Demand for artificial intelligence (AI) chips, especially ultra-powerful GPUs Nvidia (NASDAQ: NVDA) used to train the most advanced AI models seems insatiable. Nvidia’s data center segment generated more than $30 billion in revenue in the third quarter alone, up nearly 10-fold from two years ago. Tech giants are rushing to build AI data centers packed with Nvidia’s GPUs, spending untold billions on hardware.
So far, every new AI model coming out of OpenAI, one of those tech giants, or anyone else dealing with AI fever has significantly improved over its predecessor. OpenAI’s GPT-4 is much more capable than GPT-3, and Alphabet‘s Gemini AI models blow its older models out of the water. But those improvements came at a price.
Start your morning smarter! Wake up with Breakfast news in your inbox every market day. Register for free »
Training GPT-4 is estimated to have cost around $100 million, while GPT-3 may have cost only a few million dollars. Anthropic CEO Dario Amodei expects the next generation of AI models to cost about $1 billion to produce. Buying and then deploying many thousands of powerful GPUs is expensive, and collecting large amounts of training data is no easy feat either.
Major language models like GPT-4 work by predicting the next token in the output. This works quite well in many cases. The best AI models can produce high-quality text, generate compelling images, and even use fairly advanced reasoning. More training data generally yields better results, as does more time spent incorporating that data into the training process.
However, AI companies appear to be reaching a limit. The pace of improvement in AI models is slowing, even with more data and more computing power. Marc Andreessen, co-founder of venture capital firm a16z, recently noted that AI models seem to hit a ceiling of capabilities, regardless of the amount of data or computing power coming at them.
A major breakthrough could help AI companies break this ceiling, but it’s also possible that LLMs simply can’t do much anymore. The rising demand for AI chips is driven by the idea that training a $1 billion or $10 billion AI model makes financial sense. What if that’s not the case?
If AI models have largely peaked in terms of capabilities, the frantic multi-billion-dollar AI investments that tech giants are making in an effort to avoid being left behind may never pay off in terms of revenue or profits. The hangover from this overinvestment could be brutal for companies like Nvidia as demand for AI chips dries up.
Nvidia has delivered incredible returns to investors and has absolutely dominated the AI chip market. But it’s important to remember that trees don’t grow into the sky. An AI breakthrough that blows through the apparent LLM ceiling is certainly possible, but it’s also possible that AI technology, like almost every new technology throughout history, has been overhyped to some extent.
Artificial intelligence won’t disappear if AI models stop advancing by leaps and bounds, but Nvidia’s incredible growth and profits certainly will. With a market cap of over $3 trillion, Nvidia stock seems like a risky business.
Have you ever felt like you missed the boat on buying the most successful stocks? Then you would like to hear this.
On rare occasions, our expert team of analysts provides a “Double Down” Stocks recommendation for companies they think are about to pop. If you’re worried that you’ve already missed your chance to invest, now is the best time to buy before it’s too late. And the numbers speak for themselves:
-
Nvidia: If you had invested $1,000 when we doubled in 2009, you would have $355,011!*
-
Apple: If you had invested $1,000 when we doubled in 2008, you would have $44,516!*
-
Netflix: If you had invested $1,000 when we doubled in 2004, you would have $470,586!*
We’re currently issuing ‘Double Down’ warnings for three incredible companies, and another opportunity like this may not happen anytime soon.
See 3 “Double Down” Stocks »
*Stock Advisor returns November 25, 2024
Suzanne Frey, a director at Alphabet, is a member of The Motley Fool’s board of directors. Timothy Green has no position in any of the stocks mentioned. The Motley Fool holds positions in and recommends Alphabet and Nvidia. The Motley Fool has a disclosure policy.
Prediction: This Huge Risk Could Derail Nvidia Stock. Originally published by The Motley Fool
