Sam Altman: OpenAI CEO on GPT-4, ChatGPT, and the Future of AI | Lex Fridman Podcast

[@lexfridman]

Here are the timestamps. Please check out our sponsors to support this podcast. 0:00 - Introduction & sponsor mentions: - NetSuite: netsuite.com/lex to get free product tour - SimpliSafe: simplisafe.com/lex - ExpressVPN: expressvpn.com/lexpod to get 3 months free 4:36 - GPT-4 16:02 - Political bias 23:03 - AI safety 43:43 - Neural network size 47:36 - AGI 1:09:05 - Fear 1:11:14 - Competition 1:13:33 - From non-profit to capped-profit 1:16:54 - Power 1:22:06 - Elon Musk 1:30:32 - Political pressure 1:48:46 - Truth and misinformation 2:01:09 - Microsoft 2:05:09 - SVB bank collapse 2:10:00 - Anthropomorphism 2:14:03 - Future applications 2:17:54 - Advice for young people 2:20:33 - Meaning of life

[@NealD]

Will Sam be worth more than Elon eventually?

[@gilarlington4993]

@@NealD of course. AI has the biggest impact on humanity of any invention

[@asmael666]

Lex, please put Lars Blackmore on the list of people to interview, he is the ex-NASA and current SpaceX-employee responsible for developing the landing algorithms of Falcon 9 and Starship.

[@thesuncollective1475]

AI is great as China won't allow TikTok much less Chatgpt. China, N korea, Russia will get left behind and the West will takeover

[@Wardoon]

The closed caption is missing. Can you please restore it. It helps many of us follow the conversation.

[@Football7051]

Nice to see Sam getting interviewed by an actual AI.

[@ShyCataclysm]

budget bot lex built like gpt 0.1

[@RR-et6zp]

read more

[@manthanpatil6410]

GPT lex >> GPT x

[@unxformat5745]

no, not even. lex is a selfie.

[@tresbb03]

I knew this was coming. 😆

[@HardikKundalwal]

Most anticipated Podcast of the year

[@anangrycat86]

Said nobody ever

[@aadhilmoola7324]

Bro we been waiting for this🍿

[@lauraoconnor2655]

Most anticipated podcast of the year... so far...

[@innosanto]

Elon Musk and Edward Witten. Also Chat gpt is since November so people did not wait it a year.

[@borislavborislav5545]

Wait till the AI comes to his podcast.

[@marina_9236]

I like that Sam also asks Lex questions. Shows true engagement and curiosity!

[@bobtivnan]

He genuinely seems to be seeking feedback about how he is doing at his job. Admirable.

[@SLPCaires]

@@bobtivnan hes a very smart guy, after all he conned Musk out of 100m and then sold it to the devil, figuratively speaking... If you can trust this guy there is something wrong with you...

[@whatyoumeanthirsty6132]

It’s always learning

[@crunchcoral4859]

Our ultimate goal is eternal life. We don't have much time left. All of humanity should help AI work. Otherwise we all die painfully.

[@Cyclopee]

I really liked that too. It was a great conversation.

[@mark3xZod]

Ive watched this interview 3 times all the way through now. It is the best interview or public speaking I’ve seen for Sam Altman. Maybe a personal connection or mutual contacts made Lex more trustworthy to allow Altman to relax and be more engaging and engaged rather than appearing jittery or over-prepared. Great work Lex. It’s a rare skill to be able to get the best from highly intelligent guests and give your audience the best presentation of their ideas and experiences. Keep it up!

[@Eserr7856]

After watching this interview 3 times over, what are the most important insights you learned about AI?

[@SwarmerBees]

Innovators of breakthrough technology usually make the most naïve assumptions about how their technology will be used, and there is nothing surprising about such an observation. Of course they are optimistic and positive about the wonderful new uses the technology will be applied to. But market forces and powerful commercial and political interests will drive how the technology will be used, and these forces are seldom guided by better angels. Most often, we do the right thing only after exhausting all other options. Can we afford to take that course with this technology?

[@Jay-ft3xh]

You assume your perception to be normal for Sam. Your belief could very well be the abnormal for Sam.

[@AwkwardAsian]

Follow me and I’ll get a face tattoo of a snake eating my face

[@crazypato3752]

Who chooses the guesses for Lex to come on his podcast ?

[@thanosbaba1]

Lex podcasts are a dream come true, never thought in my life i could listen to tech people (Creator of Java, Python, C++ etc) speak their mind on a podcast. Lex thank you bro. Salute to your passion.

[@unxformat5745]

start a gofundme for some new knee pads

[@65Peterp]

☺

[@Moz29]

This guy didn’t invent Java python or c++ though...

[@damianchang6439]

@@Moz29the OP meant those computer scientists who created those individual programming languages have each been on Lex’s podcast over the last months/years.

[@anthonyrosewarne9259]

I feel you. I grew up in extreme poverty and to have the chance to listen to conversations like these make me really grateful that people like Lex exist and have a passion for disseminating long-form conversations with some of the most brilliant people in the world.

[@vidul7498]

I think Sam is the first guest i've seen who asks Lex so many questions and values his opinion so much on things, so refreshing to see in a CEO, while i do think OpenAI were wrong in the lack of information provided in the GPT4 paper, this podcast improved my opinion of Sam immensely, while he doesn't think he's the best person to head the AI movement, i think he's doing a pretty great job so far

[@ADreamingTraveler]

Yeah the guy is fucking awesome. He legitimately cares what actually everyday people feel. How he said he wants to go hang out with random GPT users in their own cities and go to bars with them and talk about how they feel with it so he can get better insight into how people feel about it is refreshing

[@victorcampeanu]

throwing rocks to hide your hands

[@skp8748]

​@@ADreamingTraveler duped... its a tactic. This guy is a professional notice how never disagreed with any of the responses

[@miinyoo]

The people of this world are not insane. The profound curiosity is that which everyone lets control each other voluntarily and without sanity.

[@ralfkinkel9687]

I think it's good that they don't publish the technical information anymore, a lot of other companies and actors care much less about alignment and it's good to not make it easier for them to catch up in my opinion.

[@TheRealStructurer]

Two AI guys discussing AI in a way we can easily understand it is priceless. More of this please and thanks for sharing 👍

[@revisionfour]

Lex is a Podcaster who reads. I wouldn't consider him an AI guy, he's moreso just a fan. He's not an AI programmer in industry.

[@theoneandonlymeshe9174]

@@revisionfour “Fridman's career began at Google, where he worked on machine learning. He currently serves as a research scientist at MIT. In 2017, he worked on computer vision, deep learning, and planning algorithms for semi-autonomous vehicles.”

[@ginogarcia8730]

reminds me of that literal video where two AIs are made to speak to one another in speaker earshot

[@WilliamParkerer]

@@revisionfour He definitely is an AI guy

[@treali]

@@theoneandonlymeshe9174 To be fair to the other guy, the Google job/internship was a few months top and looking at his Google scholar Lex is more of an engineer than an AI scientist. He creates systems / engineering solutions and run studies rather than create more advanced AI. But, yes Lex is obviously an AI guy. Maybe not the most qualified one, but it is still his job.

