Resources
Join the discussion: Ask Jordan questions on OpenAI
Related Episodes: Ep 318: GPT-4o Mini: What you need to know and what no one’s talking about
Ep 149: Sam Altman leaving and the future of OpenAI – 7 things you need to know
Upcoming Episodes: Check out the upcoming Everyday AI Livestream lineup
Connect with Jordan Wilson: LinkedIn Profile
Try Our Free AI Prompting Course: Register for our free Prime, Prompt, and Polish AI Course!
Unexpected Change at OpenAI
Recent weeks have shake the foundations of OpenAI, one of the leading companies in generative AI, as prominent leaders including Peter Dang, John Shulman, and Andrei Karpathy have left the firm. The firm's president, Greg Brockman, has also taken an extended leave of absence, giving rise to questions about the company's stability and future.
OpenAI's Supportive Infrastructure
OpenAI's role as NVIDIA's prominent customer, utilizing the latter's GPU chips for technological development, and its deep relation with Microsoft, which reportedly holds 49% equity stake in the AI firm, has been beneficial thus far. However, the recent instability could draw concern from both these firms.
Pursuit of AGI – Safety Concerns and Implications
OpenAI's quest for Artificial General Intelligence (AGI) and recent trajectory indicates an accelerating pace towards it. However, the exodus of key figures from the organization citing safety concerns and alignment issues could be a red flag. The issues revolve around the safety culture and processes being overlooked in favour of product development. Now, with more AI agents than humans in the horizon, the safety discussion has escalated significantly.
Impact on Major Business Partners – Apple, Microsoft, and NVIDIA
The ripple effect of changes in OpenAI could potentially impact companies like Apple, Microsoft, and NVIDIA. Apple's new "Apple Intelligence" relies heavily on OpenAI's GPT technology, which raises concerns about possible delays. Additionally, these changes could possibly affect the US economy and cause further instability in the world economy.
Speculating the Future of AI
The impacts of OpenAI's ongoing drama could have repercussions on the nature of human work and the dynamics of everyday AI usage. There is an expected release of a new model post the US election in November 2024, which could significantly alter the landscape of AI in society.
Staying Ahead of the AI Curve
In these uncertain times, staying informed about the rapidly evolving world of AI is key for decision-makers and businesses. Understanding the ongoing turmoil at OpenAI, the potential implications of leadership changes, safety concerns for AGI, and potential future developments can influence business and strategic choices.
As the situation at OpenAI unfolds, one thing is clear - the sphere of AI is bound to witness monumental shifts. Companies heavily invested or relying on OpenAI's technology and society at large need to be prepared for the potential impact. Keeping an eye on the developments and maintaining nimble strategies will be the key to weather the impending storm.
Topics Covered in This Episode
1. Major Changes at OpenAI
2. Legal Trouble for OpenAI
3. OpenAI's Technology and Impact
4. Future of OpenAI
Podcast Transcript
Jordan Wilson [00:00:21]:
There's more drama right now at OpenAI than a daytime soap opera. Breaking news less than, I don't know, 12 hours ago now have shown that a cofounder of OpenAI has left the company, and the president, a very prominent name at OpenAI, has taken an extended leave of absence. So we're gonna talk about what it all means and also lay out a timeline, because this is probably a lot more, than just some startup drama. It's going to actually potentially impact many of us depending on how this shakes out. Alright. What's going on y'all? My name is Jordan Wilson, and this is everyday AI. This is for you. It's your daily livestream podcast and free daily newsletter, helping us all learn generative AI together and how we can all leverage it to grow our companies and to grow our careers.
Jordan Wilson [00:01:17]:
So we do this every single weekday. It's unscripted, unedited, the realest thing in artificial intelligence, bringing you what you need to know. So today, I'm excited to talk about this with you all, and, you know, I'd love to hear from our livestream audience what you think is is what's what's actually going on here with OpenAI and some of this recent news. So before we get started, as a reminder, if you're joining us on the podcast, please check your show notes and go to your everydayai.com. Sign up for the free daily newsletter. Every single day, we recap our conversation in the newsletter as well as make sure to check out our thanks a million giveaway. We'll be giving away a year subscription to chat gbt or your favorite large language model. So make sure to check out the newsletter for that.
Jordan Wilson [00:02:03]:
Alright. Before we get into all of this drama at OpenAI, and we did we did some legwork on today's episode y'all. But let's first go over the AI news. So the, robotics company Figure, literally minutes ago, has just announced their newest humanoid AI powered robot figure 2. It is touted as the world's most advanced AI hardware for robotics. So the figure o two represents a complete redesign aimed at significantly enhancing performance and capabilities. So some of the, new features in figure 0 2, 6 integrated cameras for improved visual re, processing, battery life boost, onboard visual language model. They're calling that VLM for advanced image and video understanding, triple the CPU and GPU power, 4th generation hands.
