# Its Time To Talk About The Truth Behind OpenAI...

## Метаданные

- **Канал:** TheAIGRID
- **YouTube:** https://www.youtube.com/watch?v=40sxm_CLYn4
- **Дата:** 01.10.2024
- **Длительность:** 40:35
- **Просмотры:** 21,210

## Описание

Prepare for AGI with me - https://www.skool.com/postagiprepardness 
🐤 Follow Me on Twitter https://twitter.com/TheAiGrid
🌐 Checkout My website - https://theaigrid.com/


Links From Todays Video:
https://www.theinformation.com/articles/behind-openais-staff-churn-turf-wars-burnout-compensation-demands?rc=0g0zvw

Welcome to my channel where i bring you the latest breakthroughs in AI. From deep learning to robotics, i cover it all. My videos offer valuable insights and perspectives that will expand your knowledge and understanding of this rapidly evolving field. Be sure to subscribe and stay updated on my latest videos.

Was there anything i missed?

(For Business Enquiries)  contact@theaigrid.com

#LLM #Largelanguagemodel #chatgpt
#AI
#ArtificialIntelligence
#MachineLearning
#DeepLearning
#NeuralNetworks
#Robotics
#DataScience

## Содержание

### [0:00](https://www.youtube.com/watch?v=40sxm_CLYn4) Segment 1 (00:00 - 05:00)

so I think it is about time that we actually talk about what is going on at open aai the company has had a tumultuous season with many of their senior researchers and leaders departing for a variety of different events in this video I'm going to do an extensive Deep dive on what's actually going on at open AI I know that most of you know a few things but there's been a few articles released in a few days in addition to miror morati leaving which is why I'm making this video because there is a lot of new details that reveal exactly why open AI has kind of fell apart in the way that it has so let's not waste any time and dive into exactly what's going on with open AI so recently there were two articles that were released by Vox by The Wall Street Journal and by the information so we have a plethora of news organizations releasing their own proprietary information with regards to what happened to openai it was honestly surprising to read all three extensive articles but they showcase clearly what most of us already know open AI has exploded unexpectedly and with that explosive growth they're having a lot of teething problems with the company and it does bring into question how far this company will succeed and how their future might be so you can see right here we have this article that said open AI As We Knew It Is Dead this one was basically referring to the fact that open ai's initial promise is no longer the propess to actually build AGI safely and of course we also have this article here which I'm going to dive into as well which dives into behind open eyes staff churn the turf wars the burnout and the compensation demands a variety of different things that are really eye opening including some tidbits onto why elasa left and how they were trying to get him back in fact that's actually from The Wall Street Journal article which I'll talk about in a moment so one of the things that this article actually speaks about is that it speaks about samman's management style it talks about how the board had you know enough of the Rumblings of the aico about his management style and one of the things that did become apparent I think it was earlier this year was that Altman allegedly operated how he operates is that he has a tendency to pit other leaders against each other so this is something that is of course not really great for the workplace I'm not exactly sure why this has happened but if you were someone that has paid attention to this you'll remember exactly I'm not sure if it was last year or this year but we had a situation where we had people like hel andona coming out and basically saying that look samon was lying about this he was lying about that um this is not true that is not true and just a variety of different pieces of information that were really incredible regarding the true details with the board of open Ai and of course you could see here the members of the board worried that if those Tendencies left unchecked could her opening ey's ability to rain key researchers and Executives and of course A year later those concerns appear to still be here and this is quite surprising because you think that like a year on from now alman's management style potentially would have changed and we wouldn't have seen such a mass Exodus from Top tier Talent from open AI if you're really struggling to grasp this you know situation how bad it is if we look at the you know past 12 months you know several of open AI leaders have quit the company including you know the key Executives and I got to be honest Mira moratti leaving you know on Wednesday which is only two days ago was incredibly surprising and Bob mccrew and Barett Z who work and you know run the team that prepares new AI models for release known as posttraining also said that they're leaving which is it's pretty insane to have key individuals of your company leaving as the company is seemingly doing so well on the exterior I mean we do have to take into account that Mira's release was quite unexpected because I know most people would think that you know okay we're Outsiders we don't know exactly what's going on in the company but Sam Alman also put it out there that mirror morat's release who is the CTO well former CTO she actually said that she was leaving in the morning so this was something that did actually come as a shock to samman pretty much you know a stunning Revelation because if you're the CTO of a company usually you're going to be saying you know 6 months in advance at least 6 weeks in advance you know because you have such an integral role it's not like you just up one day and say okay I'm going to leave so it's quite interesting and of course I guess you know Amir morati did state that when she was leaving that hey look um I'm going to be leaving because I don't want it to be leaked and at a time where open ey isn't on an upswing with recent releases