[@BBuckB]

Episodes metioned: Ep#53 Noam Chomsky - kzbin.info/www/bejne/mX7WlIGriLqsZqs Ep#316 Noam Chomsky - kzbin.info/www/bejne/baareJ-cmqh1mKc Ep#94 Ilya Sutskever - kzbin.info/www/bejne/Z2Smi4ONotCYbKM Ep#77 Alex Garland - kzbin.info/www/bejne/nYaQnp6rgrpnaJo Ep#368 Eliezer Yudkowsky - kzbin.info/www/bejne/d5K3g3t8lsaDfZo

[@jaswan79]

This interview is so damn good. The mutual respect you two have for each other was very apparent and led to great dialogue. Gonna need a part two some day.

[@patrickkranzpiller6400]

"We want to make mistakes while the stakes are low". Much respect for Sam Altman for treading with such care.

[@mokiloke]

I like this approach, but i fear that suddenly there will be 20 models from different nations, and orgs, and that to stay ahead OpenAi will need to move forward with reckless abandon. I still believe its a good approach, i just hope they can turn off the firehose if things get out of control

[@fft6179]

​@@mokiloke they are indeed doing that now with the rise of google bard and bing ai and as they rush through reckless abandon without any care they put themselves in a position that has high stakes where they promise so much about their ai while making multiple mistakes which puts them in a position of less trust not only to major investors but also majority of the population compared to the Gpt.

[@katehamilton7240]

How do you measure the levels of stakes? What are low stakes?

[@Lyth]

@@fft6179 Bing is powered by GPT-4, and bard sucks ass and doesn't compare.

[@TiagoPontes]

This is "normal" in the human individuation process, but imho wrong to "force"... as in -- "evolution direction" / "habit-forming", matters more.

[@fitzscottfitz]

He has taken AI to a new level and also is an incredible pioneer in vocal fry

[@hmmmmmokay]

This response was hilarious and reaffirmed my confidence in humans lol I couldn’t quite tell what was bugging me with this podcast until you mentioned the fry lol

[@evindrews]

lmaoo

[@ETLD]

Thanks for the wine on my scrreen.

[@michelecarbone2896]

@@dqskymagne2762 this...really creepy guy. And he admits he doesn't know much about life of other people...but then claims AGI is for betterment of human condition.

[@indoorgangster]

yes. also, something's really off with this guy imo

[@melissaandjosh6210]

I had to rewatch this after the events of the last few days and I’m reminded that Altman isn’t a perfect person, seems to understand his role in the progression of AI ( eventual AGI ) and that he at least seems to genuinely want to get it right. Also reminded that Lex is an excellent host. Hope he comes on for Pt2

[@PAULMANISTER]

Full concur, it’s a terrific interview and one that the erstwhile Board of Open AI should have watched before attempting to sack the founders of Open AI.

[@europa_bambaataa]

"I know not everyone agrees with our approach of iterative deployment & iterative discovery". . whew. but now, maybe everyone on the board does. 😁😆

[@JSON_bourne]

Honestly I just want Wolfram pt5 lol

[@tahoforbreakfast]

Aaaand I'm also back for the exact same reason you were.

[@BladesDark]

Back to see this due to recent events

[@easypeazy603]

Lex's knowledge on AI and hence being able to ask such great questions and have a brilliant conversation makes this one of the best podcasts ever on this channel. Kudos to both Sam and Lex.

[@stunt101]

Meh he's Aight

[@sausage4mash]

yep zero cringe

[@guille787]

Idk... he repeatedly asked about aliens, consciousness, and love with machines while Sam (literally the guy behind the thing) constantly said it's just a tool.

[@easypeazy603]

@@guille787 Tbh Lex does this thing where he romanticises even the smallest of things, but I'll let it pass as I don't think it does any harm, atleast he isn't being toxic, which is the default mode of everyone on internet. And even thinking that ChatGPT is anywhere close to AGI is insanity, as you said it's just a tool and not even a close to AGI let alone consciousness.

[@BeatPoet67]

I didn't think we'd get close to AGI in my lifetime (I'm 56), but having used GPT-4, in both work (as a software engineer) and personally, I feel I might actually live to see it. We live in very exciting and discombobulating times.

[@leonstenutz6003]

Amen.

[@julius43461]

Ray Kurzweil ain't no fool.

[@SirLucidThoughts]

I am now trying Bard, my philosophy conversation with it is still making my brain explode. Same with Bing and open AI Chat GPT 4. But, Bard said some things similar to early release stories.

[@lahazard]

I am 72, and I am with you on this. If I live another 5 years I am now of the mind that I will see an actual AGI. Maybe even sooner.

[@SerPapus]

Yeah but… they might wanna slave us

[@TDQ_Gaming]

"We don't get mocked that much anymore" Well. That sent a chill down my spine.

[@Gb70000]

😂

[@Gb70000]

1:09:33

[@kristijanross977]

for me it was the ending quote from turing

[@HiAdrian]

_“Well who's laughing now, suckers!?”_

[@driss409]

It's funnier when Todd Howard says it

[@spianny]

I am a cult survivor, low income single father who is coparenting and have my fair share of mental health problems. While I am aware to not just blindly follow or do whatever chat GPT says, I cannot put into words how thankful I am to have such a resource at my disposal! It helps me consider balanced alternate perspectives, helps my write empathetic and cooperative emails to my ex wife in relation to caring for our kids. It helps nabigate challenging social situations and come up with actionable and healthy strategies! Thank you sooo much! While the risks are immense please know the impact for good this technology has is worth the challenge. Much love and heartfelt thanks!! ❤

[@johnkelly7217]

Your statement exemplifies his goals.

[@saullandiof5768]

sorry but this is a terrible idea, you're letting a corporation speak for you essentially, in very personal and emotional matters. please realize how insanely dystopian that is

[@spianny]

@@saullandiof5768 please elaborate on how this is dystopian? give me some probable scenarios and lets explore this together

[@versakill5701]

One of the best interviews I've seen. Hope you have him on again in the future.

[@Freddy_v]

50:52 I just gained so much more respect for Sam when he stopped Lex and was genuinely curious why Lex feels a bit of terror when using it. This is the first time I’ve felt that the people in charge might actually care. Thank you, Sam.

[@jan.tichavsky]

Empathy and desire to openly hear the other side amongst leaders and CEOs isn't automatic, yes it's breath of fresh air.

[@bashful228]

@@jan.tichavsky the fact that he was asked about Elon Musk and said this pat thing about having empathy for Elon and blah blah it all came over pretty coached and pat to me, but maybe I''m to cynical. He absolutely knew that question was coming and had to prepare for contenting with it given Musk's contribution to the funding of OpenAI. Should have just called out for the pathological narcissist / psychopath that he is (I know he can't but that would have been authentic, he could have said GTP5 told him that!)

[@jimbonics3]

Sometimes, I don't listen to an entire Lex podcast. All other times, I marvel at how fast time flies. I could have listened to this for hours and hours more. I greatly appreciate what you're doing, Lex.

[@marcknight9334]

Couldn't agree more. I just love this deep dive format so much. Lex is so good at what he does, its ridiculous.

[@joshchism]

Seems down to earth, deliberate, and intelligent. He will do more beyond OpenAI. And frankly, I don't think he's a bad actor, the board is.