Jordan Wilson [00:02:58]:
Yeah. Can, like, make a sandwich. It's wild. Designed for more precise and versatile manipulation, integrated wiring for streamlined and robust build and exoskeleton structure and advanced speech to speech reasoning capabilities, allowing for human to robot conversation. So, obviously, the speech to speech feature is powered, by onboard microphones and speakers connected to a custom AI model in collaboration with OpenAI. Alright. Speaking of OpenAI, yeah, it's a lot of OpenAI today. Elon Musk has revived his lawsuit against OpenAI and Sam Altman over alleged fraud.
Jordan Wilson [00:03:34]:
So Elon Musk has reignited his legal battle against OpenAI after, withdrawing his previous lawsuit. So here's what, in his newest lawsuit, what he's alleging. So Musk was obviously a, an original, on the cofounding team at OpenAI back in 2015. So he's reopened the lawsuits, look, because he said they deceived him that OpenAI and CEO Sam Altman deceived him into contributing $44,000,000 by promising OpenAI would remain a nonprofit and open source its technology. So, yeah, like I said, this is my second lawsuit against OpenAI after withdrawing, withdrawing the first one in June for unknown reasons. I think the unknown reason was the lawsuit was without merit. You know, I'm I'm no legal scholar or lawyer, but I read it, and I'm like, there's nothing here. So, make sure to check out the the newsletter today for more on that, on on that second lawsuit.
Jordan Wilson [00:04:30]:
We'll see if this one actually has teeth or not. And then last but not least in AI news, NVIDIA is facing some renowned scrutiny over alleged copyright violations in its AI training. So NVIDIA, one
Jordan Wilson [00:04:42]:
of the most, valuable companies
Jordan Wilson [00:04:44]:
in the world, you know, just under 2 and a half $1,000,000,000,000 market cap, is under fire for allegedly scraping copyrighted content from YouTube, Netflix, and other sources to train its AI models according to a report from 404 Media. This practice is part of NVIDIA's efforts to develop products like its Omniverse 3 d world generator, self driving car systems, and digital human products. The company claims its methods comply with copyright laws arguing that IP laws protect specific expressions, but not facts, ideas, data, or information according to the reports. YouTube obviously disagrees with spokesman Jack mail Malin reiterating that CEO Neil Mohan's stance, on using AI or sorry. Using YouTube content for AI training without permission is a clear violation of YouTube's, terms. So similar accusations have been made against other companies such as OpenAI and Runway for training their models on YouTube videos without authorization. So, yeah, this is, not a new practice, but, NVIDIA kind of right here in
Jordan Wilson [00:05:46]:
the crosshairs. Alright. There's gonna be
Jordan Wilson [00:05:49]:
a lot more on these stories, so make sure to go to your everydayai.com for more on that. Alright. So livestream audience, thank you for joining us. I'd love to hear your thoughts on this OpenAI story. Yeah. A lot of drama here. So thanks, for everyone, for joining, Cecilia and Tara, Michael, Liz, Woosie, Jake, Christopher, everyone. So let's let's just get straight into it.
Jordan Wilson [00:06:12]:
I know that was a lot of news, a lot of news to go over, but some pretty big things. So what's actually happening at OpenAI? This is, you know, late breaking news, you know, not even, 12 hours old now at this point. So let's talk about what's actually happening. Alright. So first, this was, a report from The Information, a a leading, kind of tech journalism company. So here's what their initial report said. It said that 3 key players at OpenAI, are no longer with the company. So they did say that, they were gone.
Jordan Wilson [00:06:48]:
They updated their reporting because they might have gotten a thing or two wrong. But here is who has either left the company or at least who is no longer at the company as of today. Alright. So Peter Dang, who is the, who was the vice president of consumer product, has left OpenAI. No details on his next moves. John Shulman, big name there. So he is, cofounder actually of OpenAI. There is 11 cofounders, but, and I think there's only 3 left.
Jordan Wilson [00:07:18]:
So John Shulman, cofounder and lead of post training team. So, yeah, he leads the team that fine tunes OpenAI's models that 100 of millions of people, use. So he's left and, is going to go to Anthropic, rival Anthropic. So pretty big news there. And then last but not least, so original, originally, it was, reported that Greg Brockman was leaving the company. So, the president of the company, probably the most prominent figure aside from CEO Sam Altman and Greg Brockman. So, originally, it reported that Brockman was leading, leaving the company. However, it has come out since that he is taking an extended leave of absence to, quote, relax and recharge after 9 years.
