Feb 14, 2023
You might’ve heard that Google had a bad week – bad to the tune of a ~$130B USD drop in the stock price in just a few days. First, they got beaten to the punch and publicly shown up when Microsoft, long a borderline comical also-ran in the search space, announced its integration of an OpenAI-powered chatbot “co-pilot” for the Bing search engine. And then, in a spectacular self-own for the ages, Google rushed to demo a comparable chatbot integration, only for it to publicly misfire in responding to a question about the James Webb Space Telescope with a bullet-pointed answer that was about 33% bullshit.
The factual inaccuracy itself wasn’t remarkable. OpenAI’s ChatGPT makes stuff up, too – all the time. And in the big picture, $100B of theoretical money going poof won’t be nearly as important or transformative as what happens next. With that in mind, I’d argue that the real significance of last week’s chain of events was in the way that Google responded to the perceived threat. In a word, that response was… rushed. Rushed, sloppily executed, and pushed out with apparently minimal regard for the downside risk of unintended consequences.
If one were looking for an emblematic moment to capture the likely character of the generative AI arms race/land grab/shitshow to come, this would be a very strong early contender. The search space just might be a bellwether, and we’re witnessing a confluence there of potentially huge first-mover advantage + powerful/adaptable technology + tremendous hype in the market – all brewing up into a perfect storm of conditions for rushing to ship and scale products with limited vetting, flimsy guardrails, and likely insufficient regard for unintended consequences.
In short, the economics of the tech industry FOMO that was so well established in the social web age now look poised to push us all into an analogous “move fast & break things” era of generative AI. What could go wrong? 🙂 (via Jeffrey)
