More than 20M of you are now using Codex! So exciting to see so many people building with agents, and we’re just getting started. Enjoy Tibo’s banked reset and make something great!
BIG TECH
-
Next-Generation Siri Uses Messages, Apps and Real-Time Screen Context
By
–
This assistant does not just respond. It understands your full personal context: → Messages
→ Emails
→ Photos
→ Apps
→ Voice
→ Camera
→ Your screen, in real time -
Lovable Cloud Deal With Google Targets Fivefold Usage Growth
By
–
Lovable Signs Multiyear Deal With Google Cloud to Up Usage 5x
#AI #CloudComputing #GoogleCloud #MachineLearning #Startup #Innovation @HaroldSinnott @fogoros @iainljbrown @NandoDF @katecrawford @drhassanrashidi @YuHelenYu -
Amazon Research Explores Treating Knowledge Corpora as Part of Agent Harnesses
By
–
Very interesting new work Amazon and colleagues. (bookmark it) This connects to the emerging theme of treating the corpus as part of the harness. Planning loops, tool protocols, and context management have matured fast, while the knowledge an agent investigates over still sits
-
Alibaba Zhenwu AI Chips Support Its Cloud Infrastructure Expansion
By
–
Yet another example of the semiconductor renaissance, Alibaba’s “Zhenwu” AI chips continue to deliver for the company’s AI services in the cloud. Compared to hundreds of billions in planned spending by Anthropic and OpenAI, Alibaba’s $56 billion (¥390 billion) over three years
-
Meta WildArtifactBench Uses Preference Judges to Evaluate Real-World Agent Deliverables
By
–
Today we’re also previewing WildArtifactBench, an internal evaluation framework designed to assess agents on complex, real-world tasks across diverse deliverable formats. By using win rates and Elo scores from human and agentic preference judges rather than strict ground-truth
-
Meta Uses Muse Spark to Caption Training Data for Generative Media
By
–
Internally, Muse Spark is deployed in several areas including media generation. For example, the model works with Muse Image for agentic media generation and produces detailed captions as training data for Muse Image and Muse Video. Muse Spark is also able to translate raw text,… pic.twitter.com/TLXX4uGQUo
— AI at Meta (@AIatMeta) 20 août 2026Internally, Muse Spark is deployed in several areas including media generation. For example, the model works with Muse Image for agentic media generation and produces detailed captions as training data for Muse Image and Muse Video. Muse Spark is also able to translate raw text,
-
Muse Spark 1.2 Supports Visual Coding, Physical Action and Audiovisual Workflows
By
–
Muse Spark 1.2 supports a broad range of multimodal tasks, from turning visuals into working code to translating perception into physical action. It also brings robust audio-visual understanding to enable video-heavy workflows common in real-world enterprise use.
— AI at Meta (@AIatMeta) 20 août 2026
Today, we’re… pic.twitter.com/ryDgr78s2pMuse Spark 1.2 supports a broad range of multimodal tasks, from turning visuals into working code to translating perception into physical action. It also brings robust audio-visual understanding to enable video-heavy workflows common in real-world enterprise use. Today, we’re
-
Muse Spark 1.2 Combines Visual Coding, Robotics Planning and Audiovisual Understanding
By
–
1/ muse spark 1.2 is a very strong multimodal model—it can do visual coding, robotics planning, and audio-visual understanding that all come together through agentic tools. pic.twitter.com/WgXGnW6crn
— Alexandr Wang (@alexandr_wang) 20 août 20261/ muse spark 1.2 is a very strong multimodal model—it can do visual coding, robotics planning, and audio-visual understanding that all come together through agentic tools.
-
Google Adds Disclosures for AI-Generated Ads
By
–
Google will now tell you if an ad was made with #AI
by Stevie Bonifield @verge Learn more: https://
bit.ly/3TqPRkc #ArtificialIntelligence #GenerativeAI #DigitalAdvertising #AIEthics