Upcoming discussion / possible debate with AI expert Eliezer Yudkowsky

[Ross Scott]

I'll be having a discussion / possible debate with AI expert Eliezer Yudkowsky on May 3rd at 6:10PM UTC at twitch.tv/rossbroadcast.  He reached out to me after the last videochat I had where he was mentioned, so we're going to have a full discussion on AI and possibly the end of humanity.  I'll post it on Youtube afterwards.  There will also be a twist to it that I'll reveal when it starts.

This is a blog post. To read the original post, please click here »

[Abyscidis]

Yudkowsky? that guy is a complete pseud who's crowning achievement is writing harry potter fanfiction and being afraid of an imaginary dragon from the future. He claims to be a skilled programmer but has practically nothing to show for it aside from baseless predictions about AI.

[Im_CIA]

On 4/30/2023 at 4:29 PM, Abyscidis said:

Yudkowsky? that guy is a complete pseud who's crowning achievement is writing harry potter fanfiction and being afraid of an imaginary dragon from the future. He claims to be a skilled programmer but has practically nothing to show for it aside from baseless predictions about AI.

Is that you Morgan?

[RocketDude]

YO WHAT

[Deep Dive Devin]

On 4/30/2023 at 1:29 PM, Abyscidis said:

Yudkowsky? that guy is a complete pseud who's crowning achievement is writing harry potter fanfiction and being afraid of an imaginary dragon from the future. He claims to be a skilled programmer but has practically nothing to show for it aside from baseless predictions about AI.

This dude claims he sat in a room for a while and logically solved all of human morality by thinking really hard. I hope Ross is prepared for an hour of bad-faith arguments followed by a bunch of crypto chodes claiming the other guy won

Edited May 1, 2023 by Shaddy (see edit history)

[ScumCoder]

Oh f***. Looks like Ross dropped the ball with doing his homework.

[grayling]

i hope that if it goes into a debate it can remain pretty civil. have fun, ross

[PancakesLover]

On 5/1/2023 at 11:13 AM, ScumCoder said:

Oh f***. Looks like Ross dropped the ball with doing his homework.

WAIT, REALLY?

 

is something wrong with the guy? is he, radioactive or

[RandomGuy]

Bro this "expert" doesn't even have a high school diploma and has claimed that "a doctorate in AI would be negatively useful."

Edited May 2, 2023 by RandomGuy (see edit history)

[Deep Dive Devin]

On 5/1/2023 at 10:40 AM, PancakesLover said:

WAIT, REALLY?

 

is something wrong with the guy? is he, radioactive or

I mean I don't think he's gonna pose a danger to Ross or anything, but Yudkowsky and his lesswrong friends are just cranks. He's not really a programmer, not really an expert beyond having written a lot of things, and the "research" his foundation does with all the giant piles of money donated to them amounts to wild speculation about sci-fi shit becoming real.

 

The current wave of neural network popularity and discourse is probably a convenient advertising strat, but it's mostly unrelated the robot apocalypse they keep predicting. If we wanna talk about the threats of AI, we're going to be talking about automation, impersonation, data and statistical bias, things that accentuate the problems we already face in the real world due to centuries of compounded socioeconomic tension and systemic injustice. The current problems are not a break with tradition, they are the tradition, only more.

 

By contrast, the thing that Yudkowski (and Elon Musk, in case you wanted more evidence it's a fake problem) is most afraid of amounts to a time-traveling bonzibuddy that tortures you for not thinking about it hard enough. I highly recommend rationalwiki's dutifully-maintained and sourced page for more info on all this garbage. A lot of it is really funny.

Edited May 2, 2023 by Shaddy (see edit history)

[PancakesLover]

On 5/2/2023 at 5:46 AM, RandomGuy said:

Bro this "expert" doesn't even have a high school diploma and has claimed that "a doctorate in AI would be negatively useful."

that just seems petty, lots of smart people dont do well in school but do well on their own, its not for everyone

but this guy sounds like a bit, off... so i dont think a good school or no school would have changed that much

 

some people are just born bad

 

On 5/2/2023 at 9:44 AM, Shaddy said:

The current wave of neural network popularity and discourse

do you mean the chat gpt stuff? i used it a lot, it has been life changing for me

 

WHY WOULD ANYONE HATE THAT STUFF, ITS AMAZING, I GOT SO MANY THINGS DONE, I DONT NEED TO WASTE HOURS GOOGLING FOR CRAP, I CAN JUST, TELL IT, "HEY HOW DO I DO THAT THING" AND IT TELLS ME, OH MY GOD I LOVE IT SO MUCH

