We're so early! even the current SoTA speech-to-speech models are so dumb compared to cascaded speech-to-speech (ASR + LLM + TTS) There's loads of low hanging fruits to pick still! Looking forward for open models to climb up this benchmark in 2025
MULTIMODAL AI
-
Rumor about OpenAI’s new o3 model
By
–
Top rumor for today’s OpenAI Release: o3 model, with multimodal input & output capabilities, available to $200/mo tier users Biggest impact showcase will be in coding (single prompt full app creation with much more interactive Canvas feature)
-
Excitement about Kling 1.6 and Blade Runner video
By
–
Sounds amazing! Excited to see how Kling 1.6 brings creativity to life without losing that authentic touch. Can't wait to check out the Blade Runner video!
-
11 Most Reliable AI Content Detectors: Spotting Synthetic Media
By
–
11 Most Reliable AI Content Detectors: Your Guide to Spotting Synthetic Media As #AIGenerated #content proliferates online, discover the leading #tools and technologies designed to detect #synthetic #text, #images, and #audio. Learn how these #AI #detectors work, their
-
AWS Bedrock Upgrades Image Services with Stable Image Ultra
By
–
We’ve also upgraded our image services on Bedrock. These services are designed to address specific generative AI needs, including: Stable Image Ultra: Powered by the cutting-edge capabilities of SD3.5 Large, Stable Image Ultra sets a new standard in photorealism. This image
-
Microsoft Rolls Out Vision to Copilot Pro on Windows
By
–
Something special is coming this holiday break As of today, Vision is rolling out to our U.S. Copilot Pro subscribers on Windows—so if that includes you, keep your eyes peeled in Edge! And if you still have some Christmas sweater shopping to do…
-
New Gemini 2.0 Flash Thinking Model Available
By
–
BREAKING : A new "gemini-2.0-flash-thinking-exp" is now available on Google AI Studio According to the description, it also should be exposing its thinking process.
-
New Gemini 2.0 Model Coming Soon
By
–
Looks like we may be getting a new 2.0 Gemini model soon > "gemini-2.0-flash-thinking-exp" – will it be 2.0 Flash-based Deep Research? h/t @swishfever
-
Genesis: Generative Physics Engine Creates 4D Dynamical Worlds
By
–
introducing Genesis project , a generative physics engine able to generate 4D dynamical worlds powered by a physics https://t.co/ohgNLpiAim
— 机器之心 JIQIZHIXIN (@jiqizhixin) 19 décembre 2024introducing Genesis project , a generative physics engine able to generate 4D dynamical worlds powered by a physics
-
Sora’s Animal Physics: Dog Leaps Similar to Generated Fox Videos
By
–
some of the animal physics shown here in the sora videos– like the way the dog leaps — are really similar to what i found with videos i've generated with sora (such as of foxes jumping in snow). https://t.co/Mimhhv2vMS
— Rachel Metz (@rachelmetz) 18 décembre 2024some of the animal physics shown here in the sora videos– like the way the dog leaps — are really similar to what i found with videos i've generated with sora (such as of foxes jumping in snow).