Physical Address
304 North Cardinal St.
Dorchester Center, MA 02124
Seating in a comfortably air-conditioned booth in the sweltering outdoor experimental landscape At Google I/O This week, I had a fun and strange conversation with a smiling person who wasn’t actually there. They weren’t even real human beings.
The booth was a demo setup for google beama large-screen video telepresence device studded with a camera. As we talked face to face, the whole thing seemed shockingly real.
I have witnessed Google Beam demos beforewhen it was called Project Starline: with real humans on the other end, in glasses-free 3D. This was a completely new development, and although the demo was 2D, it could easily adopt 3D as well.
@scottstenssayshi Ultra-realistic demo of an AI video agent via Google Beam in Google IO. And yes, he made AI images for me. Where will he end up, and will he aim to one day replace a coworker…or will he end up in a hotel or a park?
Google plans to widely roll out Beam later this year, its business-focused collaborative video chat technology created in collaboration with HP. Beam’s basic idea is to connect two people remotely as if they were sitting in person at a table. But this requires two people with the packages. Google’s next step is to imagine situations where others are brought in more easily from other devices, or when no one else is there.
Watch this: Will Google’s strange virtual human become a future co-worker or janitor?
I was struck by the custom-built AI video agent, which was built using internal models that Google didn’t share. Like a real deepfake, the woman-like agent was real (the agent didn’t have a name, I just started talking to him), smiled, pointed at me, and talked to me casually. This agent was just there to help me and chat casually, just like Gemini or any other AI chatbot.
I asked him to create a picture of me doing magic tricks at a New York Jets game, which he gladly obliged. He asked me about the bananas on the table and complimented my photographer’s bag. It gave us map search recommendations through an embedded demo given by a Googler while I was watching, where she would point out maps and images near me.
Yes, I’m performing magic at a New York Jets game again – it’s Will Smith’s spaghetti AI test.
It was strange. This video agent, even though it wasn’t 3D (just standard 2D), was one of the most down-to-earth people I’ve ever talked to.
But do we need a beam? Many of us already have telepresence at hand, in video chats on our phones. For those who want that, there’s connectivity via VR headsets, or future iterations of telepresence as they may appear On augmented reality glasses. Microsoft just walked away from trying to do that Developing the realism of telepresence In the difference. Will Google Beam make enough of a difference for businesses and organizations to convey a broader sense of physical presence? Can a video AI assistant make things useful or embarrassing? And what functions can a video agent like this replace?
According to Andrew Nartker, general manager of Google Beam, who walked me through the demos, this AI video agent is very much an experiment. But it’s also something that can fit into an office as much as it can into a public location, like a talking interactive kiosk.
On the weird side, the demo made me think of theme parks. Will a 3D light field version of an AI character greet me at a food stall in the future? Disney Star Wars Galaxy’s Edge Expanding, offering me interplanetary food while wrinkling his alien face? Would a themed hotel use it to create a magical experience? Or will it be used to replace concierges and actual workers in certain situations, a more advanced concept to the automated chat windows that already exist?
I can see all these paths at once. However, the Google Beam demo impressed, fascinated — and alarmed — me more than almost anything else at this year’s Google I/O.