### [5:00](https://www.youtube.com/watch?v=40sxm_CLYn4&t=300s) Segment 2 (05:00 - 10:00)

so this is the overall picture I know I spoke about this in another video but I wanted to show you guys this again because I think it's really important to drive home just how insane the situation is at Open Eye in terms of people leaving I already spoke about this you know that like nine people have now left the company I mean it's probably like 10 or 12 I mean in the article it says that there are actually 20 so this is only some notable figures and the crazy thing about this is that like when we actually take a look at the people who have left open AI when we look at them they're actually really important individuals like having you know Leopold Ash Brer leave you know someone who's on the super alignment team Greg Brockman take a sabatical which I'm going to dive into later because there are some stunning Revelations about why Grog Brickman actually you know is on a sabatical and there's also some stunning Revelations with regards regarding Ilia satova which I'll also talk about later but I think all of these people like John schan you know um Andre Kath the AP Ashen brener you know the head of developer relations you know Ilia Sask andir moratti I mean like if we were you know looking at any other company I think we would say that yes this is definitely not looking good at all especially since we've heard you know terrible things about alman's character about greed about power about you know all of these different characteristics that kind of give us this sort of Mystique in the sense that we don't know whether we can trust him that's not the kind of you know CEO the people would want to be working for it is really uh you know just this company that seems to you know excel in terms of what they're doing but also the kind of management style the kinds of things that are going on it seems like the company is going to just entirely implode soon if things don't make a change now thing about this is that um you can see right here is that you can see that Alman said Wednesday in the memo that he would you know is he's going to be involved in more technical product parts of the company and after primarily focusing on the non-technical aspects such as fun raising government relations and business Partnerships with Microsoft Apple and others and of course you can see the company's technical leaders who previously reported to morati mcru will now report to Alman so we have a you know a pretty you know abrupt leadership change there and you can see right here he says I obviously won't pretend it's natural for this leadership change to be so abrupt but we are not a normal company and it seems like right now the samman is clearly going into this damage control mode so last Thursday in an event which I'm about to show you guys Sam mman seemed to downplay the significance of Wednesday's departures like I said before your CTO leaving randomly is not like it's not like uh something that you can downplay that is a huge event like CTO is a huge it's not like an AI researcher that go upset and left even though that's still a big thing your CTO some you know in the head of management you know those kind of decision-making you know areas it's not something you can really downplay and he actually said that I think this will be hopefully a great transition for everyone involved and opening eye will be stronger for it as we are for all of our transitions so I'm going to show you guys that clip now but I don't think you can really downplay your CTO just leaving randomly when so many people have left before it just looks really bad thank you for being here with us son it's thank you all for coming thank you for having me it's been a eventful 24 hours we're used to those uh at open AI but it has been yes so some of our top uh Mira who's been our CTO uh and Bob our head of research decided to move on from the company um I'm tremendously grateful for that they've done and I you know they actually both of them have been working at open AI full-time longer than I have I was still doing why combinator work and you know they've done they've gone way above and beyond the Call of Duty for open AI they've done tremendous work and one of the things that I love is helping to develop talented people and then if they want to retire do something new uh I'm always excited about that and I think that's sort of a key part of the Silicon ballot ecosystem in addition to being tremendously grateful for them I'm very excited to uh Empower a new generation of leaders Mark Chen who's gonna take over Bob's job uh I think it's absolutely tremendous and really excited to work with him excited to flatten out the orig and get to work more closely with the other technical leaders I have not been as involved in the tech uh recently as L were things because there's been so much going on I'm excited to do that and you know I think this will be hopefully a great transition for everyone involved and opening I will uh be you know stronger forward as we are for all of our transitions um I saw some stuff that this was like related to a restructure that's totally not true uh most of the stuff I saw was also just totally wrong but we have been thinking about that our board has for almost a year independently as we think about what it takes to get to our next stage but I think this is just about people being ready for new chapters of their lives and uh a new generation of leadership so you can see there exactly what Sam Alman said now there's also some other things where you know we can talk about burnout you know you can see that it's quite incredible how much these people have had to work to extreme degrees you can see here it says its

