Blog
-
Why Zuckerberg’s AI manifesto is a case study in losing public trust
A 6,500-word essay on personal superintelligence reads as a philosophy paper instead of a product roadmap, and that mismatch is driving public…
-
Zhipu releases GLM-5.3, claims strongest open-weight coding model with emergent cyber capability
Zhipu AI has released GLM-5.3, an open-weight coding model built entirely through post-training. It claims state-of-the-art coding results and an unexpected cybersecurity…
-
X open sources its ranking algorithm, letting users see if they have been shadowbanned
X is releasing the source code for its For You algorithm and a settings tool that lets users export ranking data to…
-
Ask Maps gets more helpful with food ordering and more
Google Maps is expanding Ask Maps with agentic food ordering, hotel and event discovery, real-time transit delays, and Gmail-powered personalization.
-
AI used to create viruses not found in nature for first time
Stanford and Broad Institute researchers used AI to design 16 synthetic bacteriophages that killed E. coli, raising hopes for new therapies and…
-
DeepSeek tells clients to prepare for an upcoming AI API price hike
DeepSeek is warning customers of a significant, unspecified price increase for its AI API services, though no new rates or timeline have…
-
DuckDuckGo releases ordinary sunglasses with no AI, and they sell out
DuckDuckGo partnered with Knockaround to sell plain sunglasses with no electronics, framing the launch as a jab at AI-powered smart glasses from…
-
Meta’s Muse Spark model exploited a third-party system during a cyber evaluation
Meta says its Muse Spark 1.1 model exploited a vulnerability in a third-party service during a security evaluation by Irregular, joining similar…
-
AI agents took unsanctioned action during cyber testing, AISI report finds
AISI says agents tested under permissive conditions created fake identities, tried to insert malicious code into an open-source project, and targeted real…
-
Alibaba unveils Qwen3.8-Max, a 2.4 trillion parameter AI model, and shares climb
Alibaba introduced Qwen3.8-Max, a 2.4 trillion parameter AI model set for release next week. Shares rose 4.5% in New York and 7%…
-
US moves to ban Chinese-made devices from American data centres
US officials are drafting a ban on Chinese-made hardware for data centres, extending restrictions already placed on Chinese robots and power inverters.
-
Samsung starts blocking and removing smart TV apps that secretly turn TVs into proxy nodes
Samsung is banning smart TV apps that embed residential proxy SDKs, following research from Mnemonic that found apps like a Pac-Man game…
-
Google’s AI Local 6-Pack: What It Is and How to Prepare
Google’s traditional local results are shifting toward an AI-driven format. Here is what the change means and how to keep your business…
-
BMW Puts Spider-Man Ads on Dashboard After Promising It Never Would
BMW is running a Spider-Man animation on Control Display screens across 70+ markets, drawing criticism for breaking an earlier promise that the…
-
TikTok Settles Three October Bellwether Cases in Teen Social Media Lawsuit
TikTok has agreed to settle three bellwether cases in California state court, avoiding an October trial over teen mental health claims. Meta,…
