Samuel Thompson

NVIDIA is optimizing its global hardware supply chain by integrating Palantir Foundry and its cuOpt technology. This solution automates critical allocation decisions, using cuOpt for mixed-integer linear programming to minimize “Time of Ownership.” NVIDIA also fine-tuned its Nemotron AI model on qualitative data to enhance operational insights and decision accuracy, achieving significant improvements over previous benchmarks. This closed-loop system continuously learns from real-world performance for future reinforcