Elon Musk and FTC’s Potential Intervention against OpenAI’s ChatGPT Dominance

0:00Recently, over a thousand prominent technologists and AI experts, including Elon Musk, have called on ChatGPay and OpenAI to put a halt on any new updates to chat. gpt for at least six months so this is in the news a lot it's getting a lot of tension on the podcast today we're going to dive into why they're calling for this pause what the impact of this is and if this is a good idea does this stifle innovation does this lead us more time to you know prepare and make regulations and also a very interesting kind of event in this whole process is that the FTC may be getting involved after a complaint was formally filed against OpenAI in this whole case.

0:45So we're going to jump on, we're going to dive into it with the podcast. But first off, talking specifically about, you know, what this letter is saying. So essentially, the letter is warning people that AI systems with, quote, human competitive intelligence can pose profound risks to society and humanity. so you know obviously we know that ai systems can do a lot of stuff people accuse them of misinformation or um you know making crazy predictions or different things but you know we have seen that these things are getting better at predicting things they're not perfect obviously we'll see improvements so um it's going to be interesting to see how that plays out um that being said you know one of the biggest things that they say in this letter is that they say that in recent months we've seen ai at labs locked in an all-out control race to develop and deploy even more powerful digital minds that no one not even their creators can understand predict or reliably control we call on ai labs to immediately pause for at least six months the training of ai systems more powerful than gpt4 and this pause should be public and verifiable and include all key actors if such a pause cannot be enacted quickly governments should step in and institute a memorandum and so what is interesting with this whole thing is someone recently yesterday filed a complaint to the FTC complaining and saying that pretty much they wanted chat GPT to be halted so pretty much forced to be halted instead of just complying with this open letter that Elon Musk and everyone else has signed so it's kind of interesting because um essentially their justification for filing this complaint against open ai and having the fdc um shut them down until they fix it is that the model is biased deceptive and threatens both privacy and public safety um so it's someone caidp that is making this complaint um and likewise it supposedly fails to meet the commission's guidelines, calling for AI to be transparent, fair, and easy to explain.

2:50That's kind of interesting. So the center wants the FTC to investigate OpenAI and suspend future releases of large language models until they meet the agency's guidelines. And the researchers want OpenAI to require independent reviews of GPT products and services before they launch. They also hope that the FTC will create an incident reporting system and formal standards for AI generators. So this is all really interesting, right? we have kind of this double storm of people calling on them to pause. And my personal opinion on calling for a six month pause on development of these AI systems, I think it's frankly a little bit ridiculous.

3:26The number one reason why is because like, let's say we go pause this thing here in America. Do you think China or Russia or any other major superpower that could be working on these technologies is going to pause? like any of our technological adversaries, you could call them, or people that we compete with in the free market. No, those people are not going to pause. So I think pausing them just gives the United States a disadvantage if you live in the United States. I'm sure there's a lot of people listening to this that do not live in the United States. And so if you don't align with the values of the United States, then maybe you have a different opinion.

4:01But for me who lives in America and for anyone that has similar values of freedom of expression, not having censorship, etc., etc., which, yeah, is obviously very important to the internet. I don't actually think this would be the play, pausing it in America and letting countries like China, who are very censorship-heavy, continue to develop and catch up. So I think it's actually best to have a country with freedom of expression, freedom of speech, to be the one that really spearheads this AI movement. That being said, is there things OpenAI can do better and ChatGPT falls short of? A hundred percent.

4:37Does that call for a six-month pause? I don't think that's necessarily warranted. I believe that'd be a bit unprecedented. So right now what it kind of comes down to is like what regulation needs to be put in place or should be put in place. And actually, I don't disagree with a bunch of things that are being called for by the Center for AI and Digital Policy. That's a CAIDP who's filing a complaint with the FTC. I think that they have a bunch of good points. And I think that a lot of critics of ChatGPT and Google Bard and like similar models like that, they've warned of what they call problematic outputs.

