1/7 I'm really excited to announce something special we've been working on- the Conversational Video Interface. We think it's going to change how we interact with computers and LLMs completely. Check out the live demo and talk to Carter on our website 🧵

@grvydev That definitely helps For every government subsidy helping Chinese EV startups, there’s complacency and bureaucracy from our American manufacturers that has left us falling behind unfortunately (and I say this as an American car zealot)

RT @kwindla: Join us for the first Voice AI Meetup of the new year, Tuesday Jan 14th. I'll be moderating a panel discussion about Voice AI…

RT @kwindla: Video AI New Year's Resolution Coach @axbehr from @heytavus built a New Year's resolution accountability app, using the Tavus…

RT @CerebrasSystems: Tavus is building the first, real-time digital clone, now powered by Cerebras Inference, to deliver an instant and nat…

Convinced machines are already sentient

@mo_baioumy @alexocheema Will definitely need it :) I’ll DM ya

@mo_baioumy @alexocheema A new goal of mine is to get it working on a Macintosh SE with 4MB Ram. Will need to rewrite data types, probably convert the floating point ops and remove POSIX dependencies

RT @blader: in just 3 hours, we are at $75K of our $100K goal. 🥹 yi and are in tears. we are so grateful to all of you. we decided to per…

RT @svpino: Hyperrealistic replicas of yourself are here! These guys are killing it with digital twins and letting you clone yourself into…

It's been fun hearing about everyone video chatting with AI Santa (especially the TikTokers trying to brain rot Santa), but this made the team's day

@krandiash Congrats friend- y’all are killing it!

@Vapi_AI Hell yeah! Congrats friends :) Excited for what’s to come

@svpino @heytavus Thanks! Let us know what their reaction is :)

Remember writing letters to Santa? We got him on video calls now. Really fun demo from the team on using CVI to build these immersive experiences. You can fork and make your own AI Santa!

RT @sundarpichai: Introducing Willow, our new state-of-the-art quantum computing chip with a breakthrough that can reduce errors exponentia…

Yikes- you're right, we need to make it more clear here! AI video definitely isn't cheap to run, which is why we (tried) to structure our plans as usage based. We explain that on below, but it looks like it didn't make it into the in product page as clearly. We're going to fix that now :) Also, when a conversation call is started, the conversation begins being billed, as that's when the GPU starts running. We set a default 5 min timeout to prevent runaway usage. Looks like our support team has already been in touch to refund this invoice. Thanks for the feedback here, and let me know if I can help with anything else!

@AutoExpress if a cybertruck and a rolls royce phantom had a cursed child

Super interesting paper: with current LLMs, just 2 hours of interview data can predict someone's decisions with 85% accuracy. Great writeup from @techreview!