High Signal Podcasts Evidence ledger
Method
Browse
How the evidence works

Public evidence record

Max Bennett

Published podcast speaker

Claims
19
Episodes
1
Shows
1
Named items
0

Claim ledger

What Max said.

19 transcript-backed records

01 / evaluation

But for semantic memories, we're terrible. And so externalizing that with writing is 1 of the key technologies that enables us to be way smarter than we are in fact, because we now have this external device that enables us to store largely infinite numbers of memories and translate them across generations.

“But for semantic memories, we're terrible. And so externalizing that with writing is 1 of the key technologies that enables us to be way smarter than we are in fact, because we now have this external device that enables us to store largely infinite numbers of memories and translate them across generations.”
Speaker
Max Bennett
Publisher
Machine Learning Street Talk

02 / belief

I think I think there's is a reason why, so much of the neuroscience community, I mean, what I'm saying there is not really novel so much the neuroscience community has sort of rallied around this idea of predictive coding, which is very related to active inference and generative models. Because there's just so much evidence that what's going on in the neocortex, the imagination of things, episodic memories, I mean there's been some good evidence that episodic memory, other words, thinking about past and imagine the future are in fact the same underlying process happening in the neocortex, which is again consistent with this idea that there's a generative model.

“I think I think there's is a reason why, so much of the neuroscience community, I mean, what I'm saying there is not really novel so much the neuroscience community has sort of rallied around this idea of predictive coding, which is very related to active inference and generative models. Because there's just so much evidence that what's going on in the neocortex, the imagination of things, episodic memories, I mean there's been some good evidence that episodic memory, other words, thinking about past and imagine the future are in fact the same underlying process happening in the neocortex, which is again consistent with this idea that there's a generative model.”
Speaker
Max Bennett
Publisher
Machine Learning Street Talk

03 / uncertainty

I think a lot of people I've had this exact sort of argument with a lot of people and 1 counterargument that people have towards this idea, I don't know if I fully agree with it, but it's interesting, is that attention heads really are doing something more magical than we give them credit for, which is they are kind of dynamically rerouting and effectively resetting the network based on the context that the prompt is getting.

“I think a lot of people I've had this exact sort of argument with a lot of people and 1 counterargument that people have towards this idea, I don't know if I fully agree with it, but it's interesting, is that attention heads really are doing something more magical than we give them credit for, which is they are kind of dynamically rerouting and effectively resetting the network based on the context that the prompt is getting.”
Speaker
Max Bennett
Publisher
Machine Learning Street Talk

04 / belief

The most famous form of this, which I think still shows threads in modernity, is from Aristotle where, he sort of took the same kind of ideas that you see in Maclean's triune brain, where, you know, other animals might have these basic instincts.

“The most famous form of this, which I think still shows threads in modernity, is from Aristotle where, he sort of took the same kind of ideas that you see in Maclean's triune brain, where, you know, other animals might have these basic instincts.”
Speaker
Max Bennett
Publisher
Machine Learning Street Talk

05 / belief

There's only, it feels like there's a ranking and only 1 of these is gonna matter. So I think as an example of ways to reduce this sort of status seeking, I see this in sort of business all the time, which is the more you can create an environment where it's not 0 sum, where everyone's pulling their weight and together we all win, the more sort of the best version of humans emerge.

“There's only, it feels like there's a ranking and only 1 of these is gonna matter. So I think as an example of ways to reduce this sort of status seeking, I see this in sort of business all the time, which is the more you can create an environment where it's not 0 sum, where everyone's pulling their weight and together we all win, the more sort of the best version of humans emerge.”
Speaker
Max Bennett
Publisher
Machine Learning Street Talk

06 / belief

You know, sometimes I think about this as like, you know, if me and a group of 20 friends ended up on an island, how much of human and we were the only 20 humans left, not that I think about this all the time.

“You know, sometimes I think about this as like, you know, if me and a group of 20 friends ended up on an island, how much of human and we were the only 20 humans left, not that I think about this all the time.”
Speaker
Max Bennett
Publisher
Machine Learning Street Talk

07 / belief

I'm gonna now render some tests where I could differentiate them, and I'm gonna perform those tests, and then I'm gonna conclude that the world is not is not flat. So so, when 1 says that Chachapiti does not have a world model, I think some people misinterpret that as suggesting that it's just dumbly looking at the statistics.

“I'm gonna now render some tests where I could differentiate them, and I'm gonna perform those tests, and then I'm gonna conclude that the world is not is not flat. So so, when 1 says that Chachapiti does not have a world model, I think some people misinterpret that as suggesting that it's just dumbly looking at the statistics.”
Speaker
Max Bennett
Publisher
Machine Learning Street Talk

08 / belief

The second perhaps more practical reason is I think understanding the evolution of the human brain and the evolution of human intelligence is a key tool in our toolbox to understanding how the brain works and how human intelligence works.

“The second perhaps more practical reason is I think understanding the evolution of the human brain and the evolution of human intelligence is a key tool in our toolbox to understanding how the brain works and how human intelligence works.”
Speaker
Max Bennett
Publisher
Machine Learning Street Talk

09 / belief

Know, Yan Lakun talks a lot about this and are less selfish. And so yeah, I think that's a great opportunity, but there's definitely risk because the second you give an autonomous agent the opportunity to produce its own sub goals, you need to have really rich either constraints or a really well defined reward function or know, 1 ability that I think comes from mentalizing actually is, and this is an idea in alignment research, which is, if you can convince an AI agent to try and do what it thinks the human wants it to do, What you're actually doing is you're requiring it to engage in some form of mentalizing to infer the preferences of the requester and then try and do what is best for that that individual.

