Loading summary
Transcript1 lines
- [00:01]
Jayden Schaefer
Welcome to the AI Chat podcast. I'm your host, Jayden Schaefer. Today on the podcast we're talking about some news out of Anthropic, one of the number one companies when it comes to AI and their CEO has put out a very interesting manifesto blog post, as it were, about essentially the fact that he is a techno optimist and believes in the future of AI. Now, this isn't shocking considering he's the CEO of a huge tech company, but this is a company that has spent a ton of time and money on, on essentially making sure that AI is safe. And it's interesting to hear his perspective on the fact that he is not an AI doomer. He doesn't believe this is going to be the end of the world. We're seeing similar things out of a lot of big people in the industry. Yan Lecan, recently, back in May, had a big blog post or Twitter thread, essentially where he was saying that, you know, he believes we're at the very beginning of this AI thing and we do not need to regulate it, etc. Etc. So it seems very interesting, but I want to talk about some of the things Dario was saying because I think there's some really interesting insights that he has and some very interesting predictions. A lot of this is reminding me of when Elon Musk kind of announced what he was going to be doing with, with SpaceX back in the day, and a lot of people in the industry came out against him and essentially said that there's no way he's going to be able to build reusable rockets or get the price down to where he's at. It's interesting because a lot of the journalists, including ones over on TechCrunch are, are essentially skeptical of a lot of the stuff that Dario saying, some of his timelines, some of his predictions. But I'm gonna say I believe, and of course we're all gonna have to wait and find out, but I'm gonna say I'm more inclined to believe these predictions and these timelines are accurate on Dario's side. And it's probably gonna be something that we look at in retrospect, similar to all these clips we see of like, Elon, people that were, you know, skeptical of SpaceX and what they were going to accomplish back in 2013, but now everyone's showing them because SpaceX just laundered their, you know, their super big rocket back in, back in the tower came out of space. The things like the skies of a skyscraper and it was able to land itself and all of a sudden everyone's like, oh, look at all these people that don't believe in Elon back then, and now he did it. I feel like we're going to have some of the same situations in 5, 10 years with some of these predictions. So what are the predictions, uh, and what essentially did Dario say? Broadly speaking, he's talking about a world where all of the AI risks are essentially mitigated. This is similar to Yan's vision on this. And even Sam Altman came up with a similar kind of manifesto right before he was raising money. He really focuses on the fact that he believes that there's going to be unrealized prosperity, social uplift, abundance, and they're going to be able to minimize all of the downsides that AI potentially could have. So he doesn't specifically name any other companies, but he does say that there's a lot of big steps that have been taken that he believes are moving them in this direction. So Amade believes that they're going to essentially reach what he's calling, quote, unquote, powerful AI, and that they will reach it by 2026. This is a bold prediction. This isn't something he's putting out five, 10 years in the future. He's literally saying, where at the end of 2024, that within a year. So, you know, you wrap up this year, maybe get into 2026, a year and a half, we're going to be there. So what does he really define when he's talking about powerful AI? What does that actually mean? He says that it is something that is, quote, smarter than a Nobel Prize winner. And he specifically outlines biology and engineering mathematical theorems, and he says writing extremely good novels. So really something that is quite capable at talking, but also reasoning is where he's kind of putting this really good AI and something like as good as a Nobel Prize winner. So this is quite impressive. He says that this new AI that he sees coming out by then is going to be able to control any software or hardware imaginable. So that means industrial machinery, and it's essentially going to be able to do all of the jobs that humans are doing today. But he says that it will be able to do it even better. So this is really interesting, especially when you start thinking about software, right? Like if this thing can really control your computer, for example, click around and do things, it can essentially automate all of white collar jobs. And then if it's able to integrate with all of these humanoid robots out of Tesla and other places, it'll essentially Be able to automate, automate blue collar jobs as well. In a very literal sense. Like replacing a human. Yeah. So this is very interesting. Here's a direct quote from him. He said, quote, and referring to the AI, it can engage in any action, communication or remote operations, including taking actions on the Internet, taking or giving directions to humans, ordering materials, directing payments, watching videos, making videos and so on. It does not have a physical embodiment other than on a computer screen, but it can control existing physical tools, robots or laboratory equipment through a computer. In theory, it could even design robots or equipment for itself to use. Now this is a fascinating concept we talked about in the past, right? The concept of like an AI model training itself to be better and better theoretically. Right. Or teaching itself new things. So it's like brain is getting bigger. But imagine if it was then able to control machines that could build robots and it could all of a sudden build itself an arm or a leg or like just different, I mean those are examples. But like it could build itself different tools or machines that could help it accomplish whatever task it was interested in accomplishing. That would be fascinating and for some terrifying. So a lot of things would have to happen for us to be able to reach that. Even the best AI today, a lot of people are arguing these things, can't actually think. They're just predicting what words are coming next. So there's still some problems that are left to be solved. So over at TechCrunch, they believe that he is optimistic and they say very optimistic. They kind of call into question the article, there's an article in TechCrunch about this, calling into question the ability that he's going to be able to have to do this, which I, I, I believe it's going to happen. I say, here's my thing. Anytime you have these big leaders paint a broad and very ambitious vision, the first thing people are going to do is attack it and say, there's no way. This vision is way too big, it's way too crazy. But you want to know what happens with people with crazy visions? They attract the top talent, the top engineers, the top scientists to come work for them. Because why would I, if I was at the top of my game, go and work for a company with a very small vision that I didn't feel like I have a big impact. When there's another company has a big impact and, or a big vision and they say they're going to have this big impact. If I'm the brightest mind, I'm going to go work there. And they're going to be able to get the most investors, the most money, they're going to be able to compensate people the best. Like all the money and talent attracts into kind of the biggest vision. You see this with Elon Musk and SpaceX, you know, saying we're going to go to Mars. Okay, well no one's even went to the moon in like forever. So like what makes you think we can go to another planet and reuse rockets and make you know, spaceship 100 times cheaper than NASA? Like that's a very big vision. But it attracted the best talent and they were able to do it because of that. And so yeah, you see very similar things. I mean I even think of like Rivian Tesla. A lot of these players had very big visions and they've been able to accomplish incredible things and attract incredible talent because of their vision. So he believes the AI in the next seven to 12 years could help treatment all infectious disease. Now this is a very, some very interesting, very interesting things that he's you know, speculating might be possible. He says in seven to 12 years it could be able to eliminate most cancers, cure genetic disorders and stop Alzheimer's in its earlier phases. Anyways, he thinks in the next five to 10 years conditions like PTSD, depression and schizophrenia and addiction will be cured with drugs that AI was able to essentially create or he says genetically prevented via embryo screening. Which obviously is a fairly controversial way to go about doing this, right? Just screening embryos and then aborting all the babies that you think might have this issue or that issue. And for a million reasons it's a very controversial topic there. He thinks that all of this will also quote, tune cognitive function and emotional states to get our brains to have or to behave a bit better and have a more fulfilling day to day experience. To me that I don't like the sound of that. I don't want an AI to tune my cognitive function and my emotional state. Seems like manipulate is what tuning it would be. But anyways that's just me. In any case, he believes that due to all of this, what the impact that's going to have on diseases and mental states and all this, that we're going to double the human lifespan so we'll be able to live to 150 years. This is very fascinating if you know things like Brian Johnson, what they're working on, trying to make people live longer, whatever, it's very interesting with a lot of that science that's coming out too. Here's a quote from him. He said My basic prediction is that AI enabled biology and medicines will allow us to compress the progress that human biologists would would have achieved over the next 50 to 100 years into 5 to 10 years. I'll refer to this as the compressed 21st century. The idea that after powerful AI is developed, we will in a few years make all the progress in biology and medicine that we would have made in the whole 21st century. Very ambitious, but an incredible vision. This is exciting. This is something that if I was wanting to work on something, I want to get behind this vision. This is amazing. So, um, TechCrunch says, quote, these seem like stretches to considering that AI hasn't radically transformed medicine yet and may not for some, for quite some time or ever. So TechCrunch skeptical