I’ve read the QM sequence and it really is one of the most important sequences. When I suggest this at meetups and such, people seem to be under the impression that it’s just Eliezer going off topic for a while and totally optional. This is not the case, the QM sequence is used like you said to develop a huge number of later things.
The negative comments from physicists and physics students are sort of a worry (to me as someone who got up to the start of studying this stuff in second-year engineering physics and can’t remember one dot of it). Perhaps it could do with a robustified rewrite, if anyone sufficiently knowledgeable can be bothered.
The negative comments I’ve heard give off a strong scent of being highly motivated - in one case an incredible amount of bark bark bark about how awful they were, and when I pressed for details, a pretty pathetic bite. I’d like to get a physicist who didn’t seem motivated to have an opinion one way or the other to comment.
It would need to be someone who bought MWI—if the sole problem with them is that they endorse MWI then that’s at least academically respectable, and if an expert reading them doesn’t buy MWI then they’ll be motivated to find problems in a way that won’t be as informative as we’d like.
The Quantum Physics Sequence is unusual in that normally, if someone writes 100,000(?) words explaining quantum mechanics for a general audience, they genuinely know the subject first: they have a physics degree, they have had an independent reason to perform a few quantum-mechanical calculations, something like that. It seems to me that Eliezer first got his ideas about quantum mechanics from Penrose’s Emperor’s New Mind, and then amended his views by adopting many-worlds, which was probably favored among people on the Extropians mailing list in the late 1990s. This would have been supplemented by some incidental study of textbooks, Feynman lectures, expository web pages… but nonetheless, that appears to be the extent of it. The progression from Penrose to Everett would explain why he presents the main interpretive choice as between wavefunction realism with objective collapse, and wavefunction realism with no collapse. His prose is qualitative just about everywhere, indicating that he has studied quantum mechanics just enough to satisfy himself that he has obtained a conceptual understanding, but not to the point of quantitative competence. And then he has undertaken to convey this qualitative conceptual understanding to other people who don’t have quantitative competence in the subject, either.
I can recognize all this because I am also an autodidact and I have done comparable things. It’s possible to do this and get away with making only a few incidental mistakes. But it is a very risky thing to do. You run a high risk of fooling yourself and then causing your audience to fool themselves too. This is especially the case in mathematical physics. Literally every day I see people asking questions on physics websites that are premised on wrong assumptions about physical theory. I don’t mean questions where people say “is it really true that...”, I mean questions where the questioner thinks they already understand some topic, and the question proceeds from this incorrect understanding, sometimes quite aggressively in tone (recently observed example).
My opinion about the Sequences is that someone who knows nothing about QM can learn from them, but it’s worth getting a second opinion, even just from Wikipedia, since they present a rather ideological point of view. Also, when you read them, you’re simply not hearing from someone who has used quantum mechanics professionally; you’re hearing from an autodidact who thinks he figured out the gist of the subject—I’d say, very roughly, he gets about 75% of the basics, and the problems are more in what is omitted rather than what is described (e.g. nothing, that I recall, about the role of operators) - and who has decided that one prominent minority faction of opinion among physicists (the many-worlds enthusiasts) are the ones who have correctly discerned the implications of QM for the nature of reality. The fact that he espouses, as the one true interpretation, a point of view that is shared by some genuine physicists, does at least protect him from the accusation of complete folly. Nonetheless, I can tell you—as one autodidact judging another—he’s backed the wrong horse. :-)
If you want an independent evaluation of the Sequences by physicists, I suggest that you post this as a question at Physics Stack Exchange. Ask what people think of them, and whether they can be trusted. There’s a commenter there, Ron Maimon, who is the most readily available approximation to what you want. Maimon is quantitatively competent in all sorts of advanced physics, and he was once a MWI zealot. Now he’s more positivistic, but MWI is still his preferred language for making ontological sense of QM. I would expect him to offer qualified approval of the Sequences, but to make some astute comments regarding content or style.
Since it is a forum where everyone gets a chance to answer the question, with the best replies then being voted up by the readership, of course such a question would also lead to responses by people who don’t believe MWI. But this is the quickest way to conduct the experiment you suggest.
Result from Ron Maimon’s review of the QM sequence:
I skimmed a majority of the articles, and there are no glaring errors that I could find, but there is an unnecessary verbosity which is best eliminated by reading a terser introduction to the Everett interpretation. The amount of text that is presented is not commensurate with the amount of insight.
That seems less valuable. The QM sequences are largely there to set out what is supposed to be an existing, widespread understanding of QM. No such understanding exists for AI risk.
This was a big concern I had reading it. Much of it made sense to me, as someone who has had formal education in basic quantum, and some of it felt very illuminating(the waveform-addition stuff in particular was taught far better than my quantum prof ever managed), but I’m always skeptical of people claiming Truth of a controversy in a highly technical field with no actual training in that field. I’ve always preferred many-worlds, but I would never claim it is the sole truth in the sort of way that EY did.
It’s not clear how to compare said risk—“quantum” is far more widely abused—but the creationist AI researcher suggests AI may be severely prone to the problem. Particularly as humans are predisposed to think of minds as ontologically basic, therefore pretty simple, therefore something they can have a meaningful opinion on, regardless of the evidence to the contrary.
What, you mean the part where we’re discussing a field that’s still highly theoretical, with no actual empirical evidence whatsoever, and then determining that it is definitely the biggest threat to humanity imaginable and that anyone who doesn’t acknowledge that is a fool?
A poll would be good.
I’ve read the QM sequence and it really is one of the most important sequences. When I suggest this at meetups and such, people seem to be under the impression that it’s just Eliezer going off topic for a while and totally optional. This is not the case, the QM sequence is used like you said to develop a huge number of later things.