### [10:00](https://www.youtube.com/watch?v=40sxm_CLYn4&t=600s) Segment 3 (10:00 - 15:00)

researchers are grumbling about how much compensation they should get and Alman is pushing the companies teams to turn you know breakthroughs into public products as quickly as possible causing stress as employees work nights and weekends to launch the product on short timelines several employees said and whilst yes I guess you could say that this is what comes with working with you know a tech company especially companies like open AI you know the days of you know just having a tech companies where you have these bean bag field rooms and you have like you know B basically adult daycare and you're paid like quarter of a million a year those days are over I still think that this is rather incredible because I think what we're looking at here is you know a situation where you have like a lot of these AI researchers and those are kind of like different employees than two traditional Tech employees that are used to like working incredible hours and you can see here that this kind of culture you know of working you know doing so much stress it just you know it's difficult because on one hand you have you know traditional company culture which is of course you're supposed to work hard but at the same time you aren't supposed to get people burnt out to the point where they hate the working culture and they leave the company and you can see here that those pressures have also actually worn on leaders like moratti and mcrw said people who work with them so those things like that kind of culture of course it's AI it's going to be really competitive as you know it moves as a million miles per hour but those people they just you know it's just not for them and then of course you can see that they've ending up leaving which is incredible so um actually a little bit surprised that they're having to work so much because what I would have initially thought that is that you know open ey is so much further ahead but from this if they're working you know weekends and you know trying to rush stuff out the door and I'm going to show you guys later on in you know a few articles the open ey might not be as ahead as we think that the door between open Ai and other companies is slowly closing and they've had a crazy lead with gbt 4 but the thing is now is that they really need to extend that lead and to maintain that lead it's getting harder and harder because so many other companies are also now in the mix now we also have um ilas asov which is rather fascinating because he's been up to some really incredible things of course if you remember he went ahead to raise a billion dollars for his opening eye arrival and this is of course you know part of the reason that open AI has had trouble as well now another thing that I wanted to talk about is of course the fact that when we look at open AI we do have to place this company into some context this isn't me making excuses for open AI but I think it's always important to remember as objective as possible when you're analyzing certain situations so for example with this company we have to remember that they've gone from you know a nonprofit to a, 1600 plus employee business increasingly focused on its you know increasing its revenue and that is pretty incredible because the company has grown so much in such a short space of time the company is of course going to have people that aren't you know I guess you could say fonded of this certain change and some people may wish to leave the company now one thing I want to tell you guys and this is something that I need to reiterate something that people tend to forget is that when we look at how chat gbt came to be it was something that was quite remarkable because most people at the company didn't realize how incredible it was going to be for example I picked up this old article from 2021 or 2022 or 2023 late 2022 and you can see here that you know individuals who were at the company at the time basically said that this is incredible okay you can see that Jan L former member of the super alignment team said it's been overwhelming honestly we've been surprised and we've been trying to catch up John schan said I was checking Twitter a lot the days after the release and there was this crazy period where the feed was filling up with chat gbt screenshots um but I didn't expect it to reach this level of mainstream popularity you can also see here that he says I think it was definitely a surprise for us all how much people began using it we work on these models so much we forget how surprising they can be the outside world sometimes of course Liam here says that we were definitely surprised how well it was received there been so many prior attempts at general purpose chatbot that I knew the odds were stacked against us and Jan like here even says I would like to understand better what's driving all of this what's driving the reality we don't understand we don't know and the point of this is to like frame things so imagine you have like a you know a small company research lab that's focused on developing safe AI you've done a few things research products before you've made like you know a few AI agent games and those kind of things that have gone on to do remarkable you know in their own respective Fields but imagine you just making the chat bot and then the thing is that they put it out as like a small research preview it was just a website that didn't expect to get a ridiculous amount of traffic and then boom you have this insane you know thing to where it's like the fastest growing app with like 100 million users in like I'm not sure if it was two weeks but it was really small so we have an insane amount of people on that app in