“Know, Yan Lakun talks a lot about this and are less selfish. And so yeah, I think that's a great opportunity, but there's definitely risk because the second you give an autonomous agent the opportunity to produce its own sub goals, you need to have really rich either constraints or a really well defined reward function or know, 1 ability that I think comes from mentalizing actually is, and this is an idea in alignment research, which is, if you can convince an AI agent to try and do what it thinks the human wants it to do, What you're actually doing is you're requiring it to engage in some form of mentalizing to infer the preferences of the requester and then try and do what is best for that that individual.”
Speaker
Max Bennett
Publisher
Machine Learning Street Talk

10 / belief

I'm just explaining it after the fact. So so it's definitely the case that humans have that component, but it means there's also another component which we which I would argue is is unique and important, which is our ability to pause, render a plan, and then execute against that plan.

“I'm just explaining it after the fact. So so it's definitely the case that humans have that component, but it means there's also another component which we which I would argue is is unique and important, which is our ability to pause, render a plan, and then execute against that plan.”
Speaker
Max Bennett
Publisher
Machine Learning Street Talk

12 / belief

I think the Thinking about the level of granular prefrontal cortex in the new primate regions as enabling something akin to knowledge, I think is a really wonderful way to look at it.

“I think the Thinking about the level of granular prefrontal cortex in the new primate regions as enabling something akin to knowledge, I think is a really wonderful way to look at it.”
Speaker
Max Bennett
Publisher
Machine Learning Street Talk

13 / belief

However, it is not nearly as, rich because what happens in non human primates is it's primarily grounded in just the actions that I'm taking, which is much less rich than I can share not only the the data of what you see me actually do, but I can also include in the data I'm transferring to you things that happen only in my mind.

“However, it is not nearly as, rich because what happens in non human primates is it's primarily grounded in just the actions that I'm taking, which is much less rich than I can share not only the the data of what you see me actually do, but I can also include in the data I'm transferring to you things that happen only in my mind.”
Speaker
Max Bennett
Publisher
Machine Learning Street Talk

14 / recommendation

It's a great book called The Elephant in the Brain that talks about how much of human behavior can be explained by this sort of status seeking behavior. And the reason why it's so hard to study is because it's what they call, I think, a cognitive taboo or an intellectual taboo, where we don't want to admit it to ourselves.

“It's a great book called The Elephant in the Brain that talks about how much of human behavior can be explained by this sort of status seeking behavior. And the reason why it's so hard to study is because it's what they call, I think, a cognitive taboo or an intellectual taboo, where we don't want to admit it to ourselves.”
Speaker
Max Bennett
Publisher
Machine Learning Street Talk

15 / prediction

, the way this would function is that a group of humans that, are sharing knowledge with each other is gonna survive better than another group of human that's not, and that's how evolution will ensue. This is actually quite controversial in evolutionary biology because that's invoking something called group selection, which now some people the modern incarnation of this, something called multilevel selection, where, there's some consensus that, yes, there are group level effects, where that kind of can impact things.

“, the way this would function is that a group of humans that, are sharing knowledge with each other is gonna survive better than another group of human that's not, and that's how evolution will ensue. This is actually quite controversial in evolutionary biology because that's invoking something called group selection, which now some people the modern incarnation of this, something called multilevel selection, where, there's some consensus that, yes, there are group level effects, where that kind of can impact things.”
Speaker
Max Bennett
Publisher
Machine Learning Street Talk

16 / preference

Well, I think I I do I think history is such a good I love thinking about history as a tool when trying to reason about how people in the future will think about us because I think, you know, we are the people in the future to the past, which is obvious, but a useful tool.

“Well, I think I I do I think history is such a good I love thinking about history as a tool when trying to reason about how people in the future will think about us because I think, you know, we are the people in the future to the past, which is obvious, but a useful tool.”
Speaker
Max Bennett
Publisher
Machine Learning Street Talk

17 / prediction

Doesn't mean it won't work well, but we should be concerned about how well this will generalize to real tasks where we might care about this much more deeply. So for example, with a human, because part of this good evidence to suggest that part of my ability to reason about your mind is because I have a mind that works quite similarly, We are almost bound together by some common mechanistic synergy between the way in which our brains work because our brains are quite similar, which which enables a lot of data efficiency, which is I'm pretty good at predicting what people do.

“Doesn't mean it won't work well, but we should be concerned about how well this will generalize to real tasks where we might care about this much more deeply. So for example, with a human, because part of this good evidence to suggest that part of my ability to reason about your mind is because I have a mind that works quite similarly, We are almost bound together by some common mechanistic synergy between the way in which our brains work because our brains are quite similar, which which enables a lot of data efficiency, which is I'm pretty good at predicting what people do.”
Speaker
Max Bennett
Publisher
Machine Learning Street Talk

18 / evaluation

I think there is absolutely something to the idea that just because you can recognize something so that would be a feed forward network that can observe a stapler alone is insufficient to what most people would mean when they use the word understanding.

“I think there is absolutely something to the idea that just because you can recognize something so that would be a feed forward network that can observe a stapler alone is insufficient to what most people would mean when they use the word understanding.”
Speaker
Max Bennett
Publisher
Machine Learning Street Talk

19 / evaluation

Evolutionary neuroscience, so what do we know about the past brains of humans and the ordered set of modifications of how brains came to be. And then AI, which is how do we ground sort of the highfalutin conceptual discussions about how the brain works in what works in practice, I think, a really important grounding principle because it helps really hold us accountable towards the principles that we think work.

“Evolutionary neuroscience, so what do we know about the past brains of humans and the ordered set of modifications of how brains came to be. And then AI, which is how do we ground sort of the highfalutin conceptual discussions about how the brain works in what works in practice, I think, a really important grounding principle because it helps really hold us accountable towards the principles that we think work.”
Speaker
Max Bennett
Publisher
Machine Learning Street Talk
Search evidence