I wrote last week that when Google’s ChatGPT competition event was announced, it seemed like a rush job to reassure investors, but since then that event has happened, and more than anyone could have imagined. It got worse. Google’s event yielded the opposite of what was hoped for, as he fell nearly 12% from his recent high just before the event. Even Google employees are starting to notice, with CNBC’s Jennifer Elias commenting on what employees called her “hasty, failed” announcement of Google’s new chatbot. I am writing to criticize the
CNBC said it was able to see some messages from Google’s internal “Memegen” employee forums, which are usually lighthearted, but the report noted that “Bard’s posts after the announcement were more serious. He hit the tone and even went straight for Pichai.”
“Dear Thunder” began one post, “The Bard launch and layoffs were rushed, botched and short-sighted. Get back to the long-term outlook.” The post said: “Needs improvement,” as if someone were to appreciate Pichai’s performance as his CEO.
Google announced its ChatGPT competitor announcement last Wednesday, but made the bizarre decision to undermine that announcement in a blog post two days ago. We had already sent out invitations to launch our ‘New Bing’ product using , which was a day ahead of Google’s announcement, and Google wanted to be ahead.
The two events were intertwined, but couldn’t be more different. Microsoft announced the “New Bing” on Tuesday and launched the product. Despite many rate limits today, Bing + ChatGPT is open to the public and tries to answer your questions by summarizing search results into easy-to-read paragraphs. It’s also integrated into the Microsoft Edge browser, where you can answer questions, create things, and summarize pages. It is a product that you can actually try.
Google’s announcement of its “Bard” chatbot on Monday only gave an overview of the features Google was planning, and the details were so vague that anyone could write to them (answer questions in Google Chat). is a chatbot!). The Google post included him one example answer, which turned out to be wrong and resulted in an embarrassing correction by Reuters. The blog post has since been updated with new examples, highlighting the tendency of AI to generate plausible-but-wrong answers.
Google’s event did not include a public product announcement of its chatbot technology, instead opting for a private “Trusted Tester” program. The company seemed in a rush to wrap up this whole event, so of course nothing was ready to launch, but the back-to-back events only highlighted how far behind Google was. . Unlike Microsoft, Google’s event was mostly filler, with only five of his minutes out of the 40-minute event being dedicated to ChatGPT competitors, and almost all of that information was spoiled by Monday’s spoilers for him. was already present in the blog post. The rest of the event featured small updates to Google Translate, Maps, and Lens. This included a section that had to be skipped after the presenter told the audience that he “lost the phone” his Google needed to make the presentation. The live stream ended with an error his message and was marked as “unlisted” on YouTube, making it unplayable for several hours.
“Hurrying Bard to market in a panic has proven that the market is afraid of us,” one employee wrote, according to a CNBC report. Another user posted a photo of a trash can fire with the Google logo on it, saying it represented “all the emotions from last year.” Another person criticized how the company seems to be focusing on equities these days, saying, “Laying off 12,000 people would push the stock up by 3%, and a rushed AI presentation would push the stock up. 8% drop,” he said.
A real shame for Google is that they invented the key technology behind ChatGPT. The “GPT” in ChatGPT stands for “Generative Pre-trained Transformer”, and “Transformer” is a neural network architecture invented and open sourced by the Google Brain team in 2017. At the time, Google described his Transformer as “especially suited for language understanding,” but had never built a product using the technology. OpenAI, owner of ChatGPT, turns AI research into products like ChatGPT and his DALL-E that people can actually use, while Google keeps it locked up in a lab. Two events (one with a product announcement, one without) only strengthened that narrative last week.
