Google had an equal — and perhaps better — version of ChatGPT working inside its company before OpenAI released its bot to the world. But it never shipped.
So why didn’t Google ship it?
Gaurav Nemade, the bot’s first product manager, shares what happened on our new Big Tech War Stories podcast. This is the show’s first-ever episode —available to premium Big Technology subscribers — and we’re debuting with a free preview today.
The key issues holding the bot back, according to Nemade, were public relations concerns and the dark cloud cast by Microsoft’s misadventure with Tay, a teenage character bot that turned Nazi overnight.
“PR always top of mind for leaders,” says Nemade. “On the other side, OpenAI — they don't give a shit about PR. For the most part, they don't. They are like, ‘Okay, this is what we think is right. This is a reasonable way of putting it out.’ They become vulnerable, they put it out, and then they work with the community.”
Nemade also shares a list of potential products his team thought LLM technology could be used for, including imbuing non-playable characters in video games with complex personalities and bots that would get you out of awkward situations by calling your phone and talking you to safety.
Google eventually did ship Bard, an LLM-powered chatbot, but Nemade is convinced it would’ve sat on the technology if OpenAI hadn’t forced its hand. More in the interview!
Big Technology’s launch special is running for the next few days. Sign up and you’ll get Big Tech War Stories, our panel of experts reacting to major news events, Kristi Coulter’s Amazon column, and my weekly Friday stories. This deal is $90 for the year, or 50% off the monthly price. Check it out →
Big Technology AMA
Tomorrow, Tuesday, October 31, at 11 a.m. pacific time / 2 p.m. eastern, I am hosting an ‘ask me anything’ session here on Substack. You can join, ask anything, and register here: https://lu.ma/Alex-Kantrowitz-ama
PS: Here’s our Big Tech War Stories Podcast art, courtesy of Dall-3