AI Lens
AI news, hot topics, advancements, and discussions about how AI is reshaping business and society.
Your focused view on the emerging hot topics in the Age of A.I.
AI Lens
Season 1 Episode 15: The Exponential Curve- AI Capabilities Doubling Every 7 Months
Use Left/Right to seek, Home/End to jump to start or end. Hold shift to jump forward or backward.
AI Lens is your Focused view on the emerging hot topics in the Age of AI!! We provide AI news, hot topics, advancements and discussions about how AI is reshaping business and society.
Today we're exploring a captivating topic: the exponential growth of AI capabilities—specifically, how they seem to double every seven months. This isn't just a trend; it's a phenomenon reshaping industries, society, and even our daily lives. With the new Claude Opus 4.6, we have new capabilities with AI. What kind of regulation and guardrails are needed? The competitive players in AI are leading to great advancements in AI.
This is AI Lens, your focus view on the emerging hot topics in the age of AI. We provide AI news, advancements, and discussions about how AI is reshaping business and society.
SPEAKER_01Today we're exploring a captivating topic, the exponential growth of AI capabilities, specifically how they double every seven months currently, and that rate is actually increasing in speed. And this is not just a trend, it's a phenomenon reshaping industries, society, and even our daily lives.
SPEAKER_00In this episode, we'll cover several key areas, including the drivers behind this rapid growth, recent benchmark achievements like the Claude Opus 4.6, the challenges of measurement in today's world of AI, and the implications for safety and government, governance. So let's jump right in. Let's start off, John. What do you think about how we can understand better this exponential growth in the AI capabilities?
SPEAKER_01Well, let's first start by talking about what is exponential growth. And so bear with me for those who obviously know this. In simple terms, exponential growth refers to an increase that occurs at a rate proportional to its current value. So I think thinking of a metaphor would be really good here. So imagine like a snowball rolling down a hill. It gains mass and speed as it rolls further and further, just like AI capabilities are gaining speed as technology and research progress.
SPEAKER_00Well, historically, AI has seen several periods of rapid growth. And it's it's often referred to as the AI winters and the AI summers. I think actually the current phase is one of the hottest summers we've experienced yet. With the advancements in the machine learning and the deep learning, the capabilities of AI systems are really advancing at what I consider breathtaking pace at this point.
SPEAKER_01I think that's a good area to explore deeper, Liz. I, you know, drivers of rapid capability growth. So let's talk about that a little bit. I think the first thing to talk about is scaling laws. I think one of the most significant factors is our scaling laws. So as we increase the size of AI models and the amount of data they process, their performance improves dramatically. So this is akin to say, like how more horsepower enables a car to achieve better speed and efficiency, right?
SPEAKER_00Yeah, I think there are also algorithmic improvements that we need to consider. There are all kinds of algorithm algorithmic innovations. There's new architectures like transformers that have revolutionized how AI processes different types of information and it makes it more efficient and capable.
SPEAKER_01Absolutely. And again, those those algorithms are going to continue to increase in efficiency and speed, especially when the AI itself starts discovering algorithms that us humans can't even think of. I think competition is another driving factor. So you got tech giants like Google, you know, OpenAI now, Anthropic, Microsoft, XAI, you know, all these different competitors. And then you got Chinese ones like the open source, DeepSync, et cetera. So I think that they're starting to push the boundaries of what's possible, and it's driving innovation because each company is racing to develop the next breakthrough that creates like a feedback loop that accelerates growth.
SPEAKER_00I completely agree. You know, like we right now have for benchmark records, look at Claude Opus 4.6. It's really a breakthrough. This model has recently set remarkable benchmarks in various AI tasks. It's showcasing just how far we've come. It's outperformed previous models in language understanding, creative writing, and even problem solving.
SPEAKER_01Yeah, and you're starting to see, you know, it's funny because originally there were these benchmarks, and the benchmarks are getting more complex and they're measuring better and better. And you're starting to see a little bit of fragmentation with these different AI models in that they're starting to specialize somewhat and do better at different things. So, you know, based on, you know, logic and reasoning versus creative writing, for instance. So you're starting to see different benchmarks. And then they seem to change rather rapidly. So each new release seems to catapult some competitor to the top of the benchmarks. But right now, Cloud Opus 4.6, definitely setting records. But these these numbers, these benchmarks are not just numbers. I think they represent real advancements that can enhance, you know, applications in healthcare, education, and business. And it really begs the question: what does this mean for the industry? It sets new standards and expectations for performance.
