Topics
Latest
AI
Amazon
Image Credits:Maxwell Zeff
Apps
Biotech & Health
mood
Image Credits:Maxwell Zeff
Cloud Computing
Commerce
Crypto
Enterprise
EVs
Fintech
Fundraising
Gadgets
Gaming
Government & Policy
ironware
Layoffs
Media & Entertainment
Meta
Microsoft
privateness
Robotics
Security
Social
Space
startup
TikTok
expatriation
Venture
More from TechCrunch
effect
Startup Battlefield
StrictlyVC
Podcasts
Videos
Partner Content
TechCrunch Brand Studio
Crunchboard
Contact Us
Google launchedGemini Liveduring itsMade by Googleevent Tuesday . The feature countenance you to have a semi - instinctive spoken conversation , not type out , with an AI chatbot powered by Google ’s latest enceinte speech model . TechCrunch was there to prove it out at first hand .
Gemini Live is Google ’s answer toOpenAI ’s Advanced Voice Mode , ChatGPT ’s most identical feature film that ’s current in a modified alpha test . While OpenAI outsmart Google to the punch by demoing the feature article first , Google is the first to roll out the nail down feature .
In my experience , these low latency , verbal features finger much more natural than texting with ChatGPT , or even talking with Siri or Alexa . I find that Gemini Live react to interrogative in less than two second , and was able to pivot clean quickly when interrupted . Gemini Live is not everlasting , but it ’s the best elbow room to utilise your sound hands - gratis that I ’ve seen yet .
How Gemini Live works
Before speak with Gemini Live , the feature lets you opt from 10 voices , liken to just three voices from OpenAI . Google act upon with voice actors to produce each one . I appreciated the variety show there , and happen each one to sound very humanlike .
In one example , a Google product handler verbally necessitate Gemini Live to find family - friendly winery near Mountain View with outdoor domain and playground nearby , so that kids could potentially come along . That ’s a far more complicated task than I ’d ask Siri — or Google Search , honestly — but Gemini successfully recommended a topographic point that play the criteria : Cooper - Garrod Vineyards in Saratoga .
That said , Gemini Live leaves something to be desired . It seemed to hallucinate a nearby resort area called Henry Elementary School Playground that is supposedly “ 10 minute away ” from that vinery . There are other resort area nearby in Saratoga , but the nearest Henry Elementary School is more than a two - hour drive from there . There ’s a Henry Ford Elementary School in Redwood City , but it ’s 30 mo away .
Google liked to show off how users can disrupt Gemini Live mid - condemnation , and the AI will apace pivot . The company says this allow user to keep in line the conversation . In practice , this feature film does n’t work perfectly . Sometimes Google ’s project managers and Gemini Live were talking over each other , and the AI did n’t seem to pick up on what was said .
Join us at TechCrunch Sessions: AI
Exhibit at TechCrunch Sessions: AI
Notably , Google is not allowing Gemini Live to sing or mime any voices outside of the 10 it provides , according to mathematical product manager Leland Rechis . The company is likely doing this to obviate run - In with right of first publication legal philosophy . Further , Rechis said Google is not focused on get Gemini Live to realise emotional pitch contour in a user ’s voice — something OpenAI vaunt during its demonstration .
Overall , the feature article seems like a large way to plunk deep into a topic more naturally than you would with mere Google Search . Google notes that Gemini Live is a tone along the way toProject Astra , the in full multimodal AI model the company debuted during Google I / O. For now , Gemini Live is just able of voice conversations ; however , in the hereafter Google wants to add existent - time video recording savvy .
