Have you heard talk about the “AI Bubble” bursting? This is a common phrase brought up in anti-AI arguments, but what does it really mean? Is it actually going to happen?
What is the AI Bubble?
The “AI Bubble” has to do with the stock market, representing the sudden boom in the AI industry that may have valued the companies higher than they’re actually worth. Most basically, investors have suddenly spent large sums of money on investments that might not end well.
It goes deeper than that; there’s a theory that large AI companies like OpenAI, Nvidia, Oracle and more were using circular financing to artificially boost their values. Many AI related or focused companies are investing millions into each other.
The AI Bubble could also be used to refer to the fast pace at which the value of AI companies is rising. As Nvidia’s current value is higher than $4 trillion, many worry that these prices will soon show to be inflated and the price will plummet.
How expensive is AI?
Recently, OpenAI’s Sora model shut down due to costs. With the amount of money invested into OpenAI, that begs the question, how expensive is AI to run?
There are many jobs behind the creation of AI, there are people who sort the data that goes into AI, those who develop and code the AI, and more jobs, all of which demand a high salary.
Adding to the costs of paying employees, servers and GPUs are needed as infrastructure for keeping AI running. Cloud services are almost always needed, but they cost a lot to keep running.
While it’s not entirely clear how much OpenAI is losing per quarter, numbers range in the billions of dollars. Many of OpenAI’s tools like ChatGPT are completely free to use, with optional paid versions. Recently, ChatGPT has started having ads play while you use it, which might drive some users away.
What direction is AI going in?
There’s a chance that the bubble is just starting to pop, and the pattern of circular investment into AI has started to fail. A $100 billion investment between two of the AI giants, Nvidia and OpenAI, has failed. There’s a chance that money will still be invested, but not nearly as much as originally announced, which has shaken the path of investment.
AI has a severe negative impact on the environment, and while there are attempts to make a better solution that can keep AI and the state of the environment, developments aren’t very fast coming. AI is using up lots of water and other resources, some of which are limited and have no sustainability.
Another big concern with AI is that it’s “eating itself.” What this is referring to is the fact that as AI searches for more data to train itself on, it’ll find its own AI generated content, and use that as a basis for its own data. As the cycle continues, AI outputs will drift further and further from reality.
AI will also have “hallucinations” where it will make up information, or use inaccuracies from data it collected. This can also be from the nature of AI models, where they’ll try to predict the next word even if it has no data to know what it will be.
Currently, there’s a pretty big chance that the AI bubble will burst, and it might be already starting to pop. The path of circular investment is starting to fail, and not as much money is going into artificial intelligence. The bubble does exist, and it might not last for much longer.
