Major AI vendors are in an unusual sign of solidarity this week as they call for a slowdown in AI development. So the natural question we should be asking is why?
OpenAI & Anthropic have been touting for years now that AI will become so intelligent it will take away most peoples jobs. The neccessity that you need to be using AI or you’ll fall behind. AI is an unstoppable force that you need to get on board with. FOMO has been winning as many companies feel they need to use it or else they’ll fall behind.
AI is Dangerous
Both OpenAI and Anthropic have recently publicly stated that AI is potenitally getting too dangerous. With the recent AI cyber attacks coming to light, this could make sense. Add in the AI researcher who quit and went on a PR tour touting how dangerous AI is becoming and all the data points to this being a legitimate concern.
But why would any for profit company actually call to slow down their business? There are countless examples of where a company’s products were harmful, yet they continued to promote them and even worked behind the scenes to hide the dangers (cigarette industry). What makes the AI companies different?
Most businesses exist to earn a profit. CloudAI companies have been one of the most extreme examples of companies that are not just losing money, but at rates that are hard to fathom. The sheer volume of funding required to keep the lights on for these companies is insane.
It’s not altruism
What if the reason they are calling for a slowdown in their industry isn’t because of some altruistic reason to save humanity, but far simpler… to slow down their burn rate? Sure they could have done so quietly, but setting the stage as a moral imperative keeps the idea that AI is powerful and useful.
Why would they do that, you might ask? Simple. On one hand even if they slowdown new development, they still need to keep the lights on. Their burn rate is so high, the business so fundamentally unprofitable, that they still need to continue to raise money to maintain operations. Framing the spending slowdown as a fear that we need regulations to be put in place keeps the story that GenAI is powerful and useful. We still need to contunue to invest these companies, but now to keep them running until the regulations can be put in place where they can then accelerate again.
Pop goes the bubble?
Yet, there has been a growing consensus in the past month or so that there is an AI bubble. I’ve been publically calling this a bubble for nearly 2 years. If that becomes the view of investors, the money begins to dry up. And both OpenAI & Anthropic have put their IPOs on hold.
So, is the framing true that it’s because GenAI is becoming dangerous, or is it because it’s becoming harder for CloudAI companies to raise money at their current burn rates? Maybe the bubble is ready to pop & they are throwing everything at the wall to keep the funding interest.