OpenAI WHISTLEBLOWER Reveals What OpenAI Is Really Like!

So there was a recent interview that was actually quite insightful because it was actually from the OpenAI employee who made stunning predictions on how the future of AGI is going to be announced, implemented on many different levels. The interview was on the Hard Fork New York Times podcast and the person I'm talking about is Daniel Coco Taljo. Now this is pretty, pretty crazy because there are four insightful things and the whole interview is a much longer, but there are four insightful things that I do want to talk about because they actually show us, okay, some very new things that a lot of us didn't know about OpenAI in terms of the actual safety structure that goes on OpenAI. And of course, a few things that actually went on in OpenAI, including a few things on Microsoft's part and his date on when he predicts AGI. I want you all to remember the previous dates that we've spoken about in the other videos. Remember the dates that we spoke about AGI, because it seems that there is a trend from every single OpenAI employee that I've seen. I've looked at blogs, I've looked at posts, I've looked at interviews, and there seems to be one prevailing date that carries on coming up when the mention of AGI. Now, it could just be a coincidence, but I am earing on the side of maybe, maybe not. But, but without further ado, let's dive into exactly what's going on. So in this first clip here, basically surprisingly, he actually talks about something that I didn't even know happened. Okay. And this is like breaking news kind of. Basically he talks about how Microsoft, the kind of safety internal team that they had, they were supposed to basically approve Microsoft's release of GPT4 to, I guess you could say a part of India or somewhere in India. And apparently Microsoft just went ahead and they just released it, which was crazy. So take a look at this because I couldn't believe this. Like if someone, like if I saw this on a forum, I would have been like, ah, it's probably not true, but as someone who's working on OpenAI, I think, you know, taking this at face value, that is pretty, pretty crazy.

And this board had some people from Microsoft on it and some people from OpenAI on it, and it was supposed to deliberate and then eventually approve major deployments on the GPT4 scale or bigger. So it had been set up prior to reaching the GPT4 scale, but it was set up with the expectation that once we hit that level and they're on, it would be something that would gate deployments. You'd have to get approval from the dsp. While the DSP was still deliberating about this, me and various other people started hearing rumors that Microsoft had just gone and deployed it anyway, specifically in India. And so we had some conversations about this. We made inquiries to managers and so forth. I don't want to get into too much detail about who said what to whom and so forth, but I came away disappointed because it seemed like we had this self governance structure in place and there was a violation of the structure. And rather than, you know, holding Microsoft accountable for doing this, we were afraid to damage our relationship with Microsoft, given the sensitivities involved. You know, they have all this compute we're trying to.

So there we have it, ladies and gentlemen. This is, you know, a shocking revelation on the fact that, you know, a lot of people were suggesting that Microsoft's partnership with OpenAI was in fact going to, you know, I guess you could say, contend with the safety people. And you could see here that there was a clear safety agreement that they were supposed to get approval before pushing the model out for public release. However, this didn't happen. So this kind of, you know, oversight by the safety board that didn't happen is, it's pretty, pretty remarkable that this, you know, kind of happened. And it's definitely a revelation that, you know, considering all the things that have happened at OpenAI, maybe it didn't shock me entirely, but it definitely did surprise me to the extent that I think that, you know, know, a lot of people do think that, you know, every single human being and every single organization is just strict, you know, military star professionals that would never make a wrong mistake. But at the time, it's humans who make mistakes and who have incentives and who maybe sometimes are greedy. And what we can see here is that in action that humans are, you know, these flawed creatures. And just because it's Microsoft, it doesn't mean that they might rush out a product if they believe it's going to advance their company's efforts. Which, you know, in this area I think is a little bit, you know, telling for the future of how things might be now, of course, I think in the future with, you know, nationalization of OpenAI, something that has been recently predicted considering the fact that they recently did just hire someone who worked at the NSA for a very long period of time. But overall, I think this is pretty, pretty crazy. Now another thing that he goes on here, and this is pretty crazy because I thought, okay, and this is where he talks about Elias Sutskever and the culture of work after the coup. And basically he says that after the Ilya Sotskova coup, there was this entire culture at OpenAI that was basically fu. Safety people. That is crazy. Like, I would have thought that the safety people would be some of the most respected in the company, but it seems like after that there was this entire thing that kind of just, was just strange.

I guess after Ilya Sutskever and Helen Turner and Dash, after that whole fight, it was like, oh, man, things have sort of polarized. Like, I don't think that we're going to like, talk it out and, and, and pivot. And then also just sort of on a personal level, I think there was a lot of anger directed at sort of safety people such as myself, because the, the perception was that, that the old board was being dishonest and that they, they didn't really fire Sam because he was insufficiently candid towards them, but rather they fired him because they, they wanted to slow down and he didn't and they needed an excuse.

