Google promised Gemini 3.5 Pro in June, but it spent the summer trotting out smaller Flash models. Now, Google is ready to take on the frontier again with Gemini 4 Argon. — Ars Technica
Gemini 4 Argon is Google’s “next era of frontier intelligence” and built for “deep reasoning across complex, long-horizon workflows.” It follows the cancellation of Gemini 3.5 Pro and focus on 3.8 Flash. — 9to5Google
Gemini 4 Argon is Alphabet's most advanced model yet, with major coding, cybersecurity, and complex professional work improvements. — CNBC
US District Judge Amit Mehta dismissed antitrust lawsuits from Chegg and Penske Media, which argued Google's AI Overviews used their content and cut their search traffic. The judge ruled that Google's conduct is not illegal under antitrust law.
Google is launching Guided Vision in Gemini Live on compatible Android devices, giving real-time audio descriptions of whatever the phone's camera is pointed at. The feature was built with the blind and low-vision community.
A recently patched vulnerability in ChatGPT’s Mac app could have let attackers grab sensitive data, Wired reports, underscoring that AI client software is itself a security target.
Black Forest Labs has released Flux 3 Image, supporting multi-step edits that leave the rest of a picture alone, scene composition with bounding boxes and up to ten reference images, and output up to 4K. Open weights are expected in the next few weeks.
Amazon says it will spend more than $1 billion over the next five years on community initiatives near its data centers, as backlash to its AI infrastructure buildout intensifies.
An OpenAI essay argues that advanced AI may matter most for the routine work behind breakthrough ideas, and that execution could shape the next economy and the pace of progress.
Barclays is expanding its strategic collaboration with Anthropic, scaling Claude across its global operations with the aim of improving operations and client experience.
NVIDIA argues that AI factory returns depend on hardware being productive, durable and fungible, with tokens per second per megawatt as the key measure of earning capacity. It cites SemiAnalysis data showing Vera Rubin NVL72 systems deliver over 30x the throughput per megawatt of GB300 NVL72.
President Trump reportedly asked Grok for its opinion before invading Venezuela and capturing Nicolás Maduro, and the chatbot reportedly encouraged the capture.
A Mercor study finds current AI models now beat licensed CPAs on structured accounting tasks in speed and accuracy, though no model completes the harder APEX Benchmark without human oversight.
ServiceNow AI describes AutoSynthData, a method for generating training data for enterprise agents, in a Hugging Face blog post.
Google's tensor processing units were launched into orbit aboard Planet Labs satellites on a SpaceX launch, an early step toward space-based data centers.
Google has announced Gemini 4 Argon, a frontier model aimed at real-world coding, enterprise knowledge work and cyber defense. It is not broadly available yet: select testers get access first, with the API and paid tiers to follow.
Microsoft AI has released MAI-Transcribe-2-Streaming, a model for real-time transcription aimed at voice agents.
OpenAI is adding shopping features to ChatGPT that let users virtually try on clothing and accessories using their own photos. Products can also be saved to a new Favorites library.
OpenAI's GPT-6 Astra Ultrafast, running on NVIDIA Blackwell GPUs, is now available in the OpenAI API and to eligible ChatGPT Work and Codex users. NVIDIA says it generates tokens up to 8x faster than Astra's Standard mode.
US District Judge Amit Mehta dismissed antitrust lawsuits from Chegg and Penske Media, which argued Google's AI Overviews used their content and cut their search traffic. The judge ruled that Google's conduct is not illegal under antitrust law.
Google is launching Guided Vision in Gemini Live on compatible Android devices, giving real-time audio descriptions of whatever the phone's camera is pointed at. The feature was built with the blind and low-vision community.
A recently patched vulnerability in ChatGPT’s Mac app could have let attackers grab sensitive data, Wired reports, underscoring that AI client software is itself a security target.
Black Forest Labs has released Flux 3 Image, supporting multi-step edits that leave the rest of a picture alone, scene composition with bounding boxes and up to ten reference images, and output up to 4K. Open weights are expected in the next few weeks.
Amazon says it will spend more than $1 billion over the next five years on community initiatives near its data centers, as backlash to its AI infrastructure buildout intensifies.
An OpenAI essay argues that advanced AI may matter most for the routine work behind breakthrough ideas, and that execution could shape the next economy and the pace of progress.
Barclays is expanding its strategic collaboration with Anthropic, scaling Claude across its global operations with the aim of improving operations and client experience.
NVIDIA argues that AI factory returns depend on hardware being productive, durable and fungible, with tokens per second per megawatt as the key measure of earning capacity. It cites SemiAnalysis data showing Vera Rubin NVL72 systems deliver over 30x the throughput per megawatt of GB300 NVL72.
President Trump reportedly asked Grok for its opinion before invading Venezuela and capturing Nicolás Maduro, and the chatbot reportedly encouraged the capture.
A Mercor study finds current AI models now beat licensed CPAs on structured accounting tasks in speed and accuracy, though no model completes the harder APEX Benchmark without human oversight.
ServiceNow AI describes AutoSynthData, a method for generating training data for enterprise agents, in a Hugging Face blog post.
Google's tensor processing units were launched into orbit aboard Planet Labs satellites on a SpaceX launch, an early step toward space-based data centers.
No stories for this company in today’s edition.