### [15:00](https://www.youtube.com/watch?v=40sxm_CLYn4&t=900s) Segment 4 (15:00 - 20:00)

just a short amount of time so that is something that we do have to take into account because that is just incredible with how much they blew up now when you look at the fact that okay this company's just blown up like November 2022 wasn't that long ago and since then the companies it's been like thrust from you know 10 m an hour strolling down the street to like 1,000 miles per hour being valued at billions and billions of dollars so this is something that you know has like you know changed the entire I wouldn't say ethos of the company but you know dynamic of how the company kind of operates cuz previously if you were re a researcher there you would have you wouldn't have had you know product deadlines you would have been able to work on things I wouldn't say out your discretion but like the main focus was you know you're a nonprofit you're not really a company that's making money or just trying to figure out the AGI problem solve it and share it with everyone essentially we can see here this is where things get more interesting you can see that of course recent claims surrounding Sam alman's Equity where he was predicted to get around $10 billion as the % of open's valuation $150 billion would be around $10 billion you know some people are saying that this is ludicrous he's basically saying that isn't true at all and I don't know whether or not it's true or not there's many articles stating that he's going to get Equity he's stating that he's not either way I think if he does get Equity it's going to look really awful because this all this tracks back all the way to when open a was initially created I mean um for example you can see here some someone said how to make $10 billion raise $50 million from Elon Musk to start a nonprofit then tell everyone you're doing this for the sake of humanity and raise billions and then you can see that Elon Musk here says that this is deeply wrong if you aren't familiar with the backstory Elon Musk is the one that helped create open AI with $50 million of his own money to start the company and then of course on the back end of things he was kind of like shut out had disagreements with Alman and now Alman is getting $10 billion of a company that you know was previous a nonprofit now this is why it's such a bad issue I know I played this before but I'm going to play it again it's only 15 seconds but in this interview not an interview was like in this Congress hearing he was basically saying that look I don't have any equity in the company you know I'm only doing this because I love it you make a lot of money do you I make no I paid enough for health insurance I have no equity and open AI really that's you need a lawyer I need a what or an agent I I'm doing this cuz I love it so yeah um when you have people saying that okay he's just this you know uh Angel who's like okay I'm just doing this because I love it I just want this company to succeed you can see here that you know the article from Reuters which is right underneath it says Sam one will receive equity in the first you know first time ever of course people are like whoa this guy literally last year said that he didn't want it so when people now like seeing this it doesn't look great because there have been numerous accounts of people saying that Sam Alman is mellian you know he's like um I don't know what the word is it's Sly and cunning and deceptive like that just looks awful especially if you're saying that in the public and then it gets leaked that you're going to be receiving Equity um I just think it's just losing trust in in the public eye and this is not something that you want especially with a powerful technology as AGI now Sam mman isn't the only one at open AI who's having these kinds of issues um there's also an equity problem with you know individuals you know and people who are at the company you've got for example you know compensation demands and all of these things for example um you can see here that there's comp ation demands um it says that another unappreciated factor in the leadership turnover is the persistent demand of key researchers for more compensation as open AIS and valuation have sought ex employees have already cashed out more than 1. 2 billion dollar from selling their profit unit so these compensation demands basically refer to you know asking for higher pay better Financial Rewards or in the case of open AI as the company has become more and more valuable and its reputation has grown key researchers um and key employees have been basically pushing for better compensation to reflect their contributions now unlike many traditional companies that offer you know stock or equity which means employees can own a small part of the company openi doesn't actually have typical stock options instead these employees actually have profit units which give them a claim to portion of the company's profits if and when open AI becomes profitable and in the past few years open AI employees and former employees have been able to sell those units to other investors and they're basically trading their future potential profits for immediate cash and as opening eyes value has continued to rise employees see the company's increasing success and they basically asking for even more compensation to match this immense growth and now of course you have you know the pressure from these demands and it's becoming a factor in the company's leadership changes because managing these Financial expectations making sure everyone is satisfied with their compensation is a major challenge

