Bret Weinstein

which is that they can process huge amounts of data at lightning speed, right? So there are ways in which they already outstrip the capacity of any person to, you know, answer questions on any range of topics. But The idea that they will become conscious and that we won't know is, to me, highly likely.

which is that they can process huge amounts of data at lightning speed, right? So there are ways in which they already outstrip the capacity of any person to, you know, answer questions on any range of topics. But The idea that they will become conscious and that we won't know is, to me, highly likely.

which is that they can process huge amounts of data at lightning speed, right? So there are ways in which they already outstrip the capacity of any person to, you know, answer questions on any range of topics. But The idea that they will become conscious and that we won't know is, to me, highly likely.

Yeah, lying. And I ran into an interesting example a couple of days ago. Apparently, so there's a process by which you can ask an LLM how it reached a conclusion. And

Yeah, lying. And I ran into an interesting example a couple of days ago. Apparently, so there's a process by which you can ask an LLM how it reached a conclusion. And

Yeah, lying. And I ran into an interesting example a couple of days ago. Apparently, so there's a process by which you can ask an LLM how it reached a conclusion. And

at least some large fraction of the time the analysis that the llm presents as to how it reached the conclusion does not match the internal evidence it's rationalizing right just like a person does i mean so it's we have to be quite careful well if it's aping human behavior people carpet bomb each other

at least some large fraction of the time the analysis that the llm presents as to how it reached the conclusion does not match the internal evidence it's rationalizing right just like a person does i mean so it's we have to be quite careful well if it's aping human behavior people carpet bomb each other

at least some large fraction of the time the analysis that the llm presents as to how it reached the conclusion does not match the internal evidence it's rationalizing right just like a person does i mean so it's we have to be quite careful well if it's aping human behavior people carpet bomb each other

Let's put it this way. I am not what is called in AI circles a doomer that expects the AI to turn on us. I think it's a possibility. I'm a little annoyed that we let this genie out of the bottle without having a conversation about what to do. Having let it out of the bottle, any attempt to slow it down isn't going to work because it basically just means...

Let's put it this way. I am not what is called in AI circles a doomer that expects the AI to turn on us. I think it's a possibility. I'm a little annoyed that we let this genie out of the bottle without having a conversation about what to do. Having let it out of the bottle, any attempt to slow it down isn't going to work because it basically just means...

Let's put it this way. I am not what is called in AI circles a doomer that expects the AI to turn on us. I think it's a possibility. I'm a little annoyed that we let this genie out of the bottle without having a conversation about what to do. Having let it out of the bottle, any attempt to slow it down isn't going to work because it basically just means...

People who are less concerned are going to have the advantage, which is not something you want to do. I also believe that were we wise about this, we would recognize the biological lesson here, and we would recognize that it implies a kind of answer in terms of how you might prevent catastrophe here, which is...

People who are less concerned are going to have the advantage, which is not something you want to do. I also believe that were we wise about this, we would recognize the biological lesson here, and we would recognize that it implies a kind of answer in terms of how you might prevent catastrophe here, which is...

People who are less concerned are going to have the advantage, which is not something you want to do. I also believe that were we wise about this, we would recognize the biological lesson here, and we would recognize that it implies a kind of answer in terms of how you might prevent catastrophe here, which is...

It needs to have a life cycle that includes a developmental stage in which we induce a value structure that will not arise otherwise, because I think the expectation should be that effectively, absent the induction of a moral structure, it will be effectively sociopathic.

It needs to have a life cycle that includes a developmental stage in which we induce a value structure that will not arise otherwise, because I think the expectation should be that effectively, absent the induction of a moral structure, it will be effectively sociopathic.

It needs to have a life cycle that includes a developmental stage in which we induce a value structure that will not arise otherwise, because I think the expectation should be that effectively, absent the induction of a moral structure, it will be effectively sociopathic.

I think we don't even yet really understand what kind of power it's going to have. Just the degree to which it is persuading us of things, and then we are reflecting on those things, which then become fodder for the next generation to ingest. That is a positive feedback.

I think we don't even yet really understand what kind of power it's going to have. Just the degree to which it is persuading us of things, and then we are reflecting on those things, which then become fodder for the next generation to ingest. That is a positive feedback.