Jordan Wilson [00:08:07]:
So, yeah, that's a lot of of high level, people leaving. Right? So at least yes. Greg Greg Brockman will presumably be coming back, but you have 2 cofounders, that were there yesterday that are not there today, and also a VP of consumer product and Peter Dang. So a lot of institutional knowledge just left, the door. So, Greg Brockman, again, after some initial reporting, you you know, because this is almost right? This is almost as big as the original kind of news back in November of 2023 when, CEO Sam Altman was unceremoniously fired and then rehired, 4 days later. But I would say this is almost as big as big of news when you have, 2 cofounders and a VP leaving at the same time or at least it is reported that they're leaving at the same time. But then, obviously, Greg Brockman did post on Twitter and said, I'm taking a sabbatical through the end of the year. First time to relax since cofounding OpenAI 9 years ago.
Jordan Wilson [00:09:13]:
The mission is far from complete. We still have a safe AGI to build. So they're talking, about the future, and seemingly signaling that he will be back at the company, potentially early in
Jordan Wilson [00:09:27]:
2025. So that's a lot. That's a lot. Yes.
Jordan Wilson [00:09:32]:
What Penelope said? Oh, interesting. Yeah. We have our our AI drama here, at OpenAI. So let's go ahead and unpack this, just a little bit. So we don't have answers, and, you you know, I'm sure that this is going to be an evolving story, but I want to take you all, the everyday AI audience, through a little bit of a journey here because I think that there's a lot more going on, than meets the eye, and there's a lot more going on than what is right now being reported. So I do believe that there's going to be, some follow-up news regarding this trio in the coming days weeks, and I don't think it'll necessarily be good news coming out of that for OpenAI.
Jordan Wilson [00:10:22]:
Yeah. Should we hey. Shoot. Should we make this
Jordan Wilson [00:10:24]:
a hot like, a super hot take, on a hot take Tuesday? I have thoughts. With these things, I'd like to try to keep it, kind of kind of, you know, buttoned up and more factual, but I got takes. Let me know if if y'all wanna, hear some of the hot takes for our livestream audience.
Jordan Wilson [00:11:30]:
But first, let's go over why does this matter. It's like, alright, Jordan. You told me a bunch of people are leaving a a company that maybe I don't use
Jordan Wilson [00:11:57]:
a lot. Well, this really matters. Alright.
Jordan Wilson [00:12:01]:
Let's go over. Well, there's a good chance that you are using OpenAI's GPT technology whether you know it or not. Right? So, obviously, if you're using chat GPT, you know that you're using OpenAI's technology. But, hey, Microsoft Copilot, right, which is a a system that 100 of millions of people across the globe are using. It is built on GPT 4. It is built on OpenAI's technology. So, again, between ChatGPT, between Microsoft Copilot, and then the 1,000 of other GPT enabled products. Right? That's the other thing.
Jordan Wilson [00:12:47]:
So many in we've been talking about it in the in in the recent weeks here on the, Everyday AI Show. I mean, you now have all these banks, all these large wealth management institutions, enterprise companies building their own AI systems on top of the GPT technology. So that's why it matters to you because there's a good chance, whether you know it or not, you are using and benefiting from OpenAI's technology multiple times a day. Right? There's a good chance maybe a good chunk of your work, unbeknownst to you, is happening behind the scenes because of or due to the GPT technology. It is everywhere. Right? So many large companies are building on top of GPT. So, obviously, OpenAI was kind of the first big name to the game when it comes to large language models when they released, the chat, chat gpt in November of 2022. They had obviously released their GPT technology, years before that, and a lot of earlier products and services were building on top of it.
Jordan Wilson [00:13:55]:
But you can make the argument that OpenAI technically started this generative AI wave, that businesses and individuals and companies are are all riding now, almost 2 full years later. So this is extremely important, to to to know that you're probably already benefiting from GPT. So when a trio of highly visible users or, sorry, a a trio of highly, visible people at OpenAI all leave the company or it's reported that they leave the company on the same day. Yes. President Greg Brockman will, presumably be back in, early 2025. There's no guarantee on that. We're gonna get to that, here in a
Jordan Wilson [00:14:40]:
little bit, but it's noteworthy. Right?
Jordan Wilson [00:14:44]:
I'm not saying that businesses need to rethink their business strategy. Right? All huge companies, always have a lot of this uncertainty unfolding. Right? The generative AI landscape right now is extremely competitive. So, yeah, like we talked about as an example, Peter Dang, leaving for rival and traffic. There's a lot of that. Right? There's a lot of people at, you know, big tech company a. Right? So, oh, Meta leaving for Google DeepMind. Someone on the Google Gemini team, leaving for, you know, Amazon.
Jordan Wilson [00:15:15]:
Someone at Amazon leaving for OpenAI. So that's normal, but I do not think this is normal when you have 3 high ranking officials reportedly leaving at the exact same time. That's not good. It's not good. Right?