SPEAKER_00Well, and one of the problems is the measurement challenges and why it really matters. It's not just looking at numbers like you said. Measuring AI performance is actually a lot more complex than that. We first need to figure out what are the defining metrics? What do we want to measure? Are we measuring accuracy, speed, creativity? Each metric can yield different insights. And then we also have the challenges. Issues like overfitting, bias in training data, and the quality of the data sets all complicate this whole measurement process. If we don't have reliable metrics, we can really risk misinterpreting what a model's capabilities are, and we're in a whole new frontier with AI. So we don't have set standards on metrics on how to measure it.
SPEAKER_01Well, and again, the benchmarks are changing because the models are getting more complex. But the problem is, to your point, Liz, misleading metrics can lead to poor decision making. So whether that's in policy, business, or technology development, and this is actually a very important conversation for anyone involved in AI. Probably a whole nother podcast in and of itself.
SPEAKER_00Right. So, you know, a lot of people have compared what's going on with with the COVID era exponential growth of the virus, believe it or not. I don't know if I would necessarily do that, but that's what people are saying is that this is growing so fast, it's kind of like the way some sort of a virus would mutate. I think that if you compare it to a virus, it's kind of a negative connotation, whereas this can bring so much opportunity and it's an advancement and innovation. So I see it a little differently. Trevor Burrus, Jr.
SPEAKER_01But setting that aside for a second, that negative connotation, it's analogous in that, you know, we like we saw with COVID, there was exponential growth, you know, just like we're seeing with AI. And I think, you know, just as the pandemic tough pandemic taught us about rapid adaption and reaction, I think there's good lessons there, right? Because AI growth challenges us to keep pace with technology. We talk about in this podcast often how it's important to start learning about AI and embracing AI because it's changing so fast, and that change is going to come with more significant advancements and quicker advancements. So it's critical that you stay on top of these, you know, different growth, you know, it's different things with AI as it as it evolves. And the lessons learned about preparedness and response are as crucial as we navigate this frontier, as crucial as ever as we navigate this new frontier.
SPEAKER_00Well, and as we were discussing before, with rapid growth, you have to take into account the significant safety implications. As AI systems become more capable, they also become more powerful, which can lead to potential risks such as misinformation, privacy violations, and some people may use it for malicious purposes.
SPEAKER_01I think we're already seeing that play out, if I can jump in here. I mean, you have deep fakes, you know, to biased algorithms affecting things like hiring processes or different how they value different things, right? So these stakes are high, and we must develop AI responsibly, is the bottom line.
SPEAKER_00So and that's why we have governance challenges right now with the rapid growth. The current regulatory landscape is really struggling to keep up the pace with the AI evolution. Policymakers face immense challenges in creating regulations that protect users without stifling the innovation. And this is a very delicate balance, but we really need to find a way to ensure accountability and ethical standards in the whole AI development.
SPEAKER_01You know, along those lines, I was just reading today that the current presidential administration is looking to create, you know, federal federal regulatory environment that will actually preempt states. So this way, states can't come in and over-regulate AI is what it looks like they're trying to accomplish.
SPEAKER_00Aaron Powell Well, I think we do need something federal and then probably something international on this, because I mean AI goes across the whole internet around the world, and to have something that is just per state really won't work. We need something that is consistent throughout.
SPEAKER_01I think the limiting factor there is we're kind of in this prisoner's dilemma. We talked about this before, right? So each country is racing to accelerate their own AI advancements so they can be better positioned competitively in the future. So, you know, who who wants to like limit AI whatsoever, you know, and trust that other countries will as well, when there's not always the same incentives or the same willingness to act in a trustworthy manner, especially when the stakes are so high.
SPEAKER_00I really see that this year it's probably going to be large discussions and various treaties and summits that are going on, because they have to have international cooperation with some regulation. Trevor Burrus, Jr.
SPEAKER_01You know, Eric Schmidt, the fo one of the co-founders of Google, was talking about how it might take a catastrophic event tied to AI to really cause people to rethink hey, let's put some guardrails, some some better guardrails around this. And I think just I read also that ChatGPT led somebody was used like on ChatGPT and ended up creating some sort of a mass shooting in Canada. And so now Canada's looking to hold OpenAI accountable for it. So be kind of something to keep an eye on in terms of accountability going forward.
SPEAKER_00Aaron Powell Well, let's hope that more disasters don't happen that would cause the regulation. I would like the regulation hopefully to be precautionary ahead of time. Sure. Anyway, let's look at you know the kinds of other things that go on that cause the growth of AI to be so fast. And I think what you were saying before, it's a lot of it has to deal with the competitive pressure in the whole AI space. And the major players are fiercely competing to innovate, and this can lead to rapid advancements.