The negative comments from physicists and physics students are sort of a worry (to me as someone who got up to the start of studying this stuff in second-year engineering physics and can’t remember one dot of it). Perhaps it could do with a robustified rewrite, if anyone sufficiently knowledgeable can be bothered.
The negative comments I’ve heard give off a strong scent of being highly motivated - in one case an incredible amount of bark bark bark about how awful they were, and when I pressed for details, a pretty pathetic bite. I’d like to get a physicist who didn’t seem motivated to have an opinion one way or the other to comment.
It would need to be someone who bought MWI—if the sole problem with them is that they endorse MWI then that’s at least academically respectable, and if an expert reading them doesn’t buy MWI then they’ll be motivated to find problems in a way that won’t be as informative as we’d like.
The Quantum Physics Sequence is unusual in that normally, if someone writes 100,000(?) words explaining quantum mechanics for a general audience, they genuinely know the subject first: they have a physics degree, they have had an independent reason to perform a few quantum-mechanical calculations, something like that. It seems to me that Eliezer first got his ideas about quantum mechanics from Penrose’s Emperor’s New Mind, and then amended his views by adopting many-worlds, which was probably favored among people on the Extropians mailing list in the late 1990s. This would have been supplemented by some incidental study of textbooks, Feynman lectures, expository web pages… but nonetheless, that appears to be the extent of it. The progression from Penrose to Everett would explain why he presents the main interpretive choice as between wavefunction realism with objective collapse, and wavefunction realism with no collapse. His prose is qualitative just about everywhere, indicating that he has studied quantum mechanics just enough to satisfy himself that he has obtained a conceptual understanding, but not to the point of quantitative competence. And then he has undertaken to convey this qualitative conceptual understanding to other people who don’t have quantitative competence in the subject, either.
I can recognize all this because I am also an autodidact and I have done comparable things. It’s possible to do this and get away with making only a few incidental mistakes. But it is a very risky thing to do. You run a high risk of fooling yourself and then causing your audience to fool themselves too. This is especially the case in mathematical physics. Literally every day I see people asking questions on physics websites that are premised on wrong assumptions about physical theory. I don’t mean questions where people say “is it really true that...”, I mean questions where the questioner thinks they already understand some topic, and the question proceeds from this incorrect understanding, sometimes quite aggressively in tone (recently observed example).
My opinion about the Sequences is that someone who knows nothing about QM can learn from them, but it’s worth getting a second opinion, even just from Wikipedia, since they present a rather ideological point of view. Also, when you read them, you’re simply not hearing from someone who has used quantum mechanics professionally; you’re hearing from an autodidact who thinks he figured out the gist of the subject—I’d say, very roughly, he gets about 75% of the basics, and the problems are more in what is omitted rather than what is described (e.g. nothing, that I recall, about the role of operators) - and who has decided that one prominent minority faction of opinion among physicists (the many-worlds enthusiasts) are the ones who have correctly discerned the implications of QM for the nature of reality. The fact that he espouses, as the one true interpretation, a point of view that is shared by some genuine physicists, does at least protect him from the accusation of complete folly. Nonetheless, I can tell you—as one autodidact judging another—he’s backed the wrong horse. :-)
If you want an independent evaluation of the Sequences by physicists, I suggest that you post this as a question at Physics Stack Exchange. Ask what people think of them, and whether they can be trusted. There’s a commenter there, Ron Maimon, who is the most readily available approximation to what you want. Maimon is quantitatively competent in all sorts of advanced physics, and he was once a MWI zealot. Now he’s more positivistic, but MWI is still his preferred language for making ontological sense of QM. I would expect him to offer qualified approval of the Sequences, but to make some astute comments regarding content or style.
Since it is a forum where everyone gets a chance to answer the question, with the best replies then being voted up by the readership, of course such a question would also lead to responses by people who don’t believe MWI. But this is the quickest way to conduct the experiment you suggest.
Excellent idea—done. Thank you!
Result from Ron Maimon’s review of the QM sequence:
(more at the link from ciphergoth’s post)
You could also ask for an independent evaluation of AI risks here.
That seems less valuable. The QM sequences are largely there to set out what is supposed to be an existing, widespread understanding of QM. No such understanding exists for AI risk.
So why isnt that pointed out anywhere? EY seems oddly oblivious to his potential—indeed likely -- limitations as an autodictat.
This was a big concern I had reading it. Much of it made sense to me, as someone who has had formal education in basic quantum, and some of it felt very illuminating(the waveform-addition stuff in particular was taught far better than my quantum prof ever managed), but I’m always skeptical of people claiming Truth of a controversy in a highly technical field with no actual training in that field. I’ve always preferred many-worlds, but I would never claim it is the sole truth in the sort of way that EY did.
What reason do I have to believe that this risk isn’t even stronger when it comes to AI?
It’s not clear how to compare said risk—“quantum” is far more widely abused—but the creationist AI researcher suggests AI may be severely prone to the problem. Particularly as humans are predisposed to think of minds as ontologically basic, therefore pretty simple, therefore something they can have a meaningful opinion on, regardless of the evidence to the contrary.
What, you mean the part where we’re discussing a field that’s still highly theoretical, with no actual empirical evidence whatsoever, and then determining that it is definitely the biggest threat to humanity imaginable and that anyone who doesn’t acknowledge that is a fool?
This is one of the classic straw men, adaptable to any purpose.
Mockery is generally rather adaptable, yes.
I suspect a lot of it is “oh dear, someone saying ‘quantum’” fatigue. But that sounds a plausible approach.