- It's become clear, just over a week in, that Microsoft's new ChatGPT-powered Bing has some problems.
- Users have had creepy, unsettling exchanges with Bing, resulting in Microsoft limiting some features.
- Win or lose, though, Microsoft Bing has successfully pushed Google towards much-needed innovation.
Despite the hype, the launch of Microsoft's grand entrance into the world of generative AI has turned into something of a minor fiasco.
If you haven't been keeping up, here's the short version: Microsoft introduced a new version of Bing, its perennial runner-up search engine, powered by the same technology that underpins ChatGPT — the wildly popular chatbot created by OpenAI.
Billed as a new paradigm in searching the web, the new-and-improved Bing promised to help users answer more complex, specific questions than legacy search engines like Google could provide. Microsoft CEO Satya Nadella was not shy about his intent to use the OpenAI partnership as a way to make Google "dance."
A week and change later, the narrative has changed substantially. Reporters at outlets like the New York Times and the Verge were able to prompt Bing to output messages that were strange, unsettling, and at times genuinely upsetting. Getting Bing to give the weirdest possible responses became something of a sport on social media.
Microsoft published a blog post this week explaining that Bing can get "confused" if a conversation goes on too long, and on Friday rolled out an update that limited how long users can chat with the AI per day. Anecdotally from my own use, Microsoft also updated Bing on Friday to limit its responses to certain questions, bowing out of seemingly any query related to the chatbot itself.
All of which is to say, as my colleague Hasan Chowdhury reported this week, that Google doesn't exactly have anything to worry about just yet. The technology still shows immense promise, but the last week has been a crash course in how we're so close to a future powered by infallible, infinitely intelligent AI agents, and yet so far from it. As I heard someone say this week, Microsoft seems to be speed-running the Gartner hype cycle.
At the same time, I have to give Microsoft credit. Whether or not this new Bing establishes itself as the cornerstone of how we search the web in the future, it's clear that this whole episode has lit a much-needed fire under Google.
It's hard not to feel like Google search has fallen far from its beginnings at what was the most useful, most simple, and cleanest search engine in what was an extremely competitive market. Every search query gives piles of sponsored results, ads, and prompts that make it difficult to actually find a straight answer.
It remains to be seen how Google's ChatGPT competitor, named Bard, will turn out when it actually comes to users later this year. But it should be said that Google seems to be learning from the tempest in a teapot that's been the launch of the new Bing. As Insider has reported, Google is asking all of its employees to help test and fine-tune Bard's responses, including by making sure that it doesn't say anything that could be misconstrued as representing itself as sentient or human.
It's still not a great look for Google that the most exciting development in search engines in a very long time came from outside of its own four walls. But if the whole episode pushes Google to release a better, smarter AI chatbot that doesn't seem to totally lose touch with reality — or at least if it reminds Google that alternatives exist — well, that's what competition is for, right?