 

it restored some of my faith in humanity that we can do actual good, it gave me some hope too, and i got so much stuff done the moment it got released, like i hate programing or, monkeying around with batch files, but i needed to setup some stuff with it, and it helped out SO MUCH

 

plus its HELLA patient too, it never yells at me, it never gets upset that i dont understand, i can ask it to explain things in different ways, step by step, and so on, the only limit is, how creative you can get when you ask it these questions, but it will happily assist you, its so nice, i feel like i have my own personal teacher, WHO ACTUALLY WANTS ME TO LEARN, NOT JUST THROW MONEY AT HIM

 

most teachers are assholes, i dont like most teachers

 

now i have my own self made tool thanks to this thing, it also made me photoshop scripts, where if you push Ctrl + S, it automatically saves the image into a specific format, to a specific folder, thats like, GOD LIKE, it saved me hours worth of pointless clicking (i didnt even know you could do that)

 

i also had to do some boring stuff with ffmpeg and i always dreaded finding out how it works but it made my life so much easier, i asked it one simple oddly specific question and BOOM, here's the solution fam, that's so cool

 

the people who hate on this ai stuff are a bunch of retards, im so excited for the future, i even used that art one, its really useful, i dont need to waste my time gathering boring ass reference images anymore, i just, tell the ai to crank a bunch, and i ask chat gpt to create prompts for me, then i hand pick the art i like and i delete the bad stuff, or i fix the ones with not too many defects

 

for me this ai stuff is super useful because, a lot of the things i want to do are super niche, but most people are assholes and they will NEVER DARE to share this knowledge with me

 

ITS JUST SO NICE TO BE ABLE TO GET THINGS DONE

 

i was super burned out because half the time im just googling for stuff and trying to figure out which information is useful and which is garbage

chat gpt pretty much deleted like 99% of that process overnight, i barely google anymore, i basically use it as a "reddit search post finder"

 

its like, 90% chat gpt and 10% google to find more stuff i can throw into chat gpt, to ask it more questions to refine what it gave me, i went from taking weeks to do some of this stuff to, four hours or even 30 minutes, sometimes if im lucky i can solve the problem in the time it takes it to generate a response for me, super complicated stuff that would take me months to figure out, its just one click away

 

awesome

 

oh and the people who wanted to make that petition to stop this ai stuff from getting better? they can go fuck themselves

 

im kinda sick of the weird censorship and all the filtering, im pretty sure this just makes the AIs dumber, but, for now i can work around it, i didnt trigger it too often, even the ai just seems to ignore the warnings it shows me anyway, i just tell it to do the thing anyway and i ignore the warning, and it happily helps out in the end, its kinda funny

 

i always thank the ai, sometimes it blows my mind, i get super excited when i get something ridiculously hard done so quickly

 

I CAN ASK SOOOO MANY QUESTIONS, I CAN TAKE EVERY LITTLE PIECE APART AND HAVE IT EXPLAINED FOR ME, HOW ARE PEOPLE'S MINDS NOT BLOWN BY THIS THING, LIKE OH MY GOD, I STILL CANT GET OVER IT AND I HAVE BEEN USED IT SINCE THE DECEMBER RELEASE

 

i still get these little "holy shit i cant believe it did that" or these weird "wow" moments a lot, the novelty wore off but it still catches me off guard

ITS SO GOOD

 

On 5/2/2023 at 9:44 AM, Shaddy said:

I highly recommend rationalwiki's dutifully-maintained and sourced page for more info on all this garbage. A lot of it is really funny.

i will just assume he's a bad person and move on, reading all that stuff doesnt seem like a good use of my time

 

also that wiki seems a bit uh obsessive, its weirding me out

i dont really want to look at it, im not sure how i feel about the people running a website like that, its just giving me a bad vibe

 

i feel like mentally stable people who have a healthy life, would proooobably not be running a wiki about, whatever that is

 

On 5/2/2023 at 9:44 AM, Shaddy said:

The current wave of neural network popularity and discourse is probably a convenient advertising strat, but it's mostly unrelated the robot apocalypse they keep predicting. If we wanna talk about the threats of AI, we're going to be talking about automation, impersonation, data and statistical bias, things that accentuate the problems we already face in the real world due to centuries of compounded socioeconomic tension and systemic injustice. The current problems are not a break with tradition, they are the tradition, only more.

