A special type of small talk, commonly seen in office spaces around a water cooler. There, employees often share all kinds of company gossip, myths, legends, scientific misconceptions, mysterious anecdotes, or outright lies. Anything goes. In my very cool water, I discuss strange and often scientific ideas that I, my friends, or other acquaintances who have experienced in the office have left us speechless.
So, here's a water cooler idea for today's post:
AI is a bubble, like the dot-com bubble of the turn of the millennium. It has no real value and will eventually collapse and disappear.
Phew! 😅 There's a lot to remove here. Unlike the ideas I usually discuss about other cooling water, in this case, there is no simple, straightforward answer. AI is clearly an impressive technology, definitely, the achievement of our time, and a few years ago, we could only see such things within the realm of sci-fi.
Having said that, we cannot deny that there is something the worst is on the ground in it – at least in its current state. For example, MIRA Murati's (Opelai's former CTO CTO), currently Machines Lab, is currently worth $ 12 billion, basically, without the product. On top of this, Nvidia hit a stellar $5 trillion valuation a few weeks ago (roughly equivalent to Germany's GDP, roughly).
Undeniable AI models, work for expensive entertainment, with no real return so far. Companies like anthropic and aclaphai pay billions to use their fancy models (operating chatgpt alone costs $700,000 a day), but at least now, they get pennies back. So what happened? Is there a hidden piece of technology that will suddenly be revealed and give this whole machine a go? Are all these tech millennials just fools? Do they have spots? Have we all been tricked into believing in an impossible, expensive future? Or are we headed for an AI winter, and that will be it?
💹 What about the dot-com crash?
The most popular argument of people calling AI A BUBBLE is the dot-com crash of the '90s. But in fact, there were many other bubbles long before the dot-com bubble, which arose from some technological breakthrough. For example, we can think of the seaway mania of the 1840s or the radio of the 1920s.
Anyway! What people seem to forget when comparing the current state of AI with the Dot-com bubble is that most of the dotcom companies that failed back then did not have a proven business model. In particular, in the late 90s, there were many start-ups working with financing and the promise of the future called the Internet, but they did not have a real business model of how to actually make money. Online ads weren't a thing yet, and many of those companies didn't really have a solid business plan.
Source: Wikimedia Commons licensed as public domain,
In contrast, most of today's AI companies have indeed Business strategies are structured. And unlike the '40s, people – both individually and in organizations – are very familiar with the concept of to pay things online. Unlike the 90s and 2000s, the idea of free things on the Internet – like, for example, the value of the flood, or the free value in general – has gone. Now we know very well that if you want something important to get out of any screen, you need to pay; Otherwise, you will only get ads, noise, and bad information.
Even with the most popular dot-coms (think ets.com or webvan), the premise was not true. Pets.com brought pet supplies to retailers, something that exists today and is a valid business venture. Similarly, Webvan made home delivery of items within 30 windows, a market that blossomed 20 years later during the Covil-19 lockdown, with services like weber eat or Amazon Prime. Therefore, it is not the business idea or the technology itself that leads to the 'bubble', but rather its execution and how the market is ripe for such an idea. So, my take on technology-driven bubbles would be that the technology itself tends to have value, but it can happen very early, and the world is not ripe for all these opportunities to be used.
◇ Are all those tech billionaires just fools?
So, if the market is so raw for AI, why do all those companies keep investing heavily in it? Don't you know better? What do they expect to happen without the AI market crash?
The short answer is that they are all just waiting for computer power to get cheaper.
In the past, fast and cheap computers solved many tech puzzles that were thought to be insurmountable. When I look back at the first human thought about AI, people wondered if they could build a system to beat humans at chess. It's funny how chess is considered to require the highest level of human intelligence, so naturally, it would serve as a good measure for an intelligent machine. A good machine in Chess could not be a smart machine. However, the systems that beat people in chess were eventually created, not by finding a fancy algorithm, but by using the same algorithms as improved algorithms. And this is the kind of thing that everyone expects to happen with AI.
