TLDRocket
Sign in
Latest Brian Chesky interview: AI agents need their own operating system — TechCrunch AI-driven law firm Arceus Legal raises $17M to move beyond contract wo... — SiliconANGLE AI is consuming human knowledge. It should learn how to pay for it — Fortune Ken Griffin’s $3 billion gift exemplifies Florida’s new mindset: from... — Fortune There’s a pipeline of deals—for pipelines—that will help power the AI... — Fortune HPE’s $1.2 billion order from Vultr puts its AI networking strategy to... — Fortune Photon held a funeral for mobile apps. Now it has $4.5M to help replac... — TechCrunch Use Graph RAG when relationships are part of the evidence — The New Stack

The AI intelligence platform

Every AI story that matters — and the intelligence behind it.

TLDRocket reads all relevant sources, removes duplicate coverage, and publishes a short neutral summary of every story, linking back to the original. Free, no spam, unsubscribe anytime.

Add to Slack

Every story also updates live profiles event timelines weekly rankings the AI Market Index

Friday, 12 December 2025

Improved Gemini audio models for powerful voice experiences

Google DeepMind 9 months ago 31

Google released an updated Gemini 2.5 Flash Native Audio model that improves voice agents' ability to handle complex workflows, follow instructions, and maintain natural conversations. The model achieved a 71.5% score on ComplexFuncBench Audio for multi-step function calling and reached a 90% adherence rate to developer instructions, up from 84%. The update enables live voice agents across Google products and introduces live speech-to-speech translation in Google Translate, supporting over 70 languages while preserving speaker characteristics like intonation and pacing.

Spotlight on innovation: Google-sponsored Data Science for Health Ideathon across Africa

Google Research 9 months ago 7

Google hosted an Africa-wide Data Science for Health Ideathon in collaboration with three pan-African communities, where teams developed AI solutions using Google's open health models like MedGemma and MedSigLIP to address healthcare challenges across the continent. Six finalist teams were selected from over 30 submissions and received mentorship and Google Cloud credits to develop solutions ranging from cervical cancer screening to maternal health support. The winning team, Dawa Health, is piloting their cervical cancer screening tool with plans to scale to 50,000 patients next year, demonstrating how localized AI applications can address specific African health priorities.

Announcing Together Python SDK v2.0

Together AI 9 months ago 22

Together announced Python SDK v2.0, a release candidate that replaces the legacy v1 SDK with a modern, type-safe, OpenAPI-generated client built using Stainless. The new SDK uses httpx instead of requests and is approximately 20ms faster per request on average. The upgrade introduces breaking changes across multiple APIs (Files, Batches, Endpoints, Evals, Code Interpreter) and new capabilities including beta Instant Clusters support, improved error handling, and raw response streaming helpers.

BBVA and OpenAI collaborate to transform global banking

OpenAI 9 months ago 30 ● 4 sources

BBVA is deploying ChatGPT Enterprise across its workforce as part of a multi-year partnership with OpenAI to develop AI solutions for banking operations and customer service. The bank will roll out the tool to all 120,000 employees. BBVA aims to build banking services designed around AI capabilities rather than retrofitting AI into existing systems.

BNY builds “AI for everyone, everywhere” with OpenAI

OpenAI 9 months ago 17 ● 4 sources

BNY Mellon is deploying OpenAI technology through a platform called Eliza to enable employees across the company to build custom AI agents for their work. More than 20,000 employees have access to the system for creating AI applications tailored to specific business tasks. The initiative aims to increase operational efficiency and improve how the bank serves its clients by distributing AI development capability beyond specialized technical teams.

How We Used Codex to Ship Sora for Android in 28 Days

OpenAI 9 months ago 11

OpenAI deployed Sora to Android in 28 days using Codex to accelerate development through AI-assisted coding and planning. The team completed the project in 28 days using parallel coding workflows and automated code translation. This speed enabled a small team to ship a complex video generation app to a new platform faster than typical mobile development cycles.

The daily briefing

Every AI story that matters, in your inbox by 8am.

TLDRocket reads all relevant sources, removes duplicate coverage, and summarises the day in two minutes. Follow companies and topics for alerts, or get the briefing in Slack. Free, no spam, unsubscribe anytime.