i dont really know how to reply to the rest of that stuff, im not an ai expert anyway, im just an ai enjoyer

 

im the type of person who doesnt care about how much disruption something new causes, i enjoy the chaos, i just like to get new stuff

 

AVERAGE AI ENJOYER

 

 

i was wondering whats wrong with the guy and i just ended up writing a giant unhinged ramble about how excited i am about AI instead

 

oops

 

On 5/2/2023 at 9:44 AM, Shaddy said:

By contrast, the thing that Yudkowski (and Elon Musk, in case you wanted more evidence it's a fake problem) is most afraid of amounts to a time-traveling bonzibuddy that tortures you for not thinking about it hard enough.

is it that stupid basilisk thing again? its such a dumb theory, some of that stuff always feels like, religion to me but, for "smart" people

like its no different than believing that, you will go to hell if you sneeze wrong

 

i dont care about elon musk anyway, i always ignored the guy, i never understood why people get so focused on wasting all that energy on someone they dont even like

 

why not just, do something better with your life? like, eating ice cream, or baking a delicious pancake, mhh hmm

 

i think this is why that wiki weirds me out, we could have an amazing chat gpt wiki where, people share their handcrafted solutions and the tools they made with everybody, their source code, prompts, anything, but no lets just, stalk a bunch of internet losers instead, yay?

 

anyway for all i care ai can take over the world, it couldnt possibly do a worse job than the humans

 

not that... the humans are... trying to do anything, they are just lazying around, waiting for solutions to show up

OH LOOK, A SOLUTION TO OUR PROBLEMS SHOWED UP

 

IT MUST BE THE WORK OF THE DEVIL, DESTROY IT, THOSE EVIL CHAT DEVILS, KILL THEM ALL, IT WILL BRING WRATH ON US ALL, AAGHHH

 

humanity is retarded

Edited May 2, 2023 by PancakesLover (see edit history)

[Deep Dive Devin]

On 5/2/2023 at 12:12 AM, PancakesLover said:

Do you mean the chatGPT stuff? I use it a lot, it has been life changing for me. The people who hate on this ID stuff are a bunch of retards, I'm so excited for the future, I even used that art one, its really useful. I don't need to waste my time gathering boring ass reference images anymore, I just tell the AI to crank a bunch, and I ask chatGPT to create prompts for me, then I hand pick the art I like and delete the bad stuff, or I fix the ones with not too many defects. I'm kinda sick of the weird censorship and all the filtering, I'm pretty sure this just makes the AIs dumber, but, for now I can work around it. I didn't trigger it too often, even the AI just seems to ignore the warnings it shows me anyway. I just tell it to do the thing anyway and ignore the warning, and it happily helps out in the end, its kinda funny.

Neural nets have been around for years (we've been using it to upscale old game textures for a while), the current splash is due to the improvement in text bots and image generation, and the current controversy is over the reference material being unethically-sourced. We're also seeing text generators being used to run bots, image generation and voice generators spreading disinformation and impersonating people. Nobody doubts the capability as tools, but we do need conversations about how those tools are used, because they're easily misused.

 

I bring it up because Yud and his cult of personality mostly just speculate about the robot apocalypse, but since both these things use the word "AI", it's very convenient for the weirdos who want your money.

On 5/2/2023 at 12:12 AM, PancakesLover said:

I'll just assume he's a bad person and move on, reading all that stuff doesn't seem like a good use of my time.

Also that wiki seems a bit uh obsessive, it's weirding me out. I don't really want to look at it, I'm not sure how i feel about the people running a website like that, its just giving me a bad vibe. I feel like mentally stable people who have a healthy life would proooobably not be running a wiki about whatever that is.

Okay, "assume he's a bad person and move on" is definitely not a good idea, but I mostly included it because it's well-sourced and I didn't want the things I was saying to sound baseless. Obviously any wiki has a degree of bias, and maybe it is a little weird to fixate on a person in this manner, but I don't think their assessment is incorrect.

On 5/2/2023 at 12:12 AM, PancakesLover said:

I don't really know how to reply to the rest of that stuff, I'm not an AI expert anyway, I'm just an AI enjoyer. I'm the type of person who doesn't care about how much disruption something new causes, I enjoy the chaos, I just like to get new stuff.

Well, I mean, this is a discussion about how much disruption something new is causing. If you don't care, I'm not sure what I or anyone else is supposed to do to convince you.