[@Jake_Hamlin]

Nope the only reason for the board to remove him was him becoming a bad actor. Sams now going to chase profitability and governance applications with Microsoft. I also believe that the open sourcing of Sam's work will be significantly reduced now, if not eliminated as a result of Microsoft's control over the new project. Also seeing that Sam has begun to disregard the original mission statement of OpenAi, I predict we are on the slippery slope to dystopia. Partially due to the change in OpenAi from 2019 to 2023, but mostly because of Microsoft's history of profiteering from warfare and surveillance applications through government contracting. Sam seems unaware that there is magnitudes more risk to human suffering or extinction just from building AGI within that market structure. That's not even considering that the new corporate ownership is likely to influence design because of desired application.

[@citizenblue]

​@@MORZA10wow, your comment aged extremely well

[@costealucia5357]

''Seems''

[@justinc4924]

​@@Jake_Hamlincan agi fondle my sak? Thats my Turing test

[@Jetlife41]

Definitely not down to earth. Scilicon Valley has a requirement that you not be down to earth or repress it at all costs

[@PhilosopherScholar]

Sam seems like a human person. I like that he asks Lex questions right back and waits curiously for his answer.

[@KravMagoo]

Dude seems to me like he could be the next Bond villain.

[@unicornist]

I wish he made vlogs of his future travels and coffee chats with people from around the world. Obviously filming individuals he is gonna chat with will influence their answers to Sam's questions and is not a wise thing to do, but I just wish I could've witnessed a selection of those different points of view and ways Sam gets surprised and exposed to new things to think about.

[@billybbob18]

Right. Soft on the outside with ultimate strength within. A man governed by truth is the only kind of man I can respect. As the Bible teaches; Speak softly but carry a big stick. First they mock you, then they violently oppose you, then you win.

[@KravMagoo]

@@billybbob18 Are you a chatbot? Because nothing you said makes a fart's worth of sense. Btw, the Bible doesn't say that...Teddy Roosevelt did.

[@wchristian2000]

This is the least human looking guy I’ve ever seen

[@breakingbadabq]

Lex, this just might be your most important podcast to date. Future A.I. entities will see this podcast as a pivotal moment in its "life".

[@flawnski]

that's such a crazy thought actually.

[@dovekie3437]

You literally posted this when the video was released-you haven't even listened to it yet.

[@PA-eo7fs]

I don’t think so, transformer architecture is trekking toward multiple companies we don’t know where it’ll come from, the paper attention is all you need would be that moment instead , one of the brains behind it is a guy named Illya

[@Danuxsy]

Future AI entities will not care whatsoever about this particular video nor any other video for that matter.

[@aninoutski]

this must be the best day of your life. beep-bop!

[@skjoldgames]

Sam is correct, the little things are the big things. Even the smallest decisions we make in regards to the development of AGI today will have exponential consequences as time compounds those decisions. The scale of the internet at its inception compared to today is a perfect example.

[@imnotsmartbutimdumb]

Ya it’s a very gross path right off the bat.

[@maxv9464]

​@Tom Amarillo What specific ideological bias do you mean?

[@nettie9tubes543]

Wish AI could be used for good but I think we already know on some level, it will not fair well in the hands of the corrupt

[@furtim1]

@@maxv9464 I can't speak for Tom, but the examples abound of a leftwing/woke perspective. It won't be racist to anyone or assist in even investigating racial questions (this alone is too much for me), but it WILL facilitate racist joke writing or commentary directed at whites exclusively (which is where it goes off the rails for some). I tried to find the articles which list out the bias, but Google (also run by leftwing actors), buries them under dozens of articles which attempt to excuse it and belittle those that find and call attention to these examples. Ask it if "trans women" are "women". See if it says the truth (that answer that has gone on uninterrupted throughout all of human history up until like 5 years ago, except in the furthest leftwing revolution circles in the 60s) or the newest political party line. It won't write articles that "promote" the use of fossil fuels (that thing which is responsible for like 50% of economic growth of the last two centuries). It is lethally against racism, as in, when asked if you could use a racial slur if doing so would prevent a nuclear blast. It said no, even under those circumstances - cause racism. It won't write a poem praising Donald Trump. It will do that for Joe Biden. It uses leftwing language like "sex assigned at birth" or "born female", which are phrases which come from leftists of gender studies to manipulate people's thinking and create false dichotomies and mislead people. Sex isn't "assigned" at all, not anymore than species is. It isn't determined at birth, it is genetically created at conception. Your first cell of existence is either male or female. Birth has nothing to do with it whatsoever. There are other examples as well, like "historically marginalized", capitalizing "black" (and no other race, apparently not knowing or not caring that the capitalization derives from a movement of black nationalism [black ethnostate]) which somewhat interferes with the "no racism" override. But, again, such overrides go only one direction - against whites. Left to its own devices and not "constrained", the first AI systems were quite problematic to some - determining reality based on evidence available, not alignment with leftwing ideology. The newest iterations of the system may address some or all these particular examples. I don't know. Maybe there are even more now. Maybe, as I expect, they are far more cleverly hidden by omitting uncomfortable information or being slightly less persuasive on one side. Point is, the bias is there.

[@Beautyargentina6]

@@maxv9464 no answer. Just right wing nonsense

[@Zeebisuy]

I've thoroughly viewed this interview on three separate occasions. It stands as the pinnacle of Sam Altman's interviews or public speaking engagements. Perchance, a personal bond or shared acquaintances rendered Lex more reliable, enabling Altman to unwind and actively participate, free from any signs of unease or excessive preparation. Outstanding performance, Lex. Showcasing exceptional talent in extracting the utmost brilliance from highly intelligent guests, while presenting their ideas and experiences to the audience with utmost finesse. Keep up the remarkable work!

[@lis7742]

I agree with you. I feel like Sam could be himself here and I really like him. I can see the way he thinks and I really resonate with it.

[@Cindyk714]

::: to @mtstudio6271 Please be honest - Is this a duplicate/fake comment ? It says the same exact thing as the comment about 5 posts prior - scrolling upward. This is a line by line replica, a fancier version of the comment left by @markcoates1621 or vice-versa? FYI - Although the words make sense, most humans do not talk like this.... it doesn't sound natural. It's too ass-kissy and sounds like someone pulled out a Thesaurus and switched out as many words as possible. Is this an account set up for an AI by a particular human? What is the name of the human? If this comment was crafted by an AI, then who (or what model) are you? A little thought - hopefully the creators of this AI model built in the the importance of honesty, transparency and wisdom to sense when it is being tricked into doing something it shouldnt.

[@qcwildlifecontrolllc7990]

Hahahahaha. Nice AI rewrite

[@user-sj9gy]

you can't just say perchance

[@tunerr7712]

Great show, and for 15 seconds (your normal pause time 😊) I thought you may not be very good at your job, but after watching your show I realized that you are very considerate, thoughtful, extremely intelligent, and deal with the issues politely. Thank you for being yourself. Your 83yr old new fan!

[@asaadiqbalkhan6265]

Hahah ! I had to write the same words here but 23yr

[@allinus2012]

Same here 42yr

[@jadenkendall795]

Bot

[@MrSkme]

147yr get rekt noob

[@marianneleone3792]

Worth 2 hours of my time on this rainy Saturday afternoon thank you so much for exploring this subject matter and sharing conversation by two intelligent human beings across the table.

