Speaker 1: Which way should the duck go? Going left leads to a duck, which is a friend going right leads to a bear, which is a foe.
Speaker 2: Today Google announced Gemini, which is a new foundation for its most advanced ai, including its barred AI chatbot. Gemini is the third major update to its AI technology in less than a year. So the company is pushing really hard to try to catch up to its biggest rival here, which is OpenAI the creator of chat, GPT. I think the most [00:00:30] interesting thing about Gemini is that instead of just understanding text as an input type, which is pretty common, you type some text into a little text box for a prompt. It also understands video, it understands photos, it understands audio, and it specifically understands programming code. So that means that it natively understands all those different types of media at the same time. So that’s a step closer to how humans actually operate in the real world. And to be clear, it’s a demo video. We don’t know how it actually works [00:01:00] in the real world.
Speaker 2: We haven’t been able to actually try it yet with all this multimedia input, but it was a very sophisticated demonstration. It could recognize a gradually developing drawing of a duck and then offer some information about ducks. It could recognize hand gestures like this is a barking dog, which is pretty sophisticated. A lot of people aren’t going to necessarily understand that. It could follow a little cup game where there’s a little wadded up piece of paper under some cups, which cup has the paper underneath it? Try to figure out some magician sleigh of hand maneuvers. So [00:01:30] there’s some pretty sophisticated processing going on there. If you’re looking for expertise on something, it’s going to be able to handle much more sophisticated input and hopefully provide you with much more sophisticated output. Some of the examples they showed were handling a physics homework problem that had a little sketch drawing and some equations and some handwritten work.
Speaker 2: And so the Gemini AI was able to understand the picture, understand the handwritten equations, figure out what the error was in the students [00:02:00] trying to attempt to solve the problem, and then provide some answers on Here’s where you went wrong. There are three different versions of Gemini right now. The first version that we’re able to see is called the pro version. That is a version that powers the Bard Chatbot. Also arriving right now is the nano version, or actually two flavors of the nano version, and that is shrunk down into a size that’ll run on a smartphone, and in fact, it’s Google is sharing that for Pixel eight phone owners. And then the [00:02:30] third version is the most powerful version. It’s Gemini Ultra that won’t arrive until 2024. We’re still going to have to wait a little while for that while Google tests it to make sure it doesn’t have any safety problems. So that’s still in the wraps. It’s very likely that you’re going to have to pay extra to use Gemini Ultra. So there’s going to be a new version of Bard called Bard Advanced. Google hasn’t said whether it will cost extra money or how much it’ll cost, but it sure smells like me that that’s going to be a premium version that’ll cost extra in terms of a subscription. In all likelihood. [00:03:00] We’ll see how that comes out.