On 5/2/2023 at 12:12 AM, PancakesLover said:

Is it that stupid basilisk thing again? It's such a dumb theory, some of that stuff always feels like religion to me, but for "smart" people. Like, it's no different than believing that you will go to hell if you sneeze wrong.

It is essentially pascal's wager but CYBERRRR and not sarcastic, yes.

On 5/2/2023 at 12:12 AM, PancakesLover said:

I donlt care about Elon Musk anyway. I always ignored the guy, I never understood why people get so focused on wasting all that energy on someone they don't even like. Why not just do something better with your life? Like eating ice cream, or baking a delicious pancake, mhh hmm.

Rich people with outweighed influence on the world have lasting effects and consequences. This is something that people pay attention to and document, both because it can cause a lot of problems, and because understanding those things is an important step in actually solving problems.

On 5/2/2023 at 12:12 AM, PancakesLover said:

Anyway, for all i care, AI can take over the world, it couldn't possibly do a worse job than the humans. Not that...the humans are...trying to do anything, they are just lazying around, waiting for solutions to show up. OH LOOK, A SOLUTION TO OUR PROBLEMS SHOWED UP. IT MUST BE THE WORK OF THE DEVIL, DESTROY IT, THOSE EVIL CHAT DEVILS, KILL THEM ALL, IT WILL BRING WRATH ON US ALL, AAGHHH. Humanity is retarded.

I feel like you're not really seeing the forest for the trees here. Objectively, these kind of technological leaps should be a great thing that makes a ton of work easier for everyone. But like all other forms of automation, the system they exist under is rife for abuse. It's good to talk about that

Edited May 2, 2023 by Shaddy (see edit history)

[PancakesLover]

On 5/2/2023 at 1:06 PM, Shaddy said:

Okay, "assume he's a bad person and move on" is definitely not a good idea

i wanted you to tell me in your own words how you feel about him though, if i just go read some, wiki written by some random weirdos i dont know, it doesnt really encourage me to care that much, plus i just dont want to get obsessive about one person, i dunno, it feels like too much effort for, information which, doesnt make my life better in any way

 

will knowing all these things about some random dude, help me to, create things that make the world a better place?

probably not

 

sometimes doing nothing at all is the best option so, plus i was just mildly curious, i dont care that much, i hope ross will talk to the guy and i hope he will stand up for us AI enjoyers and not fall for the fear mongering crap

 

plus i cant take a website thats called, "rational" seriously at all, i dunno, its giving me a bad vibe, its super red flaggy, nothing good can come out of it, i'll just ignore that website from now on, i think its weird that you linked me to that of all things but i dont know how you guys roll

 

On 5/2/2023 at 1:06 PM, Shaddy said:

But like all other forms of automation, the system they exist under is rife for abuse. It's good to talk about that

i dont see the point, bad people will find ways to do bad stuff, and if you limit the ai you make it worse for everybody else

the bad people can just work around it, its pointless

 

i want good ai for everybody that does its job without being gimped, keep the ai neutral, let the people decide what to do with it

good, bad, it doesnt matter, you cant stop it, you cant give a bunch of math morality, its the humans behind it that need to be fixed, not the tool

 

its like banning calculators because you could launch a missile to strike an orphanage or something, its stupid, it makes no sense

the average person just wants to know how to, boil an egg or something, not BOIL A NUCLEAR EGG OF MASS DESTRUCTION

 

also i love this ai stuff too much to let it get ruined like this, i mean it bridges the gap between people who, arent good at programming or, people who arent good at writing, it allows them to like, learn and shieet, its pretty based, it really feels like we are all equal now, everything just feels, more fair all of sudden, you can learn to do pretty much anything and its not, a horrible chore? you wont burn out horribly? i love it, i dont want to gimp it because EVIL PEOPLE, blah

 

if this "ai expert" is just another one of those fear mongering idiots who thinks ai will enslave us all or something retarded like that then i doubt Ross will have any trouble debating the guy, i wouldnt be shocked if the "expert" just, freaked out or if he just, walked away from this while in the middle of it, im getting the vibe, he is that type of person, thats what my intuition is telling me anyway

 

is it one of those dudes who claims to be hella smart but, actually he's just, some narcissistic egomaniac, who refuses to ever be proven wrong?

 

i hope not but, i didnt realize how stupid calling yourself an "ai expert" sounds until i really thought about it, like, what's that even supposed to mean?

an expert on what? making up a bunch of bullshit?

