President Donald Trump ordered the Pentagon to scale back joint military exercises with South Korea, citing costs and a hostile signal to North Korea's Kim Jong Un ahead of Ulchi Freedom Shield.

Follow this section to personalize your feed and get instant alerts. Follow Go to your personalized feed WHY FOLLOW? Smart Alerts: Get notified about major news as it happens. Follow this tag to personalize your feed and get instant alerts. Follow Go to your personalized feed WHY FOLLOW?

SK Hynix has announced that it will invest 54 trillion Korean won ( 38.1 billion) to build two new memory chip facilities in South Korea. The company, which is the second largest RAM and NAND chip manufacturer after Samsung, said the new fabs will help it keep pace with the explosive demand for memory products caused by the AI data center boom. The first cleanroom will start production as early as June 2029, SK Hynix officials said. The company said it would invest 35.2 trillion won ( 24.9 billion) in a Yongin fabrication plant to produce HBM and other DRAM products, and 19.1 trillion won ( 13.2 billion) for a facility in Cheongju that will manufacture NAND storage chips. "This investment is a decision made to seize opportunities in line with the market's growth speed... we reached this investment decision after a thorough review of market demand," a SK Hynix official said in a statement.

Welcome to our monthly digest, where you can catch up with any AIhub stories you may have missed, peruse the latest news, recap recent events, and more. This month, we find out about time-series anomaly detection, delve into music generation, honour award winners, and catch up on the action from the RoboCup humanoid soccer league. We caught up with Thi Kieu Khanh Ho to find out more about her work on time-series anomaly detection, what inspired her to study AI, and what she plans to work on next. This interview is part of our series featuring the AAAI Doctoral Consortium participants. In the latest in our series of IJCAI interviews, AIhub ambassador Liliane-Caroline Demers spoke to François Pachet to find out more about his work on music generation with AI.

Follow this section to personalize your feed and get instant alerts. Follow Go to your personalized feed WHY FOLLOW? Smart Alerts: Get notified about major news as it happens. Follow this tag to personalize your feed and get instant alerts. Follow Go to your personalized feed WHY FOLLOW?

Dongkun Lee, formerly of Hyundai, will oversee its strategy. Samsung has created a new division called RX or Robotics eXperience for all of its robotics projects. Its strategy will be headed by Dongkun Lee, who joined Samsung in May as an Executive Vice President. Before he joined Samsung, Lee led Hyundai's robotics strategy, including Boston Dynamics' . Samsung CEO TM Roh himself will oversee the new division.

Diffusion Large Language Models (dLLMs) break the rigid left-to-right constraint of traditional LLMs, enabling token generation in arbitrary orders. Intuitively, this flexibility implies a solution space that strictly supersets the fixed autoregressive trajectory, theoretically unlocking superior reasoning potential. Indeed, for specific constraint satisfaction tasks (e.g., sudoku puzzles), this capability has proven to be highly advantageous. However, in this paper, we reveal that for general reasoning tasks (e.g., mathematics and coding), arbitrary order generation may in fact limit the reasoning potential of dLLMs. We find that dLLMs tend to exploit this order flexibility to bypass high-uncertainty tokens that are crucial for exploration, leading to a premature collapse of solution coverage. This observation motivates a rethink of RL approaches for dLLMs, where considerable complexities, such as handling combinatorial trajectories and intractable likelihoods, are often devoted to preserving this flexibility. We demonstrate that effective reasoning can be better elicited by simply forgoing arbitrary order and applying standard Group Relative Policy Optimization (GRPO) instead. Our approach, JustGRPO, is minimalist yet surprisingly effective (e.g., 89.1% accuracy on GSM8K) while fully retaining the parallel decoding ability of dLLMs.

Imagine you want to teach a robot to push an object on a table. The standard recipe in robot learning is to collect hundreds of expert demonstrations on a real robot, train an imitation learning policy on that data, and then evaluate the policy by running it many times on the same real robot. Both stages (data collection and evaluation) are slow, expensive, and hard to reproduce: hardware breaks, lighting changes, objects drift out of place, and every new task means more hours in the lab. A natural question is whether we can replace some of this real-robot work with a simulator. Classical physics-based simulators are powerful, but building one for a new task means manually modeling geometries, contacts, friction, and deformation, and the resulting simulator often still does not match reality closely enough for policies trained inside it to transfer.

This year, RoboCup took place in Incheon, South Korea, from 2-6 July. The event saw teams take part in competitions, training sessions, and a symposium. Take a look at what the participants got up to in our round up from social media. RoboCup 2026 officially begins today! A post shared by RoboCup Federation (@robocup.official)

The forty-third International Conference on Machine Learning (ICML) took place in Seoul, South Korea from 6-11 July. We take a look at what the participants got up to during the event. The PC Chairs are presenting the welcome remarks. One of the best conference dinners I've ever had. Try TimeChat-Captioner (at #ICML2026) -- a videoLLM that generates dense, time-aware captions for long videos.