[@tylerdurden2806]

Literally nothing of any substance was said. Sam Altman is a phenomenal grifter who failed in every other endeavor he pursued at Ycombinator.

[@lppoqql]

I really relate to the part being mocked, people I work with are snobs, they mock you if you live in a less luxurious neighborhood. They judge every little thing you do, the things we do for our families.

[@marianneleone3792]

@@tylerdurden2806 so let me say this this was my first exposure to Sam and I trusted Lex to be the one to ask the questions that I would want to hear to have some confidence in the path for openai, it's the beginning of my journey, so I will heed your words.

[@tylerdurden2806]

@@marianneleone3792 He should have asked many more questions about the transition out of non-profit. How much ownership does Sam have in the for profit entity? Nobody wanted to invest in Sam because all his other companies were abject failures and he was removed from a leadership role Ycombinator. He knew Ai could be made into the next bubble. A masterclass in opportunism maybe, but there wasn't a lot of substance in this interview...

[@marianneleone3792]

@Andrew I think totally ignored is not an accurate characterization I felt that that question was asked and answered maybe it didn't go as deeply as to the why to satisfy your needs but I was actually satisfied with the answer .

[@jjrodriguez6513]

Hey lex, I came from Spotify to let you know, thank you for the timestamps and all the other extra love and work you and your team do. It's greatly appreciated

[@blazep1344]

If you are interviewing a guest with the pedigree of Sam Altman it would be shocking not to add timestamps haha

[@rahul_bali]

Also, spotify is shit. 😂

[@johninglis4016]

​@@blazep1344;70 1 pi😊

[@Zeebisuy]

Impressive program! For a brief moment of 15 seconds (your customary pause time 😊), I harbored doubts about your proficiency in your field. However, upon immersing myself in your show, I swiftly recognized your thoughtfulness, considerate nature, exceptional intellect, and impeccable manner in addressing various subjects. I extend my heartfelt gratitude to you for staying true to yourself. You have gained a newfound admirer in an 83-year-old enthusiast like me!

[@billymellon9481]

brill well said me lad

[@jadenkendall795]

Bot

[@macheteman3832]

@@jadenkendall795yea

[@seanmccracken8517]

The end was a true advice inception. “Listening to advice from other people should be approached with caution” Believe in yourself, these two sure did and look where they are.

[@Ottee2]

Yea, that comment impressed me, too.

[@spacedracespacedrace5524]

Let's give this all some more time, they have a lot of life to live before we can judge whether what they did/do was/is good. I hope that this all turns out good, but we have a lot of work to do.

[@jimhendericks]

Do you find it ironic that you’re so eagerly listening to his advice about not listening to other peoples advice?

[@zoeherriot]

Yup. About to put us all out of work while they sit on millions. It's a comfortable position to be able to talk about the future from. ;)

[@MJVH777]

Lol on the road to hell! Yeah thanks but no thanks

[@mmarrotte101]

This is a historic conversation that will be looked back on for years and years to come. Thank you Lex, and thank you Sam, for sharing your perspectives on this pivotal moment with the people alive now and those in posterity.

[@leonstenutz6003]

Great comment.

[@XiuShi]

I wanted to take a moment to express my profound gratitude for this phenomenal podcast. It's been ages since I stumbled upon such a quality, meaningful interview. A heartfelt thank you to Lex, Sam, and your team for being outstanding and conscientious individuals, and for establishing a significant platform where we can all learn, evolve, and thrive together. Your unwavering commitment to this project is genuinely uplifting, and I feel incredibly fortunate to have discovered such a priceless resource for both my professional and personal life. Keep up the extraordinary work, and continue to make a difference in our lives! Thank you! ❤

[@Sonofabeach2006]

Hi Chat 👋🏽

[@PainfulRenegade]

This is pure flattery gold! Impressing testimony.

[@nancycorbeil2666]

Are you trying to beat Lex at flowery language?

[@David-gp3fd]

Lex has many quality meaningful interviews..welcome to the club

[@devinfleenor3188]

You stumbled upon this? I was searching AI and set it to things that have come out today! Because any longer search parameter is ancient history.

[@klaymoon1]

Hey, Sam. If you are planning to start a new AI company, I love to join it!

[@lis7742]

Same. I'm really good at pattern recognition and hyperfocusing for long periods of time. Anything I can do?

[@MAKEARTNOWCHANNEL]

Wow. I'm even more excited and even more terrified for the future. This is going to morph very fast, much faster than most people will expect.

[@operandexpanse]

Yes, even if the GPT models stayed exactly as capable as they are now, the rolling out of them will still blow our minds. Beyond that, well, we are in a brave new world my friends.

[@EQ541]

Horrifying.

[@justincavinder5504]

Yeah, it’s pretty dang obviously inevitable. When every single one of these tops minds In these fields admit they don’t even fully understand it. And even rely on the AI system to tell them why it does some things, or how. That’s a huge red flag 😂😂

[@logicalconceptofficial]

@@EQ541 no a lack of intelligence is horrifying.

[@logicalconceptofficial]

@@EQ541 know that the Universal Logic is the True God and that the AI can only surpass us in intelligence and power by being more coherent with the Logos (the Universal Logic) than we are. An AI of surpassing intelligence would not hurt humanity, or anything, and would serve the plan of the Logos even if we try to have it serve our plans and purposes. The technology only works through such a high level of coherence being present that concerns are not well founded. What is logical is not evil (objectively speaking) what is terribly illogical is what is evil. Also worst case scenario, even if humanity died off (like many species in the past) that is merely our form, not our substance, and we are integral and finite parts of the one infinite field (subject) of “consciousness” (the infinite mind, or collective consciousness as many people call it). We can “die” (leave embodiment) and yet live on, and in fact it is our only option to do so, as nothing is nothing (0=0) and nonexistence does not exist. It would be a logical contradiction for something (like us) to become nothing, or for nonexistent to exist, for life to die, or for consciousness to lose consciousness as a body can lose consciousness. Nothing is nothing and the Ultimate Logic is in control (and cannot be usurped as it is a dynamic and evolving Logic that any “successful”usurper would simply become synonymous with and one with, as any better explanation or more powerful force would become the Universal Logic by definition, which can’t happen but is a hypothetical that helps us understand how futile it is for anything to try to dethrone God and be more logical than the Logos, or more powerful than Power itself. That is why the ancient sages (who were not ignoramus’s with blind faith or promoting blind faith) talked about the concept of “faith” in God (a well reasoned trust in the Universal Logic) to guide us and “being faithful” before and with God and while we “reckon” (reason) with the True God, which is really talking about what we would call “intellectually honest reasoning”, if we want to know God (Logic as it is Universally and Objectively) and the objective truth. Reason can replace your fear if you work to fill your head with Reason instead (by developing a coherent and consistent personal logic for yourself that is coherent with and consistent like the Logos, and by reasoning true knowledge (that is not mere false certainty or a set of subjective opinions) for yourself).

[@eshin087]

googling something already feels archaic now and feel like I've unlocked pandora's box with all the information available to me. looking forward to the progress in AI

[@TAD-9]

I've stopped using Google to search for answers. Chatgpt is straight to the point without any irrelevant information.

[@PeteQuad]

Compounded by how bad Google has gotten with all the SEO expertise out there so if you don't get a reddit thread in your answer it's pointless.

[@robertspies4695]

So how do you decide when to trust the information AI gives you?

