 From theCUBE Studios in Palo Alto in Boston, bringing you data-driven insights from theCUBE and ETR. This is Breaking Analysis with Dave Vellante. Many people question if the current AI boom will end in the same way that the dot-com bubble burst. You know, that's understandable, is there a lot of similarities, especially with the exuberance this week in the stock market following Nvidia's earnings print. Well, it's easy to dismiss AI as a completely different era. There are some stark similarities that are worth remembering. Like all waves, there are also some major differences. Now, two of the most evident are the speed of innovation and the quality of today's companies that are leading the AI charge as compared to many of the dot-com firms. But like the internet, AI will be ubiquitous and available for virtually all organizations to leverage, not just traditional tech firms. Hello and welcome to this week's theCUBE Research Insights, powered by ETR. In this Breaking Analysis, we look back at the events that were leading up to the dot-com boom and bust and analyze the similarities and differences between these two transformative eras. The dot-com was characterized by irrational exuberance, that's a statement that was made famous by the then-fed chairman, Alan Greenspan. So let's take a look at the events leading up to the dot-com bubble, what the conditions were that fueled it and what happened in the aftermath. By the mid-1990s, it was clear that the most powerful monopoly in tech, IBM, had relinquished its crown. Wintel, the powerful virtual combination of Microsoft and Intel, were the defining platforms of the pre-dot-com era. Now in this chart, we show nearly a fraction of the relevant milestones and conditions that fueled the very exciting dot-com bubble. And we'll start in 1994 with a company called CMGI going public. They were a nothing shell company and ended up becoming a prominent dot-com investment firm and really the poster child for pump and dump unsustainable businesses. But what really got the dot-com started was the Netscape IPO. Less than a year after its first browser launched, a spate of IPOs ensued, including a few that were showing here like Yahoo, Amazon and eBay. Now, note also the number of IPOs we show that data starting in 1999 in the bottom of the chart there with 473 companies IPO-ing. 446 the following year in a steep reduction after the bubble burst with only 91 in 2001. And it bottomed out in 2003 at 71, way fewer than even 2022, which was a terrible year for IPOs. Well, let's back up to 1996. President Clinton signed the Telecommunications Act and that opened up competition and communications and also the floodgates for investment. Companies like Global Crossing, Worldcom, Enron, which ended up being a disaster, Quest, Lucent, Level 3, they skyrocketed and public telcos poured nearly half a billion, half a trillion dollars into building out the internet between 1996 and 2001. And in today's terms, that approach is a trillion dollars. So like today, you had all this investment in CapEx enthusiasm to the point where CMGI had a valuation of $41 billion. The company CEO was dubbed the Buffett of Dotcom Investing. And it turns out CMGI just owned a bunch of distressed assets and after the bubble burst, it lost 99% of its value. The bubble peaked in March of 2000 and it did so during a time when interest rates were steadily holding for those years between seven and 8% and also at the peak of the Dotcom, the US government had a budget surplus, quite different from today. The bubble really didn't burst until the world realized that much of the venture funding was gonna go to companies that was going to companies with no business model and these distressed assets were using VC money to buy network equipment from Cisco, servers from Sun Microsystems, storage from EMC, Oracle Licenses, Dell PCs and they were advertising like crazy on these VC-backed internet portals like Yahoo and Lycos. It was this big circular reference echo chamber that came crashing down and just when folks thought the pain might subside and markets would return to, let's call it abnormal exuberance, 911 happened and the tech industry went into a very long funk. Silicon Valley at that time, I remember driving up and down 101, it was a ghost town that were empty offices and it wasn't until Google IPO'd that the market slowly began to come back and we were deep into two wars in a financial crisis that lasted from 2007, late in 2007, well into 2009 and that momentum was halted again because of that crisis and it ushered in a long period of super cheap money. So before we go there, I think it's worth looking at the prevailing laws of the era, the dominant laws in technology. So leading up to the Doc Tom, Moore's law was in full swing. Of course it was named after Gordon Moore, we all know the mantra, doubling the number of transistors on a microchip every 24 months. The dot com era, however, was underscored by a paradigm described by Metcalfe's law. Bob Metcalfe was the co-inventor of Ethernet and a colleague of mine actually at IDG and he was the publisher at Info World and at the time and he posited, Metcalfe's law posited that the value of a network increases exponentially, even if the connected users, even as those connected users increase on a linear curve or linear line, think of this as another way when there are two people on the network with phones, there's not much value but when there are billions of people with phones on the network, that value curve bends exponentially. So now we're entering the AI era and Jensen Wong claims that the more you buy, the more you save. So we're calling this Jensen's law and he might be right, his premises that compared to general purpose CPUs, his superchips can run AI workloads, which are gonna be the predominant workloads, 12x faster and they're 20 times more energy efficient than traditional general purpose CPUs. And his argument is a direct shot at Moore's law because at a doubling of performance every two years, it'd take the better part of a decade to achieve these types of results. So whether