Mainly, AI in its current state is very expensive. It doesn't make financial sense because it costs a lot of money to run such models, and it remains to be seen who would be willing to pay that's a lot. For example, with OpenAI just to part with its chatgpt plan once, it will need to charge about $50 per month, which is more than its current price. Therefore, creating fast and cheap chips can allow them to provide the entire development process and use such low-cost models, allowing those companies to finally break away. In other words, hardware progress is the lever that can make AI sustainable.
I can't say more law right now, cliché as it may be. More's Law is the observation that the number of transistors in an integrated circuit tends to double approximately every two years, leading to exponential improvements in power, efficiency, and cost effectiveness over time. Although the ancient Moore's law is now over (progress has plateaued in the last decade), the basic mindset has created (the expectation that it will continue to be much cheaper) than the miracle of today's technology giants.
Replacing the CPUS with a GPU allowed parallel processing and removed the physical limitations that existed before. With the rise of large GPU clusters, computing power is expected to continue to grow for years to come. And that's why everyone is so weak about this.
🤔 So, where is the bubble?
A few weeks ago, Michael Burry placed a $1 billion bet on NVIDIA and Palantir by buying options on both stocks. Michael Burry was the one who quietly shorted the housing market back in 2007, and in fact, predicted the entire housing bubble of 2008 and the subsequent financial crisis. So, naturally, when he called out this alleged AI Bubble and the market, everyone seized their moment. Later in November, understandably close his investment company. A bold way to announce that we are on the market at the top!
But again, the thing is that AI is fundamentally different than the housing bubble. There's a lot we can do with real estate, but AI's look at it isn't entirely fair. It's a question of when and not, because AI will completely change the world eventually. And unlike houses, its value is not disputed; AI is a complex technology, a marvel of engineering, having a lot of nature tied to the work of great minds that worked to bring it together. On the flipside, houses have been built the same way for centuries, with little technological improvement. It's easy to dismiss AI as a speculative bubble, simply because we're not equipped to understand it.
I think it's worth comparing AI with other technological booms that are related to laying some kind of infrastructure, such as electricity, telephone, or the Internet, even railways. Oh, but Rail Mania was a virtual bubble, wasn't it? Well, what was it? The technology itself was not a bubble; Laying the rail infrastructure was unthinkable and had a real, absurd cost. The thing is that other means besides the year eventually appeared, such as cars, airplanes, or pipelines, criticizing the railway so that it could not fulfill its potential. In other words, the demand for travel is ultimately distributed among several different markets without rail. Rebellion arose from the assumption that railroads would be the only forms of transportation that ever existed. Therefore, a rational ai bubble will need other, magical technologies to contribute part of its power.
Anyway! Like the Internet, the main value of AI comes from whatever follows and sits on top of it. Placing an electrical grid allowed with a group of electrical equipment sitting on top of it was built later; Creating the Internet is allowed by a number of applications such as Google or Facebook that were created later. Naturally, whoever sets the infrastructure of what will pay the lot of money, because it is the payment that is paid to whoever is the next. However, this does not mean that technology itself is not important, and does not have the potential to lead to significant growth.
That leads us to an important caveat: For AI to make financial sense, a number of efficient and meaningful things sit on top of it need to be created. Investment in infrastructure itself makes no sense if it is not ultimately maintained as infrastructure. So, this is a real bet, a real question that one needs to answer. The technology is there, the investment is there, the infrastructure is there; Now it remains to see that enough things have actually been built on top of it.
In my mind
I know this post may come across as milky. Maybe in a few weeks or months, I'll be reading it and wondering what on earth I was talking about. However, AI is truly impressive: a technology that has the potential to change the world forever, like electricity or the Internet, and set the stage for a new technological era.
Did you like this post? Let's be friends! Join me at:
📰Put it down 💌 The medium 💼LinkedIn ☕Buy me a coffee!
What about pialgorithms?
Want to bring the power of the rag to your organization?
Pialgorithms you can do it for yourself 👉 Book a demo today