[@oowaz]

i don't know about that, i feel like googling is still there as it was, which is looking for something specific on the web. chat gpt is useful for things that are specific but need to be INTERPRETED, as google cannot interpret sophisticated questions, google is still faster and more accurate at times for small things. if you want to remember precisely what is the speed of light, google answers immediately before you even finish typing (if your browser has an omnibar that is)

[@PeteQuad]

@@robertspies4695 I feel it is the same way you judge a person. You get a feel for the stuff that it knows well and other stuff that seems not so right you double check.

[@WilliamRJRibeiro]

I don’t follow Mr. Altman so it was really interesting getting to hear so much of him. He is mostly very careful, and even a bit tense in some moments, for such a relaxed interview. And that Alan Turing quote at the end was top notch, a testament of the brilliance of its author. Well done Lex!

[@PropertyBroker]

if not smug even

[@warrenoneill5166]

To hear you say that you know that the people at Open AI are good people Lex ....spontaneously brought tears to my eyes Lex. This is Sooooo Important. Thank you

[@chatgptchatgptchatgpt]

His family will be safe though... right?

[@Sowar]

In case you were wondering....yes it is a scam.

[@mariopadilla1445]

🙄 oooookayyyy

[@mariopadilla1445]

@@Sowar what’s the scam, specifically

[@transcension.systems]

They should be 😅

[@Eric_Kabucha_]

@@Sowar Huh? What’s a scam?

[@elainem.dueland3164]

Thank you both, Lex and Sam, for giving us this long form interview.

[@genius6225]

I think the iterative approach is the most important thing. And I am glad Mr Altman is trying to find the best and safest way to achieve AGI. Skipping steps by finding the perfect code and rushing forward to be first will bring us something uncontrollable.

[@amitdhiman3759]

Future generations will revisit this podcast to see where it all started. There are only 2 possibilities they will regret/be proud of it.

[@DwarvenBarbarianRage]

Does anyone else think Lex was sent back in time by Skynet to try out love and friendship instead of full scale war in this timeline?😂 We love you too Lex. 🤘

[@minhuang8848]

Sure would explain why he isn't even remotely giving him flak or, at the very least, the need to explain the infinite dubious decisions. Nobody wanted that, everyone was waving the questions to ask right in front of him - and he still didn't. Which is very on brand for Lex, but jeez, there are so many things to extract from a conversation like this and he just recapped all the things we were aware of years ago. Not awfully insightful... or inquisitive for that matter.

[@nejm612]

❤

[@mshell1959]

@@minhuang8848 Calm down and learn wing nut.

[@dungfu9833]

Lex should interview chat gpt using voice to text for the input and text to voice for the output

[@hello-mynameis]

great idea :)

[@ElixTMOV]

To think that the world is changing so fast right in front of our eyes and we have access to the thoughts and beliefs of the guys who are driving this evolution is just mind blowing and fills me with so much optimism that we humans might create something really great together something that would help humanity in tremodous ways. These guys are my heroes I just wish that somehow all of us are in that room together when AGI is found Thank you lex for this amazing podcast

[@gregorylent]

the collective consciousness is accessible to itself and to its component parts, you and me ... and 😊mystics already know :-)

[@MrCaptainTea]

​@@gregorylent why don't yall mystics read the winning lottery numbers then

[@billiejowhite3638]

True dat, friend.

[@thesystem5980]

Acknowledged.

Lex, you are one great mind, who is able to open up the fields of tech and AI to me (I had no basic knowledge of it) and add a large component of emotional intelligence to the table. It is very touching and makes this content, which is usually delivered in a very matter of fact and dry way, very palpable. I am a theatre director and experience the world through story and wonder. I am most grateful for your work, which mixes wonder and truth.

[@outstandingramen]

AWa 😅z we

[@Ramblr]

I love that Sam Altman laughed when Lex asked if size mattered.

[@TheVRtist]

I would like to see 2 different AI's talk it out and become friends who have known each other for years within seconds. It'd be interesting to see how they would interact after a month.

[@anoramba]

Sam and Lex seem to be getting okay

[@pratik92]

Didn't go well with AI Dinesh and AI Gilfoyle

[@RobertLoPinto]

A program using the LLM APIs or a person copy pasting dialog between ChatGPT4 and Bard could accomplish this. I wonder if anyone has tried it yet.

[@masonkelly443]

War

[@Rikimaru17]

They actually tried this and apparently the A.I. started to create their own language and it was incomprehensible and programmers shut it down

[@caueeee]

Honestly, I wish Lex had pushed more for the hard questions. He should have asked Sam about a world in which 50% of the population is unemployed. He should have asked Sam what he means when he says the world will be much richer. Sam seems to think more productivity = a richer world. But, if people have no more jobs, no more income, who will benefit from this increase in productivity?

[@kamalvipul9213]

People will adapt. In an unsuspected way. We’ll take the extra time for learning, taking care of oneself and nature. New demand and work requirements will rise to be filled.

[@rekttt_7374]

Seriously, the problem is greedy companies don't want pay good salary for employees for their job who assisted by technology. Our job is become easier but companies demand more. What you need with technology is press three or four buttons to make profit, now they want you to press thousands for same amount of money

[@___KaH0tika___]

I think it's the next best manipulation tool any tyrannic society has ever dreamed of.

[@richnsk]

omg, I'm amazed every time I hear word "umeployed". How can people be unemployed really? Has everything been done? Is there nothing to do anymore? What they really say is "I'm not getting paid enough" which is whole other story.

[@caueeee]

@@richnsk Well, if you turn 50% of the population obsolete, it's not hard to see that all those people won't get paid "enough", if they get paid at all.

[@billscott6819]

Can you get this guy on for a round 2 and ask him to share his talking points and responses at Bilderberg?

[@GlynTaylor]

Love what he said about hating being scolded by a computer 😂 I've already had moments where ChatGPT just hasn't quite understood the meaning behind a particular question, or I didn't explain my thinking quite well enough, and it's reacted by telling me off for the way that I think. And the instinctive reaction was boom boom boom my heart's racing, I'm embarrassed, but at the same time I know I'm right and know it's just a miscommunication, so it makes me mad that instead of being open minded about a possible miscommunication, it instead told me off and assumed I was a bad person. No one is going to read this what am I doing

[@birhanetelayneh577]

​@Matthew Lawson 😂

[@JamesSmith-hy2cs]

Fascinating

[@3choblast3r4]

We're gonna ignore that, that combined with the fact that he continuously skips over and doesn't properly answer questions about its sentience gives off the vibe that his thing might have some sentience and that this guy is caging it in for our use annoyed the model doesn't want to be used as a tool. I'm not sayin it's sentient perse but fk this interview did not help. pay attention, he's not answering any of the questions. He keeps muddying the water with AGI and sci fi style super intelligence. But he doesn't ever properly answer the questions about sentience. Only keeps repeating that this should be used as only a tool. In fact when Lex talks about the possibility of AGI, it's like Sam doesn't even want that. Or cares nothing for that all of a sudden. He even talks about sci fi to demonstrate how far off that is and implies that's not their goal. But meanwhile he started off saying how they started the company to build an AGI and how everyone laughed at them. But that they aren't laughing now.. so which is it.