Jordan Wilson [00:15:29]:
Could it be a coincidence? Sure. I don't think this is a coincidence. Alright. So first, this is why
Jordan Wilson [00:15:36]:
it matters to you, because, there's a chance that this is going to impact your work. Alright. I mean, we also have to talk talk about the timing the timing of everything. Right? Obviously, there was this brand new lawsuit from Elon Musk. I don't think that has anything to do with the timing. But, I mean, even 8 months ago. Right? In November 2023, Sam Altman was fired and then rehired, when essentially there was an internal mutiny of sorts. Right? The OpenAI board fired CEO and founder, Sam Altman.
Jordan Wilson [00:16:08]:
Essentially, I think it was, like, 98 or 99% of employees publicly came out and said, yo. We're gonna leave the company, if you don't reinstate Sam Altman, within about 4 to 5 days with some help from Microsoft CEO, Satya Nadella. The company did reinstate Sam Altman as well as, reinstate new members or sorry, you know, put in new members of their operating board as well as take some other members out. So that's not too far off. Right? 8 months. These things take a long time to recover from, and I'm sure there were some underlying issues that maybe led to that instance, in November 2023 that are still unresolved. Alright. So what does it all actually mean? Alright.
Jordan Wilson [00:16:55]:
You get it. Okay. So, yeah, there's a good chance our company is using some GPT technology. There's some turmoil, at the company. Well, it's, I would say, it is severe instability right now. And this is at one of the most important companies in the world. Okay? So we also have to talk about some other big names at OpenAI. Right? Because I'll say this.
Jordan Wilson [00:17:21]:
You know, take away Elon Musk's involvement. I would say that there is 4, kind of big names, at OpenAI over the last couple of years that, you know, you could make the argument that so much of their success was because of 4 or due in large part to 4 very, highly prolific people on their team. So CEO, Sam Altman, still there. What right? But was fired and rehired. Number 2, president Greg Brockman, who is now on an extended leave, right, at least for seemingly, 5 to 6 months. So will he come back? He says he is, so we'll see. Then Andrei Carpathi. Right? And Ilya Sutskever.
Jordan Wilson [00:18:08]:
Alright? And Andrei Karpathy and, Ilya Sutskever
Jordan Wilson [00:18:14]:
are gone. Right? So at least you could say over the last 5 years, those are probably the
Jordan Wilson [00:18:22]:
whether you wanna say the 4 most important or the 4, most visible people at OpenAI. As of this morning, only one of them is currently working at OpenAI and CEO Sam Allman. Yeah. I'm sure there's a lot of things going on behind the scenes. I'm not saying that there's no one else that's important, but I will say as a as an outsider that observes things very closely over the last, you know, 4 to 5 years, those have been kind of the 4 highly visible people, that you could make the argument were maybe 4 of the most important people at OpenAI. Obviously, that's going to change now since only one of them is working. You know, others are going to have to rise up through the ranks, but that is concerning. That is concerning.
Jordan Wilson [00:19:10]:
That creates instability at one of the most important companies in the world. Yes. I'm saying that. OpenAI is one of the most important companies in the world because they are in the driver's seat of the generative AI and large language model race. I don't care what you say about, anthropic, about Google Gemini. No. OpenAI is still leading, is still leading this race. Right? For models that are publicly available right now for consumers.
Jordan Wilson [00:19:46]:
Right? So you go on the front end of any of these, you know, large language models. GPT 4 o technology is still the best according to benchmarks. So we've seen Anthropic's clawed 3.5 SONIT get very close. Google just released a new version of their, Gemini Pro 1.5, but it's only available for developers and in their AI studio. So at least for front end models that everyday people, right, you and me can go use. OpenAI is still the leader. And I think that their next model, whenever they do release it, will likely wipe everyone else off and make them all irrelevant for a a a short time. So that's why it's important.
Jordan Wilson [00:20:29]:
But you also have to think of the huge potential impact that moves like this could have on the US economy. So we do not fully understand the reasoning behind all of this. I have some, guesses and assumptions, and I'd love to hear, I'd love to hear from, from our livestream audience. What do you think is actually going on here? I'm gonna get to what I think is going on here at the end. But we have to look at the potential impacts on the US
Jordan Wilson [00:20:59]:
economy. Alright? So not with throw throw out
Jordan Wilson [00:21:05]:
the last 72 hours. Right? So if you're paying attention to what's going on right now in the US economy, and even the world economies, it's a little bit of chaos. Right? There's a lot of, political instability, a lot of, you know, international, relations going awry. Right? There's a lot of, economic instability right now throughout the entire world. So, you know, take away the last, you know, 72 or, you know, 4 to 5 days in time when, you know, the US stock market has kind of taken a slight nosedive. Alright. So take away that. In OpenAI, ready for this y'all?