So.

So I think that as a result there was.

So, yeah, that was something that was pretty telling. After Elia Sutskova left, I'm not surprised that the culture around the work actually did change, but considering that there was this entire culture around safety people, that was one that was surprising. But considering the fact that, you know, now looking at the fact that the Super Alignment Team has now disbanded OpenAI, basically, maybe not disbanded the team entirely, but they did actually get rid of the team. Maybe it's like, whoa, what on earth was going on at the internal workings? Maybe this is part of the reason that Jan, like, left. Maybe the reason that, you know, considering the fact that some of the researchers, when we actually think about what occurred, some of them were actually even fired, would that all play into exactly how OpenAI, you know, managed some of their decisions? Of course, it is quite hard to say, but. But it can't be denied that this may have been a contributing factor, especially after the board coup. So it's clear that there was a shifting dynamics in terms of, you know, how safety people were viewed in the AI world. So this is something that I find to be truly, truly fascinating. Now, the next point here is, of course, more fascinating. And this is where he talks about his definition. Now up the top, I've put AGI by 2027, and that's exactly what, what he says. 2027, ladies and gentlemen, is the same timeline year that I've heard from pretty much every, you know, available source of Course, people like Sam Altman have said that, you know, five to 10 years, I think he's become a little bit more conservative considering the fact that, you know, we're of course going to have to, you know, get the hardware side built out in terms of the compute aspect of things. But I think that it's kind of interesting that all OpenAI employees are predicting such short timelines for monumentally transformative technology.

I think that publicly available information about the capabilities of this models and the progress that's been happening over the last four years is already enough to make extrapolations and say, holy cow, it seems like we could get to AGI in 2027 or so, or sooner. You don't need any secrets to be able to make that inference. And then for the second thing about like, is this a big deal? Is this really dangerous? Again, you don't need any specific secret.

So yeah, he said just even looking at public, publicly available information, not even looking at things that are really, you know, private, you can kind of see that 2027 is going to be the year when AGI potentially gets here to.

Make extrapolations and say, holy cow, it seems like we could get to AGI, you know, in 2027 or so, in the labs in secret.

That is pretty, pretty incredible. I mean, it's one thing, you know, seeing it on a blog post or so or hearing three or four times, you know, but for someone that's working at OpenAI, I think it's pretty telling on the kinds of technologies that they're working on and how people expect them to evolve based on the kind of, you know, advancements that we're going to make in the near future. Now this interview was remarkable. It was honestly kind of interesting to see someone, you know, get on camera quite like we had Leopold Aschenbrenner other day talking about all the things going on at OpenAI, as well as the future of artificial general intelligence. Now I think there was also this part of the interview that wasn't by Daniel Katajlo, but the New York Times do some very interesting reporting on Sam Altman's empire. Now Sam Altman is the CEO of OpenAI, and that does make him a pretty powerful person. But what if I were to tell you that this might actually extend into a further slash different area to where Sam Altman could potentially be the most powerful man in the world. Now that might sound like an exaggeration, but listen to what they state first. Cuz what they state first is true and it's insightful and I'M not saying that being the most powerful man in the world is a crazy thing or a bad thing, but I'm just saying that I think it brings up a very, very interesting question with as to what are Sam Altman's motives for the moves that he is making in the AI industry.

What this is about for Sam Altman, right, Because this is a person who is already rich, but if he has a piece of all the biggest and most important companies in AI, then if you fast forward to the time period that Daniel's worried about when AGI is arriving, one person is going to have a staggering amount of power in that universe. Right? He will be the CEO of OpenAI and he will be a major shareholder of many of the most important companies in the space. Why do I bring this up in that kind of world? You want to know who is this person and what sort of governance structures are around him to reign him in? Because right now it does seem like we're on trajectory for Sam Altman to be one of the most powerful people in the world.

Totally. So, yeah, there we have it, guys. Sam Altman could be one of the most powerful people in the world. And I don't think that that is an overstatement considering the fact that you're at a company that could potentially be at the helm of AGI and then of course could have major partnerships with all of the other companies that depend on your technology, which is going to be integral for majority of the software companies. I mean, would you be the. The most powerful man in the world? I mean, I don't think there's going to be something as powerful. Of course you've got politics, but with artificial super intelligence that eventually will shift the power dynamics because there's no way to outsmart it. I don't see how he wouldn't be. So let me know what your thoughts on this are. I definitely think that this interview is worth a full watch. I'll leave a link to it in the description, but it's pretty insightful. It's pretty insightful that first of all, Microsoft decided to release a model anyways. Second of all, the culture of OpenAI was basically, you know, screw you safety people. And then of course, his prediction there being AGI by 2027, that is truly, truly remarkable. So let me know what you guys think about that and I'll see you guys on the next one.