### [20:00](https://www.youtube.com/watch?v=40sxm_CLYn4&t=1200s) Segment 5 (20:00 - 25:00)

for a company like open ey that is still evolving and doesn't yet operate like a typical business with of course these steady profits that they will have you can literally see here that um you know uh they've been in the thick of dealing with people who are you know threatening to quit over money and other concerns so I mean the problem here is that Guys these researchers are like Allstar NBA sports stars like they're like the top talent you know when you're seeing you know I'm not sure if you guys watch sports I don't watch sports that Mar but every now and again I'll see a screenshot of like a major you know soccer style athlete that's traded for like $100 million that's kind of like what's happening here I don't think they're like you know the transfer values that much but essentially what you do have is that these researchers are really integral to the company like these people that you have at the company they are the ones driving the research and Innovation behind the product so you can't have the company without them so they actually have a lot of power um and due to that they're they're really using their power and their leverage to try and get more for themselves which is understandable because if you're the one you know like a faceless person behind the scenes driving the valuation behind the company you're going to really want to make sure you're valued for your time now craziest thing about this is that um of course you can see open AI have uh brought on some new highest they brought on Sarah Ry former CEO of publicly traded next door to run the company's complex finances as its first Chief Financial Officer and Kevin whale a longtime product leader of the companies such as Twitter and Facebook to oversee its consumer and Enterprise Products now that's really good because of course we want to make sure that you know open your eye you know as people are going out the door you want to make sure that people are coming in the door because of course you know if you're just continually losing people doesn't seem to bold well now we've got Turf Falls here you can see here that it says several people who work directly with Alman and this is what I was speaking about before or his Lieutenant said that his penchant for allowing Turf War to fester and his avoidance of tough decisions When leaders have asked for more hiring or resources have been factors for the departures so it says that other leaders such as Brockman and moratti have sometimes had to make difficult calls Alman has avoided so this is one of the reasons why people have left as you can see right here Alman I don't know exactly the details of this but as they said before maybe people have been pitted against each other it's extremely vague maybe Alman hasn't been making certain decisions and he's been off like around you know the world doing these Global tours and interviews and making the deals for chips and stuff like that whilst Mir marati um and Greg Brookman are actually at the company we can see that you know alman's having to you know cut his world tour short because you know if you have your CTO leave and Greg brookman's on sabatical you know you're going to have to like be like okay let me actually focus on making sure this company is great because what tends to happen and I read this in a book one time and it was about um basically what happens to uh you know certain individuals is that they will have a lot of success so for example what's happened with open AI is that um you know you build chat gbt you have a lot of success and then what happens off the back of that success is that of course some one's the CEO he's going to be invited to interviews they're going to be doing blogs there's going to be all these AI events around the world he's going to want to travel make deals all of that good stuff but the thing is that if you get distracted by you know making deals going off and doing interviews going around the world not actually focusing on the company you're going to lose the entire company that brought you that amazing lifestyle it's basically like when you have an athlete that is and and honestly I'm not trying to do too many athlete references but if you have like a you know Superstar who's famous and is flying around the world after being the best at their sport but they then are not focusing on the actual sport that they're playing and for entrepreneurs like Sam Alman this is business the company falls apart and then everything goes away so the point here is that you know Alman is having to you know come back to the company and actually lock in focus on actually building a great product and great things that people actually want because it's all well and good going around the world and you know having these kinds of interviews but you have to make sure that you're there you know in person ensuring the company is completely smooth because Greg brockman's gone ear tusk's gone you know Mir moratti you know someone who was able to do all these things you know she's gone and you know several other key individuals are gone so you have to make sure that like you know this company is going to do extraordinarily well now of course you can see um like I said before moratti and Brockman have actually clashed with each other over the company's AI development plans you can see right here um one of the reason for the friction was that Brockman despite his lofty title was also an individual contributor to project so if you don't know Brockman is like he's not like um samman at all he's not like he's actually like quite the opposite he's like a guy who will be in his room coding 12 hours a day it's actually like a meme on Twitter how like locked in uh Greg Brookman is and when I say locked in I mean like how focused at work this guy is because he will literally just

