Recent AI failures are cracks in the magic
Recent AI failures are cracks in the magic
Trouble in trillions-land
GPT-4 went clinically insane two days ago, sounding more like a madman’s read of Finnegans Wake than anything else. Sora, OpenAI’s new video-generating model, can’t understand basic physics, cause and effect, or object permanence. Google’s Gemini, its answer to GPT-4, turns out to have a bunch of ominous guardrails, like refusing to generate images of Tiananmen Square, or refusing to generate images of caucasian men in almost any context or prompt, no matter how benign or innocent (even just walking a dog).
If these strange breakdowns and guard-rails are the state-of-the-art in AI, it’s no wonder that the economics of the industry are being called into question. Compared to when I wrote “Excuse me, but the industries AI is disrupting are not lucrative” in December, I think the situation is even worse than I initially thought. While it’s been known that AI is incredibly expensive due to the need to scale up by training larger models on more data, as well as the high-salaries and constant upkeep, huge water and energy demands, etc, Ed Zitron recently pointed out something more problematic for the industry:
In essence, Microsoft "invested" $10 billion in money that OpenAI had to spend on Microsoft's services, meaning that OpenAI would have to use Microsoft's "Azure" cloud computing service to run ChatGPT. When Google invested $2 billion in OpenAI competitor Anthropic, it did so in tranches—$500 million up front and an additional $1.5 billion over a non-specific period of time. Coincidentally, this funding round took place only a few months after Anthropic signed a multi-year deal with Google Cloud worth $3 billion, locking them into Google's compute platform in the process. Amazon also invested $4 billion in Anthropic, who agreed to a "long-term commitment" to provide Amazon Web Services (Amazon's competitor to Microsoft Azure and Google Cloud) with early access to their models—and Anthropic access to Amazon's AI-focused chips.
While Microsoft, Amazon, and Google have technically "invested" in these companies, they've really created guaranteed revenue streams, investing money to create customers that are effectively obliged to spend their investment dollars on their own services.
In other words, we should conceptualize the current state of the AI industry as a flow of excited VC investment being siphoned into a handful of major tech companies. Such a structure brings up fundamental questions.