Moore's law is dead or not, that's a debate for another day but clearly new approaches to silicon design are hitting the market and are needed in this AI era. So let's take a closer look at the period leading up to this current AI boom. This era was preceded by the end of a 10 year bull run which was fueled by zero interest rates policy, ZERP, otherwise known as ZERP. It was an epic time period, of course, powered by the cloud and SAS which AWS kicked off before the financial crisis. In fact, during the financial meltdown of 08 and 09, a lot of CFOs of companies that were previously reluctant to try the cloud started to migrate to the cloud as a way to reduce IT capex and all that heavy lifting. And they realized when they came out of the downturn that it was a better model and then we saw a decade of tremendous growth and that tech boom that was interrupted for a very short time when COVID hit but then it took off as organizations were called had no choice but to migrate to the cloud and then the end of ZERP halted that run and tech spending began decelerating in 2022 as the Fed tightened. But chat GPT was announced late in the year coinciding with a pause to Fed hikes and that modern AI run began and you can see in the NASDAQ started its run here. Last year, Microsoft added a trillion dollars to its valuation. Most of that from AI frothiness and it stands at $3 trillion today. Early on Friday morning of this week, the 23rd of February, Nvidia's market value hit $2 trillion. And of course the catalyst of this era was the announcement of chat GPT and the realization that Gen AI was going to impact every company, every industry and most people on the planet kind of similar to the internet. It feels like chat GPT is a bigger thing than the Netscape browser but at the time that browser was pretty remarkable. Regardless, these were the kickstarter moments for both eras and like Netscape, open AI today is top of mind as Netscape was back then. So this chart from ETR gives you an idea of just how much mind share open AI has today. This graphic is from ETR's ETS. It's the emerging technology survey of around 1500, it's more than 1500 IT decision makers of just privately held companies shows net sentiment on the vertical axis. What that is, it's a measurement of an intent to engage. And then mind share is on the horizontal axis. Like Netscape, open AI has had much greater momentum, has much greater momentum in mind share than its competitors. So much so that we had to put a red box around open AI's position on this graph because it's literally off the charts. And you see that pack of AI companies, they've raised a lot of dough, close to $10 billion. Well, Netscape's competition came from a few no-name startups but also some open source browsers like Firefox. And of course, big competition came from Microsoft and then later Safari came in. Now, many of the gen AI players in this chart, they may not become household names and they might go the way of some of these early browsers. And open source alternatives are emerging similar to the open source browsers. The most prominent, of course, is Met Islama but there are dozens in the mix. I think 15 at last count is probably more. As well, just as Google search disrupted the internet portals like Yahoo, we're seeing chat style engines like Perplexity, they're vying for mind share and adoption. And so some disruptions similar to .com will potentially evolve. Now, the other big similarity that people sometimes forget is the CapEx investment that was made back in the .com. I was talking to John Furrier about this, we were talking about this on the CubePod, he made this point. This chart here is from Charles Fitzgerald who covers this stuff pretty closely, the CapEx wars. And it shows how much investment the hyperscalers are pouring in to CapEx to support cloud and of course now AI. Amazon with its warehouses includes more than just AWS in their CapEx spend, but it's a build out that's quite similar in scale that we saw with the telcos and ISPs during the .com boom. And as Fitzy points out in his last post, NVIDIA's data center revenue was approximately equal to 50% of the hyperscalers CapEx last quarter. And by inference from NVIDIA's statements, cloud companies probably spent nearly $10 billion on GPUs last quarter, which again is fueling this AI boom. So let's close by looking at some of the most obvious similarities and differences between the .com and the AI booms here. Netscape and OpenAI, they were the catalyst innovators. Microsoft killed Netscape. They did that by bundling the browser into the operating system, Internet Explorer. They actually OEM the browser from a company called Spyglass and then basically co-opted it. Interesting. Today it's bundling OpenAI and a close partnership. So you keep your friends close. We'll see how this partnership progresses. OpenAI is a little bit better funded than Spyglass was. The web was funded again by Telco CapEx as hyperscalers of funding AI CapEx today, but there was no internet and there was no cloud at the time. So what does that mean? The difference, the big difference, the obvious differences we set up front is this time the pace is very much accelerated. And then you had the picks in shovel companies that were highly valued. Cisco was the most valuable company on the planet back then and had a trajectory and its stock price very similar to NVIDIAs. Having said that, its valuation metrics were not nearly as attractive, believe it or not, as NVIDIAs, NVIDIAs relatively cheap compared to Cisco back then. The dot-com was more bubble-ish with less quality earnings than the market today. But companies like Cisco, Sun Microsystems, I remember after the dot-com bust, their president who became CEO, Ed Zander, on an earnings call said, does anybody want to buy a server? It was that bad. Like I said, Silicon Valley was a ghost town. So there were these picks and shovel companies like Cisco, Sun, EMC, people were buying a lot of Oracle licenses at the time and running servers basically out of their own data centers. So they were spending a lot of their venture