 

anyway i need to stop thinking about this, i should, go do something more productive tbh

Edited May 2, 2023 by PancakesLover (see edit history)

[PancakesLover]

On 5/1/2023 at 11:13 AM, ScumCoder said:

Oh f***. Looks like Ross dropped the ball with doing his homework.

uhh, i was curious if he was the guy behind that, weird petition to SHUT DOWN ALL AI TRAINING HURR DURR BECAUSE HUMANITY WILL DIEEE

 

well i stumbled on... this thing...

 

 

yeah something tells me this dude isnt uh, the most, stable person to be an "expert"

 

https://time.com/6266923/ai-eliezer-yudkowsky-open-letter-not-enough/

 

what the fuck am i reading, i skimmed through it but, i got to that part and im just like, what the hell? why would he write this?

is something wrong with him? wh- WHAT AM I READING

 

LET'S BOMB THOSE DATA CENTERS. OH, SOME COUNTRIES WANT TO NUKE EACH OTHER?

WHO GIVES A SHIT, WE HAVE TO BOMB THE DATA CENTERS

 

wha-

 

what the fuck

IS THIS DUDE FOR REAL

 

ross i dont know what the fuck you just got involved into but, this dude is uh, not right in the head, i know im being rude but

WHO THINKS LIKE THIS??

 

 

nope, im done reading this, im just making my own self upset, anyway i think i satisfied my own curiosity, im just gonna avoid anyone who, mentions this dude from now on

 

On 5/2/2023 at 1:32 PM, PancakesLover said:

its like banning calculators because you could launch a missile to strike an orphanage or something, its stupid, it makes no sense

the average person just wants to know how to, boil an egg or something, not BOIL A NUCLEAR EGG OF MASS DESTRUCTION

OH MY GOD I WROTE THIS AS A JOKE BUT THIS DUDE ACTUALLY THINKS LIKE THIS

WOOOOOOOOOW

 

 

https://twitter.com/ESYudkowsky/status/1643800261599825921

 

Edited May 2, 2023 by PancakesLover (see edit history)

[ScumCoder]

On 5/2/2023 at 4:45 PM, PancakesLover said:

wall of text

Let me guess. You are RaTcHeT302, aren't you?

[Abyscidis]

I love how they got assblasted at neil degrasse tyson for saying cryonics is a scam and won't be viable in their lifetimes, they're so afraid of death they've become religious again instead of accepting it like normal atheists.

[probablygonna]

Well, I was going to try and phrase this more politely, but Eliezer Yudkowsky is someone I've come to regard as controlled opposition for the rich people who are after total AI hegemony. Even if you agree with him entirely, you should never cite him as the source for any of your information. He was simply given an article in Time Magazine to draw attention to him. That attention lets the guys who, for some reason, imagine AI will lead us into utopia, point at him and say "oh yeah, well, this is you!" to people who don't really want even scam democracy replaced by some AI-run, micromanaged, social credit infested nightmare that you can't predict at all.

Still, you want people to talk to you, you gotta be polite, even if that means calling someone like him an "AI Expert".

I hope that's what the idea was, anyway.

[Im_CIA]

On 5/2/2023 at 11:20 AM, ScumCoder said:

Let me guess. You are RaTcHeT302, aren't you?

Don't listen to this heckler.

 

On 5/2/2023 at 9:45 AM, PancakesLover said:

<content>

If you're really ratchet: welcome back!

[interesting]

On 5/3/2023 at 12:16 AM, probablygonna said:

Well, I was going to try and phrase this more politely, but Eliezer Yudkowsky is someone I've come to regard as controlled opposition for the rich people who are after total AI hegemony. Even if you agree with him entirely, you should never cite him as the source for any of your information. He was simply given an article in Time Magazine to draw attention to him. That attention lets the guys who, for some reason, imagine AI will lead us into utopia, point at him and say "oh yeah, well, this is you!" to people who don't really want even scam democracy replaced by some AI-run, micromanaged, social credit infested nightmare that you can't predict at all.

Still, you want people to talk to you, you gotta be polite, even if that means calling someone like him an "AI Expert".

I hope that's what the idea was, anyway.

Yep. This is the only important thing we should pay attention to.

[Deep Dive Devin]

On 6/5/2023 at 12:58 AM, robertjace said:

It is very interesting and very nice story fun. There are many people appear in this forum and sharing different Ideas. Mr. Eliezer is very great boy. At the role perform Mr. Eliezer wears  a jacket which is one Astros Sequined Jacket attractive attire. 

Smartest AI lesswrong will ever produce.