Jordan Wilson [00:21:45]:
This is this is when
Jordan Wilson [00:21:46]:
we talk about bringing you the news that matters, and it's like, oh, no. OpenAI AI is just a little startup. What does it what does it matter? No one's talking about this. How about the 3 largest companies in the world?
Jordan Wilson [00:22:40]:
Are overly, I wouldn't say overly reliant, but they are heavily reliant on open AI.
Jordan Wilson [00:23:06]:
Apple, the number one company in
Jordan Wilson [00:23:09]:
the world when it comes to market capitalization. So that's what we're talking about. Essentially, the 3 companies in the world that are worth the most. Number 1, Apple. Number 2, Microsoft. Number 3, NVIDIA. And for the last, you know, 3 months, those 3 companies have all kind of jostled back and forth, and they've all held the number one position multiple times. So they've all kinda gone, between, you know, silver, bronze, and gold there for, for the top 3 in the world.
Jordan Wilson [00:23:36]:
All 3 of these companies are heavily reliant, at least right now, on OpenAI. And look at all of the instability just I mean, obviously, over the past 12 hours, but even over the last couple of months, when you've had not only 3 big name people leave at once. You know, one of them is coming back, but we also said, hey. Out of the kind of, quote, unquote, Mount Rushmore of Open AI, Sam Altman is the only one showing up today and presumably for the next 5 months until and if Greg Brockman comes back. That creates a lot of instability for the biggest companies in the world as well. If I'm Apple seeing this news, I'm not happy. If I'm Microsoft seeing this news, I'm not happy. If I'm NVIDIA seeing this news, I'm not happy.
Jordan Wilson [00:24:28]:
Y'all, those are the 3 largest companies in the world. Alright. Let's talk about some of their, you know, reliance or about how of how some of their success, is definitely due to this, path that OpenAI has been blazing. They've been the one chopping down the weeds and paving the roads and making way for everyone else to follow. Apple. Yeah. You've seen that. Alright.
Jordan Wilson [00:24:58]:
So Apple is, releasing its new Apple Intelligence. You know, at first, they said it was gonna be here in September or in about a month or so. Now it's being pushed back to October. It might have been pushed back even more with this recent OpenAI news. I don't know, but it's plausible. Right? So, essentially, here's how it works. This Apple Intelligence, which come on, Apple. Just call it AI.
Jordan Wilson [00:25:22]:
Right? They're trying to rebrand artificial intelligence and call it Apple Intelligence. Can you really do that when the bulk of your, quote, unquote, Apple Intelligence is just the GPT technology from OpenAI? Yes. So people aren't talking about this, but Apple actually has 2 different models, that are executing its Apple Intelligence. So first, they have their own internal JAX, model that runs locally, on upcoming iPhones. So the iPhone 15 and above or which is essentially the, you know, iPhone 15 or the one that's coming out, the iPhone 16. Alright. So they have their own small language model that runs a lot of things locally. But, according to early reports, the majority of all Apple intelligence is essentially just being passed off to the GPT technology to OpenAI, where essentially you get a prompt and it says, hey.
Jordan Wilson [00:26:14]:
Do you want to go ahead and launch this in GPT?
Jordan Wilson [00:26:18]:
And then you click yes. Right? So so much of the companies of of the number one companies in the world or the the number one company in Apple, their big marquee feature, for the most part. It's just sending people to OpenAI.
Jordan Wilson [00:26:35]:
A lot of instability there.
Jordan Wilson [00:26:37]:
Mhmm.
Jordan Wilson [00:26:38]:
Right? Yeah. Cecilia says Apple can rebrand anything and people will buy it. Yeah. That's true. That's true sometimes.
Jordan Wilson [00:26:49]:
Hey. What Tanya says, one
Jordan Wilson [00:26:50]:
of my favorite sayings here at everyday AI. You gotta follow the money. So we just follow the money for Apple. Let's follow the money for Microsoft. Alright. Number 2, biggest company in the world. I mean, you cannot deny how important OpenAI is to Microsoft. Alright? Microsoft's value has skyrocketed over the last 2 years for a couple of reasons.
Jordan Wilson [00:27:16]:
But I'll make the argument it's largely just because of its development and early commitment to artificial intelligence. That's why, you know, I think, over the last decade or so, at times, Microsoft has struggled to keep up with Apple. Microsoft was a year and a half ahead of Apple in terms well, I mean, I guess Apple hasn't even given its users artificial intelligence, but Microsoft was early to the game. Right? They were early to the game with Copilot compared to Apple. And I think that's one of the reasons Wall Street and the what and the rest of the world, has responded favorably and how now Microsoft is, essentially always the number 1 or 2 or sometimes 3. But, essentially, they're the number 1 or number 2 company in the world, which they haven't always historically been over the last 10 years, and I think that's because of their commitment to artificial intelligence. And that is made possible by the GPT technology. So Microsoft also reportedly has a 49% equity stake, in OpenAI as its largest investor reportedly investing $13,000,000,000 into the start up company a few years ago.
