AI News, 17 November 2026

Tuesday 17 November

(3rd LD) PM announces package of measures to prevent cheating on college entrance exam

AI-generated summary Prime Minister Han Seong-sook announced measures to shield the nationwide college entrance exam from cheating using artificial intelligence. The measures include banning all smart devices from test rooms, deploying additional personnel, and covering all buttons of test takers clothing with stickers.

Earlier:

AI progress outpaces governance, Stanford study warns

AI-generated summary Stanford HAI administered the third AI100 study, which warns that artificial intelligence is transforming society faster than expected while governance struggles to keep pace. The study reports that progress relies on ever larger datasets and computing resources, while facing constraints from scarce high quality data and prohibitive training costs.

Earlier:

10 AI and Workforce Transformation Events You Can’t Miss at SFF 2026

AI-generated summary The Singapore FinTech Festival 2026 highlights artificial intelligence and workforce transformation across multiple sessions scheduled from 18 to 20 November at Singapore EXPO. CFTE notes that workers possessing artificial intelligence skills command a 62 percent wage premium, while one in three surveyed employers report replacing entry-level roles with artificial intelligence.

Chat-based AI lets users run complex alloy calculations on their own computers

AI-generated summary Researchers at Empa have developed an artificial intelligence framework called OptiMat Chat that lets users run complex alloy calculations on their own computers using plain language queries. The platform utilizes a conversational agent named OptiMat Alloys to select domain-specific tools, execute predefined workflows and summarize results for expert review.

4-bit training: Six times faster on paper, 8 percent in practice

AI-generated summary Thinking Machines Lab explained in its blog how it trained Inkling 2 Small, a mixture-of-experts model with 594 billion parameters. The lab says the model was trained using a native 4-bit forward pass from the first step of pretraining to the last reinforcement learning run.

Earlier:

Swiss neurotech Indivi acquires Cogitat to advance AI-powered brain monitoring

AI-generated summary Neurotech company Indivi has acquired Cogitat to advance AI-powered brain monitoring using deep learning models that decode cognitive function from EEG brain signals. Indivi states that the acquisition will help detect the effects of experimental Alzheimer's treatments and strengthen research and development across the life sciences.

Tess 4.5: Open 327 MB Coding LoRA for Inkling 2 Small

Original midnight violet editorial illustration of a glowing amber adapter prism connected to a large glass neural matrix with fine code-like light trails.

AI-generated summary Continuity Research developer Migel Tissera has released Tess-4.5-594B-A17B, which is a coding-focused derivative of Thinking Machines' Inkling 2 Small. The developer reports training on 46.9 million tokens from anonymized engineering sessions and releasing a separate LoRA adapter of about 327 MB alongside a full merged checkpoint.

Earlier: