META's Big Plan to Catch Up on A.I.
Showing mention at 0:40 — highlighted below
Transcript
20 segmentsNobody does it better than Regent's Seven Seas cruises. Enjoy all-inclusive unrivaled luxury with unlimited shore excursions, indulgent cuisine, personalized service and more aboard spacious all-sweet ships. Visit rssc.com to experience the unrivaled. On the legacy retirement group.com phone line, he is our tech reporter, Mike Duboskey, ABC News. See, good morning, Michael. By the way, I see that Mark Zuckerberg at Meta, of course, the parent company of Facebook and Instagram, kind of doubled down on their open source AI strategy at Meta. And they're just trying to stay competitive. Break this down for us. Exactly. I think he hit the nail on the head there. This is a new strategy from meta outlined in a 6,500 word essay from Mark Zuckerberg yesterday that basically outlines what they think the future of the AI space is going to look like. It's a title or the title of the post is called The Future is for Everyone. And basically the big theme is that sharing is caring. AI should be accessible to everyone and it should be powerful. That's the only way that it's going to usher in, according to Mark Zuckerberg, extraordinary new things. new businesses, new ideas, new concepts, and advancements in health and quality of life. What does sharing is caring mean? Well, he says that they are going to start offering artificial intelligence basically for free or at extremely low cost to everyone. This is part and parcel of the sort of open source way of thinking about technology. We shouldn't necessarily keep it to ourselves. We should make the source code, the sort of building blocks of our artificial intelligence available to everyone around the world. They can build things on top of it and make their own advancements. A rising tide lifts all boats, in other words. This also means just monetarily, they are launching a bunch of new open source models that anyone can download for free, start building things with. They're not the most powerful models that meta has ever made, but they're kind of in that conversation. They're called Muse Glimmer, is what their titles are. And they are relatively lightweight as well. They can run on laptops and things like that. So that gives them a little bit more accessibility, at least according to meta. This also gets into some of the social cost of AI, which Mark Zuckerberg outlines in this essay. We're talking here about data centers, which have seen huge pushback across the country for putting pressure on local electrical grids and water systems. So META says they're going to start building energy generating infrastructure wherever we invest. That could potentially give some low-cost energy back to local communities. They also say they're investing a billion dollars to support police and fire. fighters in their sort of communities where they start building data centers. All of this, again, kind of is an effort by meta to not just outline what they see as the future, but also to compete better with the best of the best in the AI industry. That is to say, open AI, anthropic, and increasingly Chinese firms like Moonshot and Alibaba, which have unveiled some pretty powerful models recently. There remain some big questions, though, about whether this strategy will actually do that, will actually vaunt meta into the... the same tier as the big guns in the space, especially when you consider some of the existential problems facing, not just the AI industry, the sort of social pushback to data centers, the struggles to monetize any of this technology, but also some of the micro concerns that meta itself is just dealing with yesterday.
day that Mark Zuckerberg published this big essay, a judge denied the company's appeal of a group of lawsuits that claim it and other social media companies intentionally create addictive features that harm people, especially young people. So they're going to be hit with a whole bevy of new lawsuits over the next few months that they're going to have to fight and pay for. So that is going to take some of their capital and some of their attention away from this AI game. You said a key word, struggling to monetize. AI. So why would meta give away AI models for free when their competitors are trying to make money? It's a little bit like you and I are across the street from each other and we've got lemonade stands and you're selling a glass of lemonade for a buck. I'm giving it away. Well, where are people going to go? So it's a good question. So this is kind of an ancient struggle in Silicon Valley, this idea that should technology be open and free for everyone to build upon and to learn from, or should we keep it to ourselves and try to monetize that? If you go to that lemonade example, more people are going to go to your lemonade stand right at the end of the day. Even if you're not making money off it, you have a bigger user base. So that is a potential avenue to monetization. Maybe there's something they can do with advertising. maybe low-cost subscriptions that they start turning on a couple years from now after they've demonstrated to people that, hey, this technology works, this is crucial for your life. That's a potential avenue that they could go down. It's also worth mentioning that this open source, low-cost idea is kind of how the Chinese AI firms have been talking about AI Moonshot and Alibaba both have unveiled these open source or open weight is technically what we should be saying in terms of artificial intelligence technology, these sort of open models. And, you know, they have been pitching that as like a competitive advantage. Hey, look at what you can come do with our technology, that open AI and Anthropic. are not offering so yeah it's sort of remains to be seen exactly how they can monetize that but i think they have an eye on you know garnering the users first and then figuring out how to make money off at second yeah i mean that's a data collection is the is the business model here really i mean it's like hey if i can find out more about you and then i can flip that around and sell you other things i mean that's the That's the model there. I mean, that's how they did that with their social media platforms, right? It doesn't cost anything to sign up for Facebook, but they make their money off of the data that, you know, you give to that platform that they can then sell to advertisers. And maybe they have a similar strategy with the AI. I mean, right now AI is free. I mean, you can go to chat GPT right now. I mean, that's a different. That's a chat bot essentially, but I mean, you don't pay a nickel for that. So, you know, would they, I guess, you know, it's like when you, they started charging for Twitter or X, if you wanted a blue check, you could pay a couple of bucks a month. And then, you know, now it's all of a sudden a not a free platform for you.
Yeah, now, you know, X still does have the free tier. It kind of remains to be seen whether like a lot of people actually signed up for the premium version of that platform. Right. You know, that was a way for them to sort of flip the money switch, basically. You know, they knew they had a lot of users. They knew a lot of people spent a lot of time on the platform and cared about things like blue checks. But I think what they came to find out over the course of. you know, the rollout of this new premium version of X was that, hey, maybe not as many people value this, at least to the degree that they would shell out however much money a month, you know, to actually have us turn a profit on this. And to a degree, that's what we have seen in the AI space as well. A lot of people use artificial intelligence. Apparently, chat GPT uses more than 900 million, has more than 900 million monthly users closing in on a billion. However, very small amounts of people pay for it every month. You can. You can get more advanced AI features. You can go and do more things with if you shell out 20 bucks a month for a subscription. But, you know, the AI sector has struggled to message to the, you know, sort of everyday person that... that extra cost will actually come back to you, right? Like that if you shell out, like it's actually valuable enough for you to justify the cost. That argument hasn't really been made in a convincing way to enough people for this to be a profitable industry.