of all of this. But in any case, Amadei, he had some other ideas. He, he went so far as to say I could solve world hunger. He thinks that it can help address climate change. He thinks that it can transform the economies. This is very interesting. He said, be able to transform the economies in, you know, sub Saharan Africa and their gdp, their per capita GDP to be about the same as the per capita GDP of China right now, which is China is $12,700 in 2022 and Sub Saharan Africa is $1,700. So I mean almost a 10x probably seven and a half if I'm my mental math is, is right 7.58x and he says that it could do that to Africa in the next five to ten years. Very interesting. Bold predictions, bold vision. Personally, I think this is interesting. One thing that he did say is that he thinks this is going to, quote, require, quote, a huge effort in global health, philanthropy and political advocacy. And he thinks this is going to be in the world's best economic interest. So it's going to happen. He thinks this is going to. That there would be a dramatic change in human behavior and so that all of this is going to kind of be embedded in there. He thinks that AI are going to be paid much lower than a minimum wage, while he says their employers are going to get tens of millions of dollars. Or maybe that's just my own prediction, who knows? In any case, something that TechCrunch likes to always bring up whenever you talk about employees and employers and AI is the fact that these AI companies are obviously raising a lot of money and doing well and they hire people for below minimum wage in third party countries or in third world countries and they always are saying that they're paid below minimum wage. Which paying someone in $3 in the Philippines is below minimum wage in America for sure. But that could be a great salary in the Philippines, depending on what, you know, region of the Philippines you're from or any other country. So in any case, TechCrunch likes to complain about that. That is a fact of the global economy. People make different amounts of money, different places. So I don't know, I don't think that's the big thing to bring up, but maybe that's just me. In any case, talking about some of the danger, talking about some of the that side of things. He makes some interesting predictions. He thinks that there's going to be a coalition of democracies which are essentially going to secure AI's supply chain and they're going to block anyone that's an adversary. So you can imagine like the EU and the United States and Canada and some others blocking China and Russia or other political adversaries from getting the supply chain, getting like the Nvidia chips and the semiconductors and all of that stuff. So this is very interesting and we see this sort of, sort of from the United States are now blocking China from getting the most advanced chips for AI. He also says that if AI is in the right hands, it could be used to quote, undermine repressive governments. So big shout out to the CIA if you want to topple some new governments. AI sounds like this might be your tool of choice, can go undermine repressive governments. I mean, in all honesty, that statement would be fantastic if you're actually undermining repressive governments. Somehow I feel like it's just going to be done wrong and we're going to go topple governments that we don't like the, you know, the, the leaders of, regardless of whatever. Yeah, whatever. You know, the CIA, classic CIA moves. Okay, I think that's pretty non bipartisan between Republican and Democratic administrations over the last 100 years. I think that's, I think it's pretty, a nonpartisan issue. In any case. Anyone at the CIA listening, big shout out. Okay? So this he says, is quote, a truly mature and successful implementation of AI has the potential to reduce bias and be fair to everyone. TechCrunch says they think that AI is going to exacerbate biases. He thinks that it could fix it. It's kind of interesting because TechCrunch says it's going to exacerbate it. You need to do more to fix biases in AI. These AI companies are doing stuff. Some people say they're doing too much, some People are saying they're not doing enough. Elon Musk says that it's dangerous to add our own biases from the people training these. Sam Altman and Dario think that you need to inject your own beliefs to, you know, drowned out beliefs that are bad, make these things have guardrails and behave good. So, anyways, there's different, different takes on that. But all of this to say, I think this is going to be absolutely a phenomenal time to, to see essentially what's going to happen in the future. Sam Altman did a very similar thing, a very similar post when he was raising, rounding, finishing up his $6 billion raise. So I recommend you go check that out if you're interested. But it seems like we're coming at a place where a lot of these, a lot of people in the space are very optimistic about what's going on in AI and they can see a lot of upside. Of course there's going to be downside, I think, but I think there's going to be a ton of upside. I personally am very optimistic and very excited about this. I'll keep you up to date on everything happening in AI so if you enjoyed the episode, make sure to like and subscribe, and I hope that you have an incredible rest of your day.