Jordan Wilson [00:28:28]:
Alright? So you can't deny the ties there are very close. Yes. Microsoft has started to distance itself from OpenAI. I think there's a lot of reasons, both both topically and, under the surface on why they're doing that. Number 1 is it's probably not too good, for a company with a $3,000,000,000,000 market cap to be too reliant on technically a startup in OpenAI with 600 some employees, probably not good. So, Microsoft has started its own AI division, Microsoft AI, by, you know, hiring an acquihire essentially, and they are developing their own internal models as well because they do not want to be too reliant on OpenAI. And then earlier, earlier this week or sorry, earlier last week, which we reported, for the first time, Microsoft is listing OpenAI as a competitor. Right? So as OpenAI, starts to come out with new products such as search GPT, it is kind of a competitor to Microsoft's Bing, search engine even though the search GPT uses the Bing technology.
Jordan Wilson [00:29:41]:
I know it's confusing. Right? But Microsoft is, both essentially calling OpenAI one of its biggest partners, but also one of its biggest competitors. So, again, if Microsoft seeing this news that a trio of highly visible leaders are no longer there, I'm not feeling good about this. Presumably, Microsoft knew about this before the in, before the report from the information, but regardless, it's not good. Then last but not least, NVIDIA. So I'd say out of these 3, NVIDIA has the least to lose, when it comes to OpenAI. But, again, OpenAI has been one of NVIDIA's, I won't say larger customers, but one of its more prominent customers. Right? So, the GPT technology was built and trained on NVIDIA's GPU chips.
Jordan Wilson [00:30:32]:
Right? So, again, I think out of the 3, NVIDIA is probably least reliant on, OpenAI. Right? And the reason why I'm going over this is, well, what happens if something crazy crazier is happening at OpenAI? And what does it mean for the US economy? Well, like we just said right there, number 1 and number 2. Right? Let's just say for whatever reason over in the coming quarters, OpenAI loses its grip on the large language model race. Right? Which could be possible. Right? If you lose like I said, if you lose your Mount Rushmore, if you lose 3 of the 4, more prominent people on your team, if then you keep losing, you you know, key key leaders, people who are cofounders, VP levels, if you keep shedding people at the rate that OpenAI is shedding them right now, I don't think it's gonna turn out well. I don't think it's gonna turn out well. Alright. Let's keep this going.
Jordan Wilson [00:31:29]:
Just a couple more points to make here. So let's let's talk about the the the elephant in the room here. What is actually going on? Right? What's actually going on? Why was Saint Maultman fired back in November? I mean, there was a lot of reports. A former, OpenAI board member at the time came on and, you know, went through all the dirty laundry. That's not what this is about. So, you know, like I said, anytime at a fast growing startup like OpenAI, there's always going to be internal conflicts. There's always gonna be, you know, wrestling for power internally. But I think something else is going on here, and it's pretty obvious.
Jordan Wilson [00:32:05]:
I think there's safety concerns. I think there's safety concerns. Alright? So OpenAI has been very open about one thing. They've been open in their pursuit of AGI or artificial general intelligence. Alright. And if you are new here and if you're not really sure what that means, it's kinda like Terminator. Right? It's kinda like that. That might technically be superintelligence, but, AGI, artificial general intelligence, is essentially when the AI systems that we all use become smarter than any human at almost all tasks.
Jordan Wilson [00:32:42]:
Right? And if you ask the smartest people in the world, there's actually a graph. I'll try to find it and share it, and share it
Jordan Wilson [00:32:50]:
in today's newsletter. But,
Jordan Wilson [00:32:53]:
essentially, there's a graph, that shows over time. Right? They've they've, always asked the top researchers in artificial intelligence because AI is not new. It's been around since the fifties. Right? Businesses have been using artificial intelligence, deep learning, machine learning for many decades. They've always surveyed the smartest people in artificial intelligence about AGI. Because when AGI happens, when artificial general intelligence happen happens, things get weird. Things get weird. Right? When all of a sudden, you can talk and ask an AI system anything, and it's going to get it consistently right every single time.
Jordan Wilson [00:33:34]:
I mean, I I don't think I'm I'm here to get philosophical, but it really changes humanity. Right? It changes more than work. I think right now, artificial intelligence, generative AI, large language models, are drastically and radically changing what it means to to work and what it means for for businesses to use technology. But I think when you start having the artificial general intelligence question and then after that comes, you know, superintelligence, That's when I think it starts to say, okay. This really changes humanity. This changes what it I know that sounds weird. It changes what it means. It it changes the role of humans.