capital money on those picks and shovels and those stocks ran up as well. They were spending a lot on advertising on the portals that were VC funded and then that dried up and poof. But the point is that these picks and shovel companies is kind of an analogy to NVIDIA today and even the hyperscalers, AWS, really an infrastructure services company and Microsoft and Google. Meta as well, room to be getting into the business. But NVIDIA and these hyperscalers have, in our view anyway, much more durable business model. So it's going to be to be determined how durable NVIDIA's model is. I personally think it's got quite a bit of runway left and we'll see. As well, the injection of AI into SaaS is going to accelerate adoption. So you're seeing companies like Snowflake and Databricks and Workday and Salesforce and ServiceNow are a big backstop to a really value generator as AI gets injected into their platforms. Back in the .com, AOL was the granddaddy of the internet. They were a very, very odd company and they had this closed proprietary walled garden kind of like NVIDIA and the hyperscalers are today. But you know, ultimately the open web and open web standards won out. And look, open source software is there today like Lama. So that's kind of similar. But the big difference is that AOL had a very weak value proposition. It was ripe for disruption. It was really on thin ice. You know, maybe if they opened up and took a different strategy, they could have evolved but they didn't. But I feel like NVIDIA and the hyperscalers have these massive moats much stronger than the fragile .com companies that we talked about earlier. But open source still looms, so we'll see. Now the market back then, markets were irrational with lots of investor fear of missing out, FOMO, anything .com just went up to the moon. And we're seeing some of that irrational exuberance again today, but look, seven companies really led the AI boom in 2023. And it's even narrowing now. So it's a much narrower set of companies. But as I said earlier, the supporting cast of SaaS companies are gonna sort of play in this wave but they're much more durable than the frauds of the .com. Again, Snowflake, Adobe, Workday, Salesforce, ServiceNow, Oracle, SAP, and many, many more are gonna power this AI era. The stock market was very bubble-licious, probably even more so back then than it is today. But the companies back then had no earnings. You know, the other differences, IPOs were booming during the .com. And as we showed earlier, it's not the case today. And the leaders, the other big differences the leaders today are highly profitable in today's markets. So they've got much, much stronger balance sheets and much more durability. But look, there's a lot of similarities. The embryonic markets, they have very similar patterns. The one .o gets better, leads to a two .o and a two .5 and a three .o and more and more innovation. And we're seeing the same thing today. The difference, again, this time around is the pace of innovation is gonna be much faster because you've got the ubiquity of the internet. It's there, it's a platform and the cloud. And all that investment that's put in there is leverageable for AI. Like the .com days, we're witnessing a completely new set of user experiences. So that's another similarity. The difference this time around is for the first time we're talking about the replacement of cognitive functions. Okay, so that's a big, big difference. But, again, some similarities to the .com are looking at a virtually unlimited and incalculable TAM. At the same time, post.com, we saw a lot of risks. Social media came in after .com and who knows what our artificial general intelligence brings. Another thing not shown in this chart is the regulatory climate is much different today. It's more constricted because the tech market is much more mature and so that remains another wild card. So the question on everyone's mind is whether the market is overblown. What's going to happen to NVIDIA? Is the market going to blow up at some point? And the answer is, yeah, probably. When that's going to happen as anyone's guest, but unlike the .com bubble, it won't likely be because guys like the cloud companies are running out of money and they can't buy GPUs. Now, they may stop buying maybe because maybe they overbought, maybe they're ordering too much and you might see some fluctuations in supply and demand, but it's not likely going to be because these firms are not good companies and they run out of money. Today's AI firms are much more durable than the .com frauds. Rather, the disruption is more likely going to come from external factors. What if China invades Taiwan and chokes semiconductor supply? There's a lot of wars going on. What about acts of terror? Climate disaster. What about other financial crises that aren't expected? Or some other external event that really shakes the economy. Now, hopefully that event will not decimate humanity and will recover to see the wonderful promises of AI coming to fruition. I don't know. What do you think? Are the similarities to the .com greater than the differences? Or is it kind of full speed ahead and dam the torpedoes? All right, that's it for now. Thanks to Alex Meyerson and Ken Schiffman on production and they also do our podcast, Kristen Martin and Cheryl Knight help get the word out on social media and our newsletters and Rob Hoef as our editor-in-chief over at siliconangle.com. Remember all these episodes, they're available as podcasts wherever you listen, just search Breaking Analysis Podcasts. I publish each week on the cube research.com and siliconangle.com. When I get in touch, you have some thoughts on this or other episodes you can email me at david.volante at siliconangle.com or DM me at dvolante. Pitch me, I love to get pitches. Bring you on the show if you got a good story or comment on our LinkedIn posts. Please do check out etr.ai. They get the best survey data in the business. This is Dave Vellante for the cube research insights powered by ETR. Thanks for watching, everybody. We'll see you next time on Breaking Analysis.