In today’s Pill, we will talk about how this recent AI rush is different than the dot-com bubble or any economic bubble for that matter.
Before the story begins, here is a request from the author of this page – It takes a lot of effort to write genuine high-quality articles with a full-time job and it would be great if you could support us by following us on LinkedIn and subscribing to our website and more importantly if you find it informative, share it with your colleagues and friends!
With that being said, let’s dive right into today’s story.
The Background
Nvidia, the chipmaker reached $3 trillion in market cap this year in June, surpassing Microsoft and Apple briefly to become the most valuable public company. What’s so special about that you ask?
Well, Nvidia is the company that is powering much of the AI revolution with its chips being used in a majority of the data centers where giants like Google, Amazon, Meta, and Apple are training large language models (LLM), which require huge amounts of computing resources and data to train and they underpin generative AI applications like ChatGPT.
However, recently the company lost a staggering 30% of its market, almost a trillion dollars. It wasn’t just Nvidia though, it was in total seven companies - Microsoft, Apple, Tesla, Amazon, Meta, Alphabet (Google’s parent company), and Nvidia – also named Magnificent Seven, who are heavily investing in AI.
This begs the question – Are we in an AI bubble? Something eerily similar happened during the dot-com bubble burst when investors poured hundreds of millions of dollars into a new technology, prices of every internet company skyrocketed but when the profits didn’t turn up and ‘speculators’ pulled money, the bubble popped.
There’s a case for and against it. Let’s start off with some reasons why the AI bubble might be real.
The Ayes
The obvious reason is companies haven’t yet realized their profits and it is not clear if or when they will. The Dot-com bubble or any economic bubble for that matter takes shape when the underlying asset’s value is inflated far beyond its expected value. To put this context, according to Sequoia Capital, the entire AI industry to be sustainable has to generate $600 billion in annual revenue however the industry’s biggest player, OpenAI, clocks just $3 billion a year. This means the investment made today by these giants will take decades to materialize, that is if they ever do. This has created panic among investors or ‘speculators’ and hence they seem to be pulling out. Moreover, for companies to ever recoup their investments, enterprise spending on AI has to go up because currently it is only limited to small trials. They need to convince the big corporations from each industry why spending on AI is worth it and also what would happen if they don’t.
The second possible reason could be the pullout of some prominent names in the industry. Character.AI raised $150 million to build AI chatbots. Adept raised $415 million to build AI agents. Inflection raised a whopping $1.525 billion to build an AI chatbot named Pi. All these companies still exist in some form but their founders have long exited the companies and agreed to work for tech giants (Google for Character, Amazon for Adept, and Microsoft for Inflection).
Why you ask? Most startups realized that even billion-dollar fundraises can’t get even close to what the big boys are ready to spend on AI. Hence the most profitable and logical thing to do according to them is to take the cash when they still can.
Lastly, the rate of innovation seems to be slowing down. When ChatGPT was released, it was groundbreaking. Especially the ChatGPT 3 was a big leap forward from ChatGPT 2. The latest GPT-4 and GPT -4o are even better – they’re faster, more efficient, and less likely to hallucinate than GPT-3. But we are still using them the same as we used ChatGPT 2 and the difference in their responses are basis the data they are trained upon. Hence the returns in innovation, ChatGPT for example, is diminishing and until the next GPT 5 releases, this would be open to debate.
In a nutshell, the companies being not able to realize their investments made in AI, startups with billion dollar fundraises throwing the towel early, and the rate of innovation being slowed down are possibly the reasons why we maybe in an AI bubble and it’s not completely wrong, if you think, but what is against that?
The Nays
It is not completely new for emerging tech to be profitable from the start. Historically, tech companies have been unprofitable for very long periods. Amazon remained unprofitable for nine years, Uber, after 15 years of its existence, turned profitable only this year. This is because whenever a new technology emerges it takes time for public to understand it and truly realize its benefits and something similar is happening with this AI boom, it will take time for everyone to understand how it is going to benefit them.
According to big tech CEOs, AI is the largest opportunity in this generation and in their view, a pull back now would mean to risk ceding the race to their competitors. Google’s CEO Sundar Pichai told investors on an earning call last month - “In tech, when you are going through transitions like this.... the risk of underinvesting is dramatically higher than overinvesting”
Startups failing is not uncommon. Just because handful of startups like Character.AI threw in the towel doesn’t means entire industry is failing. Just look at image generator Midjourney, which was expected to generate $200 million last year and has reportedly been profitable since its earliest days.
And innovation? We have actually been focusing too much on how the next ChatGPT model will look like rather on focusing how much more innovation can be made from current GenAI models we have. For instance, look at Google’s DeepMind, just last month the company used a custom model to win a silver medal at the International Math Olympiad. Also, they build an amateur table tennis robot that beat every beginner who was ready to give it a challenge and of course lost to every advance player. And you know what? these innovations came just couple of months after DeepMind unveiled AlphaFold 3, which predicts protein structures with amazing accuracy. Now we don’t know how long it will take Google to turn DeepMind into profit but one thing is for sure – GenAI can make billions in critical applications like robotics, healthcare and medicine.
The Verdict
The dot-com bubble was formed because valuation of companies was inflated based on nothing other than just hype. Investors poured billions into companies with no established business model and when the profits didn’t come along, they pulled out, salvaging what they can. One of such company was Webvan - a grocery delivery service (Yup Groffers and Big Basket aren’t that unique), which went public in 1999 and its stock was doubled on first day of listing giving it a valuation of whopping $6 billion. However, the company was making a mere $10 per customer while spending $27 to fulfil the order. Over time the company accumulated $1 billion in loses and was shut down in 2001, eventually acquired by Amazon.
On the flip side if you closely look at AI boom – there is demand, innovation, and use case and they actually add great deal of value to various industries. This is evidence that AI boom is a lot different than dot-com bubble. However, that still leaves us with this question – Why AI companies and projects are not able to turn profitable if there is demand, innovation and use case? – Two things actually – Cost and user base.
And in this race of bringing the cost down and expanding user base, the big guns have an edge. They can afford to pour billions in developing and training LLMs and provide subsidized AI services while they work to bring costs down and scale user bases up. To fund this, they have their profitable businesses in e-commerce, advertising, and hardware.
While for startups? The road seems to be quite bumpy. They not only have to deal with deep pockets with likes of magnificent seven but also keep the venture capitalists happy and provide answers on how and when they will turn profitable.
However, the nastiest part about being in a tech bubble is that you don’t even realise you are in it until it bursts. And let’s say in worst case scenario, predications of AI critiques do come true and we are in bubble, will that be the end of AI?
Well, think about it this way – did the internet come to an end after the dot-com bubble burst?
Like if you loved it, Subscribe to support us & Share if you learned something new.
Sources
Comments