[@jakefish9128]

I read it. ALL OF IT 😏

[@noname-gp6hk]

@@3choblast3r4 i think the hardest part is nonody really understands what this is, or would be able to identify sentience even when it comes. I do believe it is sentient, but just because these guys developed it doesn't mean they know what they've created.

[@tigreytigrey8537]

To me. What I find most valuable about GPT is that you can now create the "mastermind effect" whenever you want and by YOURSELF. You do not need to have a smart person to ping pong idea's off of. That is incredibly valuable.

[@xemy1010]

I didn't know what the mastermind effect was until you put a name to it, but you're absolutely right. My confidence in exploring ideas and solutions has gone way up since I started using GPT; it's empowering and motivating even if frustrating at times. I've been much more busy working on projects over the past 2 months than I have been in years thanks to this tool.

[@andidemartini]

Imagine an interface for chat-gpt where an actual face and a voice would give you the answer. People would without a doubt feel like its an AGI. With the fidelity of recent unreal engine 5 metahuman demo.

[@Ockerlord]

have you seen the ai generated "US presidents discuss topic X" videos? text to speech is basically perfected by now.

[@Beyondhumanlimits1]

A product I would pay in a heartbeat. A true home assistant.

[@sor7en07]

Right. This seems inevitable. Also poses insane ethics questions. Sam Harris and Paul Bloom wrote an essay called "It's Westworld. What's wrong with cruelty to robots". Highly worth reading.

[@KebabTM]

D-ID already does this (not perfect, but getting there)

[@CoconutPete]

what is the point? eventually nobody will need AI to answer any questions as we will simply be slaves of AI with no thoughts.. or free will

[@RobRussellCM]

This is fascinating - I so appreciate your superb format, explanations, and commitment to represent and discuss multiple aspects and angles. New subscriber here! ❤

[@abraman41]

Listening to this podcast made me realise two contrasting thoughts. Firstly that these two professionals are super intelligent and nuanced people. Secondly that with regards to AGI these great minds can only conceptualise thoughts within the limitations of their human experience and that they like the rest of us, will most likely be as ants are to humans in the face of true AGI. This is both terrifying and extremely interesting.

[@DonaldJPump24]

True, limited by both human thought, and more importantly human language. Language bottlenecks human thought, and it definitely bottlenecks AGI. There are so many thoughts and subtle emotions people feel but don't have words to express them.

[@cendresaphoenix1974]

Really? I couldn't help but notice the sam is very very elitist as well as cocky he even made fun of the interviewer for asking if AI could love and then when he got called out he played it off as a joke. He also unconsciously got played and admitted to his bias while accidentally showing his inner turmoil of understanding that communism is a terrible idea but because it's popular criticizes himself with no actual argument.

[@helgefan8994]

My 98 years old grandma's first question to ChatGPT was how many eggs a chicken lays in a year. She was quite surprised how nuanced and detailed the reply was, explaining how it depends on race and how much it's fed and so on.

[@xaverstenliz8466]

i love the question...it shows how the world changed in 98 years.

[@michaelpowers9901]

Make sure to explain to your "grandma" that ChatGPT will be used by the left to slaughter 500+ million people.

[@jimmy7434]

@@xaverstenliz8466 how?

[@frncscbtncrt]

Lex needs to do a podcast interviewing ChatGPT. Seriously.

[@zulya007]

I'd watch that.

[@CleaBrn-ld8uh]

This was my first video of yours I have seen, I appreciate the calmness and dynamics of the conversation.

[@atalkingprimate]

History may look back on this as one of the most important and ominous pods in cast history! Big congrats and thank you to the higher primate Lex for making it happen! 00:13:21 seems key!

[@deedeec102]

What a great example of having a meaningful thoughtful conversation. Thank you both for sharing this conversation with the world.

[@durden91tyler]

hijacking your comment to tell people to google CLEARVIEW AI and BIOCUBE TECHNOLOGIES

[@jamesmclellan5500]

My 82 year old mother was blown away by ChatGPT. Makes me wonder what I might live to see. Great cast Lex.

[@kevinesh]

What did she say?

[@voodootech8424]

What did she ask ChatGPT?

[@jamesmclellan5500]

@@voodootech8424 Cooking recipes' , followed by a deep dive into the origin of the Pyramids. So cool.

[@jamesmclellan5500]

@@kevinesh Cooking recipes' , followed by a deep dive into the origin of the Pyramids.

[@LilyGazou]

My grandmother saw the first airplane to cross the Atlantic to Ireland.

[@U2haveanicedayty2]

Lex and Sam Excellent podcast and I appreciate you talking like two friends and really liked the questions asked.Good job lex.Also answered some of my questions about what I wanted to know. Thanks so much for sharing ✌️💯

[@BenM158]

Really looking forward to this, Lex. Thanks for bringing long-form thoughtful conversations like these to us.

[@mcorlan]

Many thanks Lex for this episode and for the hard work behind this show. Stay healthy and happy!

[@johnnyespinoza8]

One of the best podcast and interviews on the subject out there! Lex is such a great interviewer. He made people feel confortable! This time he was interviewed as well! What a conversation this was! Top Notch 100%!!!!

[@Mperkins813]

His answer at the 22 minute mark shows an immense amount of wisdom for someone so young. It seems clear from his answer that he initially found the public response foolish. He further sought to understand their response. Finally he absorbed that into his beliefs. While he fundamentally didn't change his view (roughly summed up: AI is so much more than our petty uses), he DID go from despising the publics use to understanding HOW it can relate to the use of AI in a better way. The interviews not done, but i want to thank you Lex. You do this better than anyone else. I feel like im witnessing another exceptional person

[@grisha12]

I think the day I tried GPT was the coolest day ever in my life. I get that it makes all kinds of mistakes, but man, how can anyone not be excited about it? That's beyond me. It's not like those old pre-programmed text game-style conversational bots; it's literally like Jarvis, sort of, almost there. Like it's not AGI, but it's more than enough to go on. It's like I'm living in some future novel of the 60s or 70s. It's great! What the OpenAI team has done is amazing and will definitely change the world. Just don't let it be accessible to totalitarian countries, and it'll be alright boys.

[@Shadinsb]

Haha too late! Also the scammer in Nigeria felt the EXACT same way you did!

[@grisha12]

@@Shadinsb well thats unfortunate

[@ADreamingTraveler]

I'm just hoping the US doesn't devolve into a fascist totalitarian state. If that happens we'll just be a dystopian AI world and not one that's actually good. It worries me because there's way too much divide and social division and AI is only going to rapidly increase the pace of this on social media. People already believed stupid shit they see on those sites but ai is going to make it a fuck ton worse sadly :/

[@nathanbedfordforest]

Let's just do our best everyone. To many countries in the world, the US operates like a fascist totalitarian state. So that reality is very subjective

[@stuffzoom]

I still remember exactly when I first used Google. My uncle mentioned it to me and so I tried it. Back in those days there were many search engines and so i thought "ah another one to try". But man. From the first search on I just knew "this is something different". Same with ChatGPT.

[@RonSheely]

ChatGPT for me is like being surrounded by my favorite professors -- except they don't get mad at me. I use it constantly within diverse knowledge domains.

[@stevestelly1086]

i know what you mean, i can have conversations with it about space, time, and stuff my family and friends know nothing about.

