The inappropriate laughs reminded me to this recording of a speech from David Foster Wallace: This Is Water.
Is unwarranted, incredulous laughter a sign of a too big cognitive distance between the speaker and the audience? I.e., if the speaker is too smart or too dumb compared to his listeners, are the latter going to find the whole situation so disorienting as to be funny?
I didn’t see the laughs as inappropriate; they appeared at moments which would, in a normal TED talk describing a problem, be queued as jokes. I even read them that way, but it was a short notice unpolished talk, so there was no time for strategically pausing to allow the laughter to express.
Eliezer was clearly being humorous at a few points.
People tend to laugh at things that have become to worrying to ignore, but that they do not wish to act upon, in order to diffuse the discomfort and affirm that this is ridiculous and that they are safe.
I think Eliezer being invited to TED, and people listening, most applauding, many standing up, and a bunch laughing, is a significant step up from being ignored. But it is still far from being respected and followed. (And, if we believe the historic formula, in between, you would expect to encounter active, serious opposition. The AI companies that Eliezer is opposing initially pretending he did not exist. Then, they laughed. That won’t fluently transition to agreeing. Before they will make changes, they will use their means to silence him.)
I think it was less a matter of intelligence differential, but that the talk presupposed too much in specific arguments or technical details the audience simply would not have known (because Eliezer has been speaking to people who have listened to him before so much that he seems disconnected from where the general public is at, so I could fill in the dots, but I think for the audience, there were often leaps that left them dubious—you could see in the Q&A they where still at the boxing the AI that does not have a body stage), and would have profited from a different tone with more authority signalling (eye contact, slow deep voice, seeming calm/resigned, grieving or leading rather than anxious/overwhelmed), specific examples (e.g. on take-over scenarios) and repeating basic arguments (e.g. why AIs might want resources). This way, it had hysteric vibes, which came together with content the audience does not want to believe to create distance. The hysteric vibes are justified, terribly so, and I do not know if anyone who understands the why could suppress them in such an anxiety inducing situation, but that doesn’t stop them from being damaging. (Reminds me of the scene in “don’t look up” where the astrophysicst has a meltdown over the incoming asteroid, and is hence dismissed on the talk show. You simultaneously realise that she has every right to yell “We are all going to die!” at this point, and you would, too, and yet know this is when she lost the audience.)
In that vein, no idea if I could have done better on short notice; and de facto, I definitely didn’t, and it is so much easier to propose something better in hindsight from the safety of my computer screen. Maybe if he had been more specific, people would have gotten hung up on whether that specific scenario can be disproven. It is brave to go out there, maybe some points will stick, and even if people dismiss him, maybe they will be more receptive to something similar that feels less dangerous in the moment later. I respect him for trying this way, must have been scary as hell. Sharing a justified fear that has defined your life in such a brief time span in front of people who laugh frankly sounds awful.
The inappropriate laughs reminded me to this recording of a speech from David Foster Wallace: This Is Water.
Is unwarranted, incredulous laughter a sign of a too big cognitive distance between the speaker and the audience? I.e., if the speaker is too smart or too dumb compared to his listeners, are the latter going to find the whole situation so disorienting as to be funny?
I didn’t see the laughs as inappropriate; they appeared at moments which would, in a normal TED talk describing a problem, be queued as jokes. I even read them that way, but it was a short notice unpolished talk, so there was no time for strategically pausing to allow the laughter to express.
Eliezer was clearly being humorous at a few points.
Some of it is likely nervous laughter but certainly not all of it.
People tend to laugh at things that have become to worrying to ignore, but that they do not wish to act upon, in order to diffuse the discomfort and affirm that this is ridiculous and that they are safe.
I think Eliezer being invited to TED, and people listening, most applauding, many standing up, and a bunch laughing, is a significant step up from being ignored. But it is still far from being respected and followed. (And, if we believe the historic formula, in between, you would expect to encounter active, serious opposition. The AI companies that Eliezer is opposing initially pretending he did not exist. Then, they laughed. That won’t fluently transition to agreeing. Before they will make changes, they will use their means to silence him.)
I think it was less a matter of intelligence differential, but that the talk presupposed too much in specific arguments or technical details the audience simply would not have known (because Eliezer has been speaking to people who have listened to him before so much that he seems disconnected from where the general public is at, so I could fill in the dots, but I think for the audience, there were often leaps that left them dubious—you could see in the Q&A they where still at the boxing the AI that does not have a body stage), and would have profited from a different tone with more authority signalling (eye contact, slow deep voice, seeming calm/resigned, grieving or leading rather than anxious/overwhelmed), specific examples (e.g. on take-over scenarios) and repeating basic arguments (e.g. why AIs might want resources). This way, it had hysteric vibes, which came together with content the audience does not want to believe to create distance. The hysteric vibes are justified, terribly so, and I do not know if anyone who understands the why could suppress them in such an anxiety inducing situation, but that doesn’t stop them from being damaging. (Reminds me of the scene in “don’t look up” where the astrophysicst has a meltdown over the incoming asteroid, and is hence dismissed on the talk show. You simultaneously realise that she has every right to yell “We are all going to die!” at this point, and you would, too, and yet know this is when she lost the audience.)
In that vein, no idea if I could have done better on short notice; and de facto, I definitely didn’t, and it is so much easier to propose something better in hindsight from the safety of my computer screen. Maybe if he had been more specific, people would have gotten hung up on whether that specific scenario can be disproven. It is brave to go out there, maybe some points will stick, and even if people dismiss him, maybe they will be more receptive to something similar that feels less dangerous in the moment later. I respect him for trying this way, must have been scary as hell. Sharing a justified fear that has defined your life in such a brief time span in front of people who laugh frankly sounds awful.