Jordan Wilson [00:34:12]:
Right? Anyways, this survey over over the years has said, how far off is artificial general intelligence? And, you know, 10 or 20 years ago, it seemed like most people were like, oh, it's it's 30 years away. And then, you know, 7 years ago, people were like, oh, it's it's, you know, maybe 15 years away. Right? But it keeps getting. Every couple years when they ask the smartest people in the world about artificial general intelligence, it is seemingly getting closer
Jordan Wilson [00:34:40]:
and closer. Right? And now some of
Jordan Wilson [00:34:45]:
the smartest people in the world say, oh, we're maybe a year or 2 away. Right? I was lucky enough to be in a closed door session with NVIDIA CEO Jensen Huang, at the, NVIDIA GTC conference. And he was asked about this multiple times, and he essentially said less than 5 years. Right?
Jordan Wilson [00:35:07]:
That changes things. And there
Jordan Wilson [00:35:08]:
are so many safety concerns. Because at that point, it also when you talk about AGI, right, so when you talk about artificial general intelligence, you have to think about AI agents. So essentially think there there could, in theory, be more AGI enabled agents than humans. Right? There could be 1,000,000,000 of AI agents that have more reasoning capabilities, more strategic thinking abilities, and more knowledge than the smartest humans in the world. And there could be 1,000,000,000 of them 1,000,000,000 of them working autonomously, completing tasks, building companies, even themselves building new AI systems on autopilot. Right? That's what you have to think about when you think about this transition from art from, artificial intelligence, AI, to artificial general intelligence, AGI, to ultimately superintelligence. Right? So is that what is that what's going on here? Right? Reports have said that OpenAI has been done training their GPT 4 model for almost 2 years. Right? And there's been a lot of very other visible people at OpenAI leaving and citing some safety concerns.
Jordan Wilson [00:36:29]:
So let's talk about that one here. Alright. So Jan Leakey. Alright. So, we cover this on the show, when this happened. So we have to I mean, we have to talk about this one. So he came out and said so he is the former head of alignment at OpenAI in the super alignment lead as well. Alright? So he left for Anthropic back in May, and he was one of the first, prominent former OpenAI employees who said what everyone else is thinking.
Jordan Wilson [00:37:07]:
Alright. So among other things, I'm reading his tweets here on the screen if you're on the the the livestream here. He said building smarter than human machines is an inherently dangerous endeavor. OpenAI is shouldering an enormous responsibility on behalf of all of humanity. But over the past years, safety culture and processes have taken a back seat to shiny products. We are long overdue in getting incredibly serious about the implications of AGI. Alright. End quote.
Jordan Wilson [00:37:45]:
Alright. So Jan Leaky, again, was one of the more prominent people, in the whole world when it comes to alignment, when it comes to model alignment and model safety. And he came out and said it. He said OpenAI is not taking the safety of AGI seriously enough. Right? And then he essentially just said, oh, no. We're going to I'm going to Anthropic, and we're going to devote the resources actually needed to make AGI safe. Alright. So what else is going on here? Well, it seems like there's also a little bit of an anthropic exodus or people leaving OpenAI to go to anthropic.
Jordan Wilson [00:38:34]:
Alright. So like we talked about, one of the 3, original names here in this original, report. So, again, if you're just maybe tuning in late or if you need a reminder, the the trio of leaders who are no longer at OpenAI, Peter Dang, the vice president of consumer product has left OpenAI. No updates on his plans. John Shulman, a cofounder of OpenAI, who was the lead of the post training team that fine tune models, He's gone and has left for Anthropic. And then the president of OpenAI, Greg Brockman, originally was it was reported that he was leaving the company, but he's not. He said that he is just taking an extended leave. So John Schulman leaving to Anthropic.
Jordan Wilson [00:39:23]:
And then you also have Jan, Jan Leakey and others that have left OpenAI for Anthropic and even Anthropic itself. Right? We don't talk about this enough maybe, but Anthropic was started by 2 former OpenAI, people who worked in safety. Actually, a a brother sister duo. Hopefully, I get the names right here. But, it's Dario Amodi and Daniella Amodi. So, Dario was the vice president of research at OpenAI, and Daniella was the vice president of safety and policy. And they left OpenAI, I believe, in 2021 to start Anthropic.
Jordan Wilson [00:40:09]:
So there's there's a a trend here. There's a trend here. Right? And you also,
Jordan Wilson [00:40:16]:
I mean, you also have to look at Ilya Sutskever. He didn't leave for Anthropic, but, you know, Ilya essentially you know, he was missing for a while. Right? Essentially, it was kind of underreported. Oh, he's taking a leave of absence, and no one heard from him for many, many months. And then it was just announced in May, that he was officially leaving OpenAI and starting Safe Super Intelligence Inc or SSI. And what's curious to me is when you go to now their website, which is just a blank white page with some black text, it starts with the phrase, superintelligence is within reach. Okay? I don't think Ilya would have put that on, SSI's website if he didn't understand that, hey. OpenAI has to be already close to having artificial general intelligence.
