Our vision for building a universal AI assistant
We’re extending Gemini to become a world model that can make plans and imagine new experiences by simulating aspects of the world.
SpeakStream: Streaming Text-to-Speech with Interleaved Data
With the increasing integration of speech front-ends and large language models (LLM), there is a need to explore architectures that integrate these modalities. While end-to-end models have been explored extensively, cascaded models that stream outputs from LLMs to TTS seem to be oddly under-explored, even though they are potentially much simpler. Using traditional text-to-speech systems …
Read more “SpeakStream: Streaming Text-to-Speech with Interleaved Data”
Revolutionizing earth observation with geospatial foundation models on AWS
Emerging transformer-based vision models for geospatial data—also called geospatial foundation models (GeoFMs)—offer a new and powerful technology for mapping the earth’s surface at a continental scale, providing stakeholders with the tooling to detect and monitor surface-level ecosystem conditions such as forest degradation, natural disaster impact, crop yield, and many others. GeoFMs represent an emerging research …
Read more “Revolutionizing earth observation with geospatial foundation models on AWS”
Create shareable generative AI apps in less than 60 seconds with Vertex AI and Cloud Run
Want to turn your generative AI ideas into real web applications with one click? Any developer knows it’s a complex process to build shareable, interactive applications: you have to set up infrastructure, wire APIs, and build a front-end. It’s usually a complex process. What if you could skip the heavy lifting and turn your generative …
Read more “Create shareable generative AI apps in less than 60 seconds with Vertex AI and Cloud Run”
FLUX.1 Kontext enables in-context image generation for enterprise AI pipelines
FLUX.1 Kontext from Black Forest Labs aims to let users edit images multiple times through both text and reference images without losing speed.Read More
WIRED Talked to a Fired DOGE Staffer About Who Was Really in Charge
Sahil Lavingia, who says he was fired from DOGE after speaking out about his experiences there, told WIRED about how he communicated with the group, who appears to be in charge, and what might be coming next.
Horses ‘mane’ inspiration for new generation of social robots
Interactive robots should not just be passive companions, but active partners — like therapy horses who respond to human emotion — say researchers.
Four-legged robot plays badminton with humans
A small team of roboticists at Robotic Systems Lab, ETH Zurich, in Switzerland, has designed, built and tested a four-legged robot capable of playing badminton with human players.
What faceswap software would this be
Saw this on Instagram, link bellow, and was stunned by how good it is, I’ve been looking for softwares like those for private content creation, I record my self and use faceswapper to make my self a video game character(mainly from rdr2) for the fun of it, but this is next level. Where can I …
Tokenizers in Language Models
This post is divided into five parts; they are: • Naive Tokenization • Stemming and Lemmatization • Byte-Pair Encoding (BPE) • WordPiece • SentencePiece and Unigram The simplest form of tokenization splits text into tokens based on whitespace.