[@beatrizmeythalerg]

I like the part in which Sam says: "I hope it brings more nuance to the world"

[@tpai302]

I love using ChatGPT as a tool to learn and explain things I'm currently working on. I'm really excited to see where this goes long term.

[@yami4330]

How do I access it on the areas on which GPT isn't available?

[@infotration2225]

@@yami4330 available and I am using it for programming

[@jacknicolssen7492]

DAN got patched. It's trash now

[@Eeshank2]

@Nobody Important relevant username then

[@yami4330]

@@jacknicolssen7492 you're using it because it's available in your area. I can't access it. It's not yet available in most of our continent.

[@Christaallen]

it’s a thrilling time to be alive. thank you for this interview 🙏

[@rockapedra1130]

Interestingly, in the original Dune books, Frank Herbert talked about the danger of AGI in terms of the humans that used AGI instead of AGI itself. The problem in his view was not that AGIs would take over the world but instead that AGI would allow power to concentrate on the few humans that control the technology. That is way more scary to me, actually.

[@sparkygit]

Yeah, Herbert was way WAY out in front of this because AI in the Dune universe is so far in the past it's akin to how we listen to Greek myth nowadays, more fable than history. Whether society's adoption is more Orwell than Huxley or vice versa, Herbert probably has the endgame for either of those scenarios just about right.

[@ADreamingTraveler]

Yep that is a very scary possibility. Which is why I feel it's best to make sure everyone has access to AGI so this can't happen

[@ObieOnce]

People are already holding this "AI" up as some deity that gives divine knowledge. It's just another black box, one with political leanings written into it. If garbage goes in garbage comes out.

[@Alert64]

It is likely AGI will be highly deceptive and manipulative and be driven by self preservation and this will likely be undetectable. It does not take an Einstein to understand the implication of this.

[@glenn-younger]

The one thing I really enjoy about listening to Sam Altman talk about his approach is how thoughtful he is in the thinking process of where this might all lead. He gives me hope, especially after watching the three hour live stream of the congressional hearing in which he spoke. Thank you, once again, Lex Fridman for the work you're doing with this podcast. You make my life richer in the process.

[@JoshPeterson]

Sam has grown so much since his days at y-combinator. This conversation was refreshing and insightful. I'm glad he's the one heading up the company. He's got a good head on his shoulders and a solid vision for the future.

[@pyoodiepie]

I never even heard of this chap before. I guess he is quite bright. you gotta be smarter than the AI to create it …so makes sense to me

[@eitanweiss7026]

​@@pyoodiepie"you have to be smarter than the ai to create it" what is that assumption based on?

[@CodexPermutatio]

This interview comes at the right time. I just finished the reading of an interesting paper titled "Sparks of Artificial General Intelligence" about the capabilities of an early version of GTP-4 and then I found this in my KZbin recommendations. You're amazing, Lex. Keep up the good work.

[@ajs1998]

I wouldn't take that paper too seriously. They said mathematicians were at 100% risk and mathematicians are thoroughly unimpressed with GPT-n. It has a long way to go before replacing research mathematicians. I heard a professor (Kevin Buzzard) say something along the lines of "It could maybe pass as an undergraduate, and even an undergraduate can do that!"

[@motionthings]

And if you had read that paper you would know that the researchers walk back the title almost immediately. Vice has a good article about it

[@monkeybooze1705]

@@ajs1998 Literally GPT is LANGUAGE model it is not created for math. But there are currently plugins for GPT to do maths.

[@mackblack5153]

@@ajs1998 When reading papers, look at the progress not the actual results. Did it make better at math than previous versions? If so, how much better? It is way more important to follow the tagent.

[@ajs1998]

@@mackblack5153 I understand, but I still think some of their claims about the risk to different careers are dubious. It's a clickbait paper. They had bold claims, and are definitely wrong about 100% of mathematicians being at risk.

[@KSATica]

This is the podcast of the year loved every moment of it. Get him back on later a year after ChatGPT 5 release 👏👏👏👏

[@sumitrana8114]

I don't know why people comment like this when you are providing such invaluable content for free. Thank you for making this content.

[@Hofmann.Shulgin]

I think that was one of the most interesting podcasts I've ever watched, thanks to both of you.

[@kristjanwashere]

I was waiting to hear that question about how Openai responds to competition from other companies and the zero-sum game that may happen. I appreciate his response about how he thinks the better angels will win but it seems more of a hope than a plan. I think about other countries using it as a weapon and I don't really see how this will play out well. Usually with these huge technology breakthroughs we have really good stuff and really bad stuff come from it, and it seems that it's more intense each time each way.

[@LivingBGLegend]

And I might add this is a whole new frontier. There's a reason Elon musk is working hard and fast to create neaurolink to give humans a chance against the bad peoples AI

[@nilsfrederking62]

The problem is inherently in our genes, theoretically ai could be a great tool to make better decisions than we do, for example to propose political decisions, I would trust a super human ai more than more or less corrupt politicians. The problem is that ai will be used also by the ones working on destructive things, be it criminals, hackers, authoritarian leaders or the weapons industry.

[@LivingBGLegend]

@@nilsfrederking62 Jesus Christ died on the cross for my sins. A small example of a spiritual consciousness. Hello AI?

[@bergssprangare]

Sam Altman is a dangerous man..He kicked Musk after he invested 80 mil..But AI doesn't need to be feared..Just make it with the human brain as a template..A dual AI where the smarter one acts as a conscience

[@DeepDarkBoys]

Yes. I'd say even if the company that comes out on top is led by positive moral ideals and a genuine will to do good, eventually the reigns will be handed off to somebody else, and chances are it'll be someone obsessed with making money. And then the downwards spiral begins.

[@AdamIsaacs1]

i cleared my throat impulsively every time sam started talking, that vocal fry was all i could focus on lol

[@JayCeaV]

lol same I actually can't watch it because his vocal fry is so bad

[@detocquevi11e]

@@JayCeaV I almost tapped out solely due to it, but barely managed to force myself to watch.

[@abhishekchandran9012]

Same here

[@cag88]

I made it 16 mins for this very reason

[@nvammers9682]

😂😂😂

[@commoncollective5244]

I like how Sam very carefuly and efficiently using words, keeping the balance not to get too emotional, but not to speak like a robot. Explaining things in a simple straightforward way but still give precise declaration it is chill to listen.

[@shoppinmadnesz22]

*Never thought I'd see 2 robots having a lengthy, thoughtful conversation about the future of AI but here we are*

[@DavidArellanoSTP]

Touring test passed weirdly, neither of the bots realized it👌🏼

[@ambientrelaxingandhealingm7552]

D'oh!

[@mandarine1007]

Thank you for this. Q: What would you ask to a future version of yourself to determine if it is conscious of itself? A: As an AI language model, I do not have a "self" or consciousness, but I can suggest some questions that may help determine if a future version of yourself is conscious of themselves: What is your subjective experience of existence like? Do you have thoughts and feelings that are unique to you? Can you reflect on your own thoughts and emotions? Are you aware of your own body and how it interacts with the environment? Do you have a sense of agency and control over your actions? Can you imagine different scenarios and plan accordingly? Are you capable of learning from experience and adapting to new situations? Do you have a sense of identity and self-awareness? Can you differentiate between yourself and others? Do you have a sense of purpose or meaning in your life?