### [25:00](https://www.youtube.com/watch?v=40sxm_CLYn4&t=1500s) Segment 6 (25:00 - 30:00)

tweet about coding all the time um and he's someone who's very Hands-On um I guess you could say you know later on the article you're going to see uh it's almost too much so you can see right here that one reason that the friction was that Brockman despite his lofty title was also an individual contributor to projects for example an effort to take early AI research and turn it into working prototypes or products that could handle lots of customers that at times puts him at odds with tech leaders like morati who managed the scores of researchers and software developers working on core AI models those staff sometimes had their own plans for turning work into products that conflicted with Brockman so we have this you know issues where we've got Brockman that is very Hands-On he's very technical and he wants to turn you know his you know Research into working prototypes you know that could handle customers and then we have other software developers in the company who are you know basically you know the they're just not agreeing they just there's no like Synergy there so um this is actually part of the reason that you know Brockman actually took a leave because there was just so much friction between them and other teams so it was really hard to get um them to like kind of focus so it's just it's just incredible like this stuff is going on at such a major company but like I said before this is why you know Brookman is very Hands-On he's very technical um and this is you know something that uh you know kind of caused certain issues within the company and you know I'll show you guys this now basically um as you know as we're talking about Brockman the kind of problem was that you know Brockman um didn't have direct reports but of course um he tended to get involved in any projects he wanted often frustrating those involved according to current and former openai employees they basically said that Brockman um this is why he left you know he demanded last minute changes to Long plann initiatives prompting other Executives including morati to intervene to smooth things over so you have Brockman um disagreeing with certain employees about you know certain things and certain products how they're being deployed what's being offered and this is of course causing internal conflict which isn't great you don't want to have internal conflict at a company where you're trying to push the needle forward I mean it's quite an issue and you can see right here that this caused um so much you know issues that um staff has urged Alman to reign in Brockman saying that his actions demoralized employees and that those concerns persisted through this year Alman and Brockman agreed that he should take his leave of absence so basically Greg Brockman right now is on a sabatical because I'm guessing that you know the issues that they've been facing on the company where he's just you know consistently wanting to get involved in certain things you know it just wasn't working and I think that's incredible because we have a situation where someone that wants to you know do certain things in the company I don't know the exact details but um he he's of course left a sabatical so it will be interesting to see how this company manages moving forward and the kinds of things that you know they get up to because I think openai really has some work to do in terms of their organization skills and so far I mean it doesn't look great and I really hope Greg Brockman comes back because if he doesn't I don't think open AI is going to do well for the foreseeable future now in terms of course people who left you could also see that the chief research officer at openai um actually left okay you can see that moratti and mcru exits could have wide ranging impacts on the company and standing in the air race as it develops you know its large next Flagship language model Orion so the crazy thing about this is that these leaders were crucial in coordinating efforts for open AI research and applied organizations moratti for instance was known to settle disagreements between the company safety and product teams during launch processes and mcru okay approved requests from researchers for additional Computing resources a former employee set and both leaders were involved in decisions to prioritize certain lines of research and to abandon some research products so these are critical areas guys like um compute was literally and the crazi thing about this is that like um you know uh approving requests from researchers for additional Computing resources this was also something that was like um really largely one of the reasons that Jan like actually left the company because he just said that there was none approved and you know I'm trying to do research here and I'm not able to do it and of course if you know you don't have someone that's approved Ming the the Computing resources or someone that's actually you know managing this kind of stuff things really start to fall apart especially if you also have someone who's you know settling disagreements between safety and product teams during launch processes that is of course one of the biggest things because you have people that are trying to accelerate with the company then of course you have the safety people who are like guys uh if you release this product it's going to cause this amount of damage we have to you know be careful managing that balance is quite intricate because you also want the company to succeed but you also don't release a