5:13If you read like there's an article on Engadget when they talk about problematic outputs, they say inaccurate statements, hate speech and bias. And while, you know, those things, none of those are good. I don't really think that even those are necessarily the most dangerous or things like that we should. I don't know. Personally, I'm not super worried about that. Everyone has some sort of biases. Are they good? No. should we work on uh being unbiased of course but um i i don't don't think if ai has a bias that means that intrinsically we need to shut it down um because that's kind of what they're calling for here is for the ftc to halt it until it fixes everything so i don't i think that's a bit ridiculous that being said um i do think that there is some i do think they have some good suggestions you know having independent reviews having uh the ability to look at what biases OpenAI has for them to be able to explain them.

6:02I think transparency is going to be really important. They've obviously went with humans and manually tuned a lot of things this AI says, sometimes training it to lie about things, sometimes training it to skirt around the truth or inconvenient facts, as I'm sure you've seen on Twitter. In any case, this obviously is not going to be the most, I don't think, effective solution. Inevitably, if OpenAI doesn't do this, because the FTC forces them to. There's going to be other AI models that come out that people will choose to use that are more transparent. And in one final point on this whole thing about shutting down OpenAI, ChatGPT, and other language models for six months, there was someone that recently signed a, one of the people that signed that thousand person letter, which included Elon Musk about pausing it, was a man named Gary Marcus.

6:56So he's a New York university professor, um, and Mark Emeritus who signed the letter. And he said in a blog post that he disagrees with the other people who are worried about the near term prospect of intelligent machines. So smart that they can self-improve themselves beyond humanity's control. Cause that's kind of like the, that's like the big concern with AI, right? Is like, if the machine is smart enough to be able to improve itself and we don't even understand how it's improving itself, but it improves itself beyond humanity's control. Right. So it's decided and then, you know, it's of course it goes down to like it becomes so smart.

7:30It can do anything. It teaches itself to be a hacker. It decides that it has to protect the world. It decides the only way to protect the world is to destroy humanity. So it hacks into the nuclear databases, fires nuclear missiles off over the whole world and blows all the people up so that the Earth is saved and not destroyed by humans. That's kind of like the, you know, extreme fear, ultimate fear of like AI. That's like the extreme fear where it goes to. so this New York professor is not worried about that what he's more worried about is what he calls mediocre AI so something that he's worried something that's widely deployed by criminals or terrorists to trick people or spread fake news, whatever currently I think that that also is not a big concern or perhaps it's a concern, right?

8:17maybe that is a concern, maybe those are dangerous things but I don't think that shutting down OpenAI or ChatGP is going to help because there's already a ton of open source AI models. People can go train their own now. Stanford cloned ChatGPT for$600 using Facebook's new Lama model, training model. So honestly, at the end of the day, I think if anyone can train an AI model for$600, there's no way shutting down ChatGPT is gonna do anything. People can get their hands on this technology. It already exists. There's open source versions. there's um you know code bases online that people can download and build this stuff so if we're really worried about the hackers of the world using this to exploit people they're definitely going to do it but they use all technology to do that we're obviously not going to shut down the internet or mobile devices or um you know voice over ip because so many call centers use it for spam calls right like we can't just shut down a whole technology segment you just have to figure out how to regulate it how to make it safe um and ultimately make it you know something put safeguards in place to kind of shut down those things.

9:23There's always those kind of concerns with technological advancements. I don't think that that could ever justify shutting down those technological advancements because bad actors, whether that's states with censorship, hackers, whoever, are not going to abide by whatever rules the FTC or anyone else decides to impose. And so inevitably, I think having good actors continue to develop the technology in good faith with you know regulations I think is ultimately the way to go so it'll be interesting to see if this has any impact on the industry if Chai Chimptee or OpenAI do decide to slow down rollouts and if this is going to slow down the advancement of AI so we're going to have to follow the industry to keep up with that and find out what happens next.