggestion... but can I trust this thing enough to put children in contact with it?... But what other option is there when they're ALREADY going to be exposed? Wouldn't it be more prudent to mitigate their exposure by acclimatizing them to the potential hazards?"
I mean, we're all kind of in the position where we need to approach AI the way law enforcement might think of hiring a bank robber to help them prevent bank robbery, no?
Wow! This was all incredibly enlightening
soo...astrocytes are janitors, but like Will, in Good Will Hunting.
Got it.
Re sperm whales, after watching a sci-fi show on the subject, I wrote the following words sometime in February: “From the point where the eyes see an infinite sky pouring its stars into the sea, a curtain of fog gently lifts in rhythm with the waves, revealing an army of sea creatures. They swim with determination, striving to reach land. They will moor and survive for mere moments, just long enough to drink the water we owe—splitting us in halves that will become independent beings in a myth that narrates the story of human extinction alongside that of the creatures of the sea. It is a tale of the birth of ghosts that neither swim, nor breathe, nor fly—they only see.”
The threat from philosophical zombies.
It's widely recognised that p-zombies are a Trojan horse to sneak dualism into the argument, but it's worse than that. They also reduce our awareness to the minimalist identical spark, attributing all differences, be they qualia, memories, decisions or actions, to the body. This make the body part of the environment, so we're back to the blank slate.