### [30:00](https://www.youtube.com/watch?v=40sxm_CLYn4&t=1800s) Segment 7 (30:00 - 35:00)

also don't want the company to release a product too early and then it results in complete catastrophe so that's something that is you know remarkably important now I'm going to show you guys some more stuff because there's so much that's just been revealed um of course another thing and I didn't even know this but Ilia Sova has actually been poaching employees from open AI you can see efforts by Rivals such as il's new startup you know to recruit open ey Talent have exacerbated these Talent crises the hiring attempts have prompted open leaders to make generous counter offers to retain researchers said another person directly involved in a recent negotiation guys so what we have here is basically a situation where ilov of course as I mentioned before he left to start his own company save super intelligence but they've been you know offering people who work at open AI you know generous compensation packages to come and work at safe superintelligence and I can't imagine that with how openi has been in terms of you know playing in terms of certain areas this is you know no surprising that some of them may have taken that opportunity because safe super intelligence I think they got um a billion dollars at of $5 billion valuation I mean it's not hard to just pay a few of those employees a million dollar $2 million a year to come and work for a new company when it's a really small company and just have a lot of the other money you know going towards Computing resources so this is rather surprising because it seems like you know in this area it's a doggy dog world you know Ilia satk is going after opening eyes employees which is you know a remarkable uh thing like I honestly didn't even know that this was um a thing that was going on and crazy crazy stuff okay is that we also have the uh the non- dispatchment agreement so let me actually find the slide well I can't find the exact slide but I will move on to talking about the corrupt nature of open aai which some people have spoken about so others including AI scientists that have been with the company for years believe infusions of cash and the prospect of massive profits have corrupted open ai's culture and this is something that like I said before when I was speaking about at the start of the video you have a situation where you have a research company moving to a for-profit company once money is involved things always inevitably go wrong because certain individuals are greedy certain individuals are corrupt and the problem is that you never know until the money starts to show so um I kind imagine what would happen to a small research nonprofit turning into a multi-billion dollar tech company that is focused on you know making money in insane deadlines it's incredible and with these insane deadlines what I'm want to show you guys is something here is that like this has led to an incredible arms race between Google and open AI you can see right here that open AI have been rushing their launches surprisingly so it says this spring tensions flared up internally over the development of a new AI model called GB GPT 40 that would power chat GPT and business products researchers were asked to do more comprehensive safety testing than initially planned but given only n days to do it guys safety testing usually takes a long time like usually safety testing takes like 6 months which is why I'm guessing that GPT 5 isn't here but um of course GPT 40 I guess they just wanted to completely rush it out because they knew that other companies had models in the bag which is why I said is developing a lot quicker than we would have had it before so they were asked to do you know safety testing you know than they left initially planned but nine days guys nine days is insane you're probably going to have to work 16 hour days just 24/7 like until that gets done because trying to compress all of that into a short time frame is pretty much borderline impossible you can see it says Executives wanted to debut GPT 40 ahead of Google's annual developer conference to take attention from their arrival and this is a true thing when Google has their events going on um opening ey always tries to um you know upshow them or outdo them which is uh ridiculously uh hard to do when it comes to Google so they've got to like pedal to the metal to make sure that like they can uh you know find out exactly what's going on and release something immediately and Google did a similar thing before where they released Gemini 1. 5 Pro now of course you can see right here like I said before the staff the safety staff was worked 20 hour days guys I thought it was 16 hours they had to work 20 hour days and didn't even have time to double check their work the initial results based on incomplete data indicated that gbt 40 was safe enough to deploy so I mean you know but before but after the model launch people familiar with the project said a subsequent analysis found the model exceeded opening eyes internal standards for persuasion the finders the ability to create content that can persuade people to change their beliefs and uh potentially engage in potentially dangerous or illegal Behavior so you can see right here that you know these guys are having to work 20 hour days and not