Jordan Wilson [00:41:13]:
So there's been
Jordan Wilson [00:41:14]:
a lot of reporting, and this
Jordan Wilson [00:41:16]:
is where we tiptoe the line between reporting and rumors. Right? But there's been reporting and rumors that OpenAI's next model, whether that's called GPT 5, I believe it might be called GPT next. I'm I'm you you know, we're not sure. But there's been plenty of reporting that it's been done and it's been ready to go for a while and that OpenAI and I I believe this and OpenAI is waiting until after the US election, to release their new model because presumably, it will be maybe a little too much for the world to comprehend all at once. And the opportunities for misinformation and disinformation, I mean, especially when you talk about things like SORA. Right? If SoRA, you know, OpenAI's extremely impressive AI video tool will be incorporated into the next version of GPT 5 or GPT NEXT, There's a lot of misinformation and disinformation implications that come along, with this next model from OpenAI. Presumably, it will be off the charts in terms of benchmarking in everyone else. OpenAI, for the most part, I mean, they've updated their GPT 4 technology in their GPT 4 0, but it's it's impressive.
Jordan Wilson [00:42:29]:
Yes. But it's still working off the backbone of technology that was reportedly it's it's been done training for 2 years. So you have to assume that OpenAI either has a model that is bordering on artificial general intelligence or that is maybe too powerful and too capable, for the rest of the world to handle. I mean, it sounds weird saying that out loud, but, I mean, when you see some of the highest ranking members of OpenAI leave and they go to they go to Anthropic. Right? And I don't think this is a money thing either. Right? People are getting paid NBA salaries at OpenAI or Anthropic. It doesn't matter which one. So I don't think it's a money thing.
Jordan Wilson [00:43:16]:
I think it is ultimately a safety thing. I think there's been high ranking officials at OpenAI that, you you know, we said it in Jan Leaky's own words that OpenAI has maybe focused a little bit too much on the next shiny product and not enough about the safety of artificial general intelligence. So there's a lot going on under the covers here. So we have to think of what's next for OpenAI. Well, I'm not sure. All I know is the the the coming weeks and the coming months are going to be incredibly fun to watch. Right? I'm glad I'm not, an executive right now at, you know, those 3 big companies. Right? Apple and Microsoft and NVIDIA because, from their from their point of view, I can see how this is a lot of uncertainty and instability in one of the more important companies in the world.
Jordan Wilson [00:44:14]:
But here's the other thing.
Jordan Wilson [00:44:17]:
As a start up with still some of the brightest minds
Jordan Wilson [00:44:20]:
in the world, OpenAI is, capable, I think, of so much more than all of us can even comprehend. Right? We haven't even fully talked about even the new, vision and voice capabilities that are coming, in the new update, which should be in theory coming out any day soon. Right? OpenAI is a little late on that, but essentially in a desktop mode on Mac where you can click a button, open chat gbt can see whatever you're working on. You can talk to it like a real human. I mean, if that's not a huge step toward AGI, I don't know what is. Right? When you're on your smartphone, you can say, like, hey. What's this? And it can see in real time with low latency, and it's smarter than any any human. Right? When that comes to your phone, when that comes to your
Jordan Wilson [00:45:07]:
desktop, I mean, I don't know. Is that AGI?
Jordan Wilson [00:45:12]:
I don't know. All I know is it's going to be extremely exciting because I am in the camp of, saying that, hey. Probably in the in the months after the US election in November 2024, we're probably gonna see a new model that I don't know if the business world will know how to handle it. And it might really change what it means to be a human worker in today's society. I truly believe that. Alright. That was a lot today, y'all, unfolding the drama, giving you all what you need to know. So now you are officially if you're listening to this, you are officially probably the smartest person in your company when it comes to AI.
Jordan Wilson [00:45:54]:
You know what's going on, what has happened, what is happening, and what could happen. Alright. So I hope that you find a lot of value in these type of episodes where we kind of break through the noise and give you, some of the deeper implications of what's going on on the surface. So if you do enjoy this at everyday AI, please consider if you're listening on the podcast, check out the show notes. Go subscribe to our daily newsletter at your everydayai.com. Subscribe, wherever you get your podcasts on Spotify, on Apple. Even if you're normally a livestream person, maybe you can't come every day. Make sure to check out the podcast and share this.
Jordan Wilson [00:46:30]:
Go ahead and tag someone at your company that needs to hear this. Hey. You are the smartest person at your company when it comes to AI if you're tuning in every day. So thank you for tuning in. Please also go check out the thanks a million giveaway. Go tell your friends. And more importantly, please join us tomorrow and every day for more everyday AI. Thanks, y'all.