SPEAKER_01Yeah, it's um it's interesting though. It can also lead to some feuding, as you see with Anthropic and OpenAI right now. Those two seem to be going back and forth. You saw the ads on the Super Bowl commercials where Anthropic kind of took some some some might say cheap shots at OpenAI for their proposed advertisements for their free version. And OpenAI has taken some shots back at Anthropic as well in subsequent interviews and whatnot. So but I think the pressure can also lead to corners being cut in safety and ethical considerations. So again, the stakes are so high and the money's so significant and it's so competitive that you know it's important that we also foster a culture of collaboration alongside competition so that we can ensure the advancements benefit society as a whole and we don't take shortcuts and it can end up costing society.
SPEAKER_00I I completely agree. What really needs to happen going forward, well, first we need to establish ethical standards for AI development. It's going to involve collaboration between the big tech companies, governments, and academia for this to happen. We're also going to need to invest in education and research to cultivate a workforce that's capable of navigating this complex landscape. And I know that the education system at first was very much resisting AI, you know, as because it was first seen as a tool to cheat. I think now more and more it's coming on, the schools are coming online with working with AI and teaching the students. Hopefully, by doing so, we can harness some of this and use AI for good rather than have it, you know, be captured and used for malicious malicious purposes.
SPEAKER_01Yeah, Morgan Stanley came out and was I'm not sure exactly who of Morgan Stanley, but they were saying that AI is not gonna lead to necessarily early retirement initially, that basically people are gonna have to retrain, especially white-collar individuals, they're gonna have to retrain and learn AI because that's gonna become the future of work. So, you know, you talked about how we talk about how the future is gonna change quite a bit in this podcast. You know, I think that you're gonna see, you know, we talk about the build-out that's gonna happen over the next 10 years with AI. It's tremendous and remarkable and it's gonna grow exponentially. Can't emphasize that enough. And therefore, you're gonna see some crazy breakthroughs, especially in the next three to five years. We're gonna look back on this time and just be shocked at how far we've advanced during that time. And you're gonna see natural language processing, computer vision, and even emotional intelligence being built into AI systems. You're starting to see it already, in fact.
SPEAKER_00Yeah, well, and the implications for various industries like we've been discussing are going to be profound, and we're gonna be able to revolutionize so many areas that's going to make life just an incredible. This is a very exciting time to be alive, a very exciting time to experience this whole change. I mean, I I've been hearing a lot of people saying, you know, we went through our industrial phase for a long time, and they're saying we're not in it anymore. This is a whole new thing. And it's part of the intelligence revolution and with predicative analysis analytics to transform education with personalized learning experiences to healthcare revolution, to discovering cures to diseases, I so many things are happening that's gonna change this world. It's gonna be incredible.
SPEAKER_01You know, one thing that's interesting too is you sit there and you watch like these like uber successful individuals, people that are on the cutting edge of predictions or market makers, those types of individuals, they all agree that AI is gonna be just completely redefining society at the most elemental level. I mean, it's gonna be profound, it's gonna be broad, it's gonna be deep. And so I think we're past the point of thinking that somehow AI is just a fad or a bubble and all those sorts of things that's used to hear once in a while. And at this point, it's simply gonna be something that is gonna just, again, grow in ways that that we we're gonna look at this time, and it feels like it's grown quickly over the last two years, but it's gonna look like we were moving at a snail's pace. And I think that the pace is accelerating. The new updates are coming out faster and faster, the advances are getting greater and greater with each of these, and it's only gonna continue to grow that way.
SPEAKER_00And so Well, I would say if you haven't checked out Claude Opus 4.6, be sure to do it. I I believe X uh Grok is coming out with something. It was a grok, no. Open AI, I think, is coming out with something new. Was it called Garlic?
SPEAKER_01Garlic, yeah.
SPEAKER_00Yeah, that's supposed to, I think, I think it was it's supposed to come out late this week. So it, you know, it could be any day now. So keep an eye out for it and try them and see what's going on. And and really, it's just such an exciting and very complex time to be alive. And we're gonna continue to navigate this dynamic landscape. It's very crucial to stay informed and to stay engaged. And when the new models come out, try them, see you know what you think. And I know my preference for, you know, and whether I'm going with open AI, anthropic, rock, all of the whichever one, it's really it changes almost weekly. So anyway, until next time, we have had a blast looking at this today. And I know that both of us are really excited for these new advancements. Thanks for listening to AI Lens. Remember, we're your focused view on the emerging hot topics in the age of AI, providing you with news, advancements, and discussions about how AI is reshaping business and society. If this episode about the exponential growth and development of AI made you curious or skeptical, make sure to follow, subscribe, and share this episode with someone. Just remember, exponential growth means we're con it's constantly changing, and we are very likely to have another episode giving you mu many, many, many more updates. Until next time, stay curious, stay informed, and keep your lens focused on the future.