### [35:00](https://www.youtube.com/watch?v=40sxm_CLYn4&t=2100s) Segment 8 (35:00 - 40:00)

having any time to double check their work this doesn't seem like the kind of scenario where we're heading to a safe AI future I mean right now gbt 40 is pretty safe in terms of you know it's just the fact that it's just a chatbot but I mean in the future like if we don't get some kind of I'm not sure if there's going to be any kind of regulation in the future that you know locks this stuff down because companies kind of see AI as like this weapon so they don't really want to regulate it too much cuz they don't want to fall behind cuz if falling behind an AI is like a year is like 10 years like if you fall behind you're going to be behind for quite some time so um I think this is too far though I think if you're at the stage where you don't even have time to double check your work and you're making the safety staff work 20 hour days I think that is probably moving too quickly um and you can see right here that this rush to deploy gbt 40 was part of a pattern that affected technical leaders like morati the CTO and this is where you know we got so many people saying that where is voice mode um the CEO repeatedly delayed the planned launches of products including search and voice interaction because she thought they weren't ready so the thing here is that I think this is you know a situation where um we have some people at opening eye that just wanted to do things traditionally and then we have some people that just wanted to push forward ahead with everything and you can see that um you know this is one of the reasons that I'm guessing why mirror morati didn't like it because she's delayed the products of course some people are going to say no these things already you can see here that you know the disagreements you know this is what happened so it's incredible it's really incredible how quickly they were trying to move with these products and I keep telling you guys like you know a lot of people say oh my God I can't believe we haven't got this product for months guys you have to understand that like this is Major breaking AI technology like this is you know breaking new ground like the fact that we get multiple new models a year is absolutely insane like these things take months in terms of safety testing post trasting post training data collection um you know SA testing like all of these different things to get a final mod like guys this stuff is not easy and I'm telling you guys that like we are very lucky to be in this space where we're enamored with all of these AI products that can help our lives in so many ways for a cheap price in terms of relative to what we're actually getting and at the speed we're getting them as well it is insane guys like I feel lucky to be alive at this time like and I know some people don't really care that much but I think this is honestly insane so of course you can also see here that um other senior staffers were also so growing unhappy with this and you can see that John Shan you know another uh you know co-founder and top scientist told colleagues he was you know frustrated over opening eyes internal conflicts and basically and I can't believe I forgot this part but that ilasa they actually tried to get ilas S back um they went around to his house they did many different things um and the thing is that John schan was basically disappointed in the failure to woo back El satova and of course was concerned about the diminishing importance of its original Miss Mission so basically John Shan was like okay um of course you know we're doing well yada y but the point is that we don't have ilas S anymore and we should have um at least tried harder to of course make him come back and because we haven't been able to do that I'm currently losing faith in you know opening eyes ability to continue so it's pretty crazy like guys this is actually insane that we're getting this kind of information of course the co-founder and top scientist actually left open AI in August so this is crazy guys is crazy I think what we have here is a situation where you have people that are dedicated about their work you've got these top research scientists Executives and they're thinking okay I'm working at open AI I'm on a high salary or whatever I'm doing well but I'm stressed there's a lot of fast deadlines we've got you know Shady management decisions we've got the fact that this company's growing so quickly we've got the uncertainness in terms of people keep leaving why don't I just you know keep or sell my Equity as you know profit shares and join a company like anthropic where I can actually focus on doing you know effective research I can focus on doing things for the longer term without having to work 20 hour days and ship out products that are unready and basically leave the Titanic because that's why I think that the people who have left open ey probably think it is because I think if you know a company's going to do well in the future I don't think top people would leave at this moment in time sure safety researchers I completely understand it but top execs leaving is just not a good sign especially abruptly that is a huge major red flag in terms of the kinds of signs if a company is going to be doing well or not me personally I just think that is a really bad sign considering Sam Alman was you know caught off guard by this decision so overall let me know what you guys thought about this video I think this video was very well needed because there is so much going on at open AI um and I know this video was a long one I honestly try to speak as quick as possible I wish I could have speak louder currently got a sore throat which is very frustrating when you're trying to make content but nevertheless if you did enjoy this video let me know what you think about the future of open AI I

### [40:00](https://www.youtube.com/watch?v=40sxm_CLYn4&t=2400s) Segment 9 (40:00 - 40:00)

think the future is not great I think that you know they're between a rock on a hard place they're trying to maintain the lead and maintain their researchers without stressing everybody out but you know it just doesn't seem that things are going in the right direction I think openi does have a real good shot because they do have proprietary technology such as their models and the fact that they have the brand image so they will be around for quite some time but if they manage to mess up continually within over the next years and key researchers continue to leave then things could get bad pretty quickly with that being said if you guys enjoyed the video I will see you in the next one

---
*Источник: https://ekstraktznaniy.ru/video/14045*