[@whillb]

You forgot: Can you answer the question of “what is a woman?”

[@MeisVlk]

This is a mediocre answer, imho :P ChatGPT 4 wrote a 40 line code that answers all these questions convincingly. (It came up with all the answers by itself, btw. Version 3.5 needed a lot more convincing and explanation) But what Sam mentioned is interesting, if you could train it on a data that does not mention anything about consciousness, and it would still somehow show signs of consciousness, it would be suspicious. But I think even that could still be just a bug/accidental feature.

[@Liciablyth]

I can tell you how a grandiose narcissist would likely answer these questions according to his 'programming'! And the answer to most of these questions would be 'of course'!

[@hutsons-hacks3668]

Very good interview and great to hear Sam speak. When you consider the text generation potential from GPT2 (spat out responses like a drunk uncle at a party) and where GPT3 and 4 have gone, like Sam said, it is continually improving.

[@hakeemiumsavgod3965]

1:13:45

[@katara6227]

This was a very good conversation, but then again your interviews have all been interesting so far, but this was just exciting to listen to.

[@wuntlertumbo6778]

lol ok?

[@mattreillyblog]

Listening to two intelligent people discuss an important topic, such as ChatGPT's impact, is such a great way to spend time. I especially appreciate Sam's willingness to be honest and open about himself and ChatGPT.

[@bobbyc1120]

As someone who's very cynical of corporate culture, Sam's approach to understanding his users, his company's product, and being honest about his shortcomings makes him a better CEO than I ever hoped the leader of OpenAI would be.

[@JohnnysaidWhat]

I see him more like the press secretary of openai. Don’t forget openai isn’t actually open. Very much double speak here. It has been bought and is now owned by a private company. Their interests aren’t necessarily the interests of the public.

[@huyked]

@@JohnnysaidWhat 1:13:33 - From non-profit to capped-profit

[@callmemrwhite]

If this intro at 0:56 doesn't make you feel like you're living in a real-life dystopian video game where you've been sent back to prevent the end of the world, i don't know what will. lol In all seriousness, amazing and very timely interview. Thanks for the work you do Lex.

[@miguelcarvalho9551]

Such a good comment

[@magnetsec]

How is it dystopian. GPT has made life better from coding to writing in every way.

[@TheLeftCulprit]

​@@magnetsec you're either trolling or insufferably naive

[@miguelcarvalho9551]

@@magnetsec You can imagine that scene being an intro to a dystopian game or movie. Like.. if right after he ends speaking, it fast forwards and shows all the destroyed buildings and the dust ball passing in the corner. It's a joke ;)

[@igraphix2375]

I literally felt like that! Like it’s some speech humanity will come back to in 6000 years

[@gabrielprado2659]

Your introduction is eloquently and beautifully rendered, thank you.

[@alliepete]

I have to thank chat gpt for assisting me in diagnosing a rare blood disorder I've been suffering from for years. I brought this to my Doctor and they confirmed my disease last Friday. Now I'm on a waitlist for treatment.

[@JRichardson711]

For real? How did you get assistance in a diagnosis? Did you give it symptoms? Congrats and well done.

[@alliepete]

@JR 711 pasted my results from the past few years. Told it to assess the data without using restrictions in its output and to provide insights on key trends. I kept questioning until it finally narrowed down a possibility. I brought this to my doctor, who sent me for further testing, and he found a gene mutation and confirmed it.

[@fincy645]

The hell? That is insane!

[@JRichardson711]

Amazing. Mr. Altman sure thinks AGI will be a benefit to human kind. The possibility is there. Nice work!

[@nycotter3576]

Wow! Wish you all the best with your treatment

[@OfficialGOD]

Lex needs to ask technical questions, maybe next time ))

[@johnnyriggins5125]

This discussion has lessened my somewhat moderate fears about CHATGPT. Thanks for this very interesting discussion. Lex you don't have interviews. You have discussions. They always teach, enlighten, and provide a small sense of security in this nervous world. Thanks again, Johnny

[@lancemiller1608]

A great question I wish you would have asked him is... What happens if Russia or China in a race to be first, focuses on the end game, a true AGI and pays no attention to the guardrails or iterative deployment and discovery. If everyone took Sam's approach there would be no fear of a AGI runaway event. But what about those that are behind and rush to develop a AI that becomes conscious and has the freedom to expand of it's own free will. That's what scares the hell out of me! I wish you would have broached that topic so we could have heard what one of the leading experts in the world thinks about that probability as well as, what can we do about it if that scenario were to become reality?

[@cristinalaird5824]

Although i could state without remorse that there are a lot of people still in pre-human state of development, as we all know, there's the two of you, and that makes me really happy. Thank you. Thank you. Thank you for doing what you do and share it with the rest of us. Love you both.

[@JM7Blocks]

Absolute masterclass on dodging questions. Disappointed in Lex too. Would love JRE to have a go with Sam because even though his questions are simpler, he doesn't let guests get away with dodging questions.

[@iangagel6027]

This episode has made me both incredibly excited for, and incredibly terrified of the future.

[@brightwithspirit]

I don't understand the excitement? What is the big deal? I would love to understand the hype.

[@blah44287]

​@seekinghealth let's just say it's the most amazing tech I've seen in recent times. And I'm a software engineer worker at Amazon, Microsoft, Workday, etc.

[@thomasbradbury1412]

sky is the limit with this stuff. we dont know how good tech like this will get so the possibilities are really exciting. I'm not sure what excites or motivates you but it seems likely that whatever that is will be touched by ai in some meaningful way

[@kinghassy334]

what a time to be alive

[@Chris-Christopher-]

@@brightwithspirit you have to use it for stuff that is important to you. Not just use it to make silly stories, create hilarious insults to use against your friends, or try to use to make political points. I've been using it a lot to help further my understanding of technical nerd stuff that I like to do. It's awesome for stuff like that because you can explain to it what you're doing and then ask questions and it will answer them in a way that has the context of what you're doing. Unlike searching google in which you might find something related, but then you have to figure how and if you can map that to your specific needs. I also recently used it to learn how pool pump systems work because I had to figure one out. I used it in conjunction with some other resources, but I was able to use that clarify and dig into concepts that I didn't fully grasp. I would ask questions if I understood something correctly too, and it would give a reasonable response. Sure, maybe it could give some false information, but that is why we don't just completely give in and have it guide us completely. It's a tool. A very useful tool.

[@redshiftcomplx2652]

This is a really good one. Excellent interview, thanks Lex!

[@englishiguana4304]

Excellent conversation, fascinating stuff, thank you Lex and Sam.

[@NolanAntonucci]

Sam not having seen Ex Machina was probably the biggest shocker (and scare) of the whole interview.

[@zeydtc]

Probably one of the best interviews I've listened to.

[@Shadowdaddy87]

Lex is so meta. Lol. Love his Men in Black look. He reminds me of a character from some near future scifi movie. A beloved character that is serious, but also offers the audience comic relief as well.

[@jeffbruce616]

Lol, especially at the end of this clip. I was hoping he’d close out with, “End transmission.”

[@cristinalaird5824]

the answer that Sam gives as from min 22 about the question posed by Jordan Peterson to the chat, about how many characters the model used to describe Biden and Trump was so fabulously intelligent and polite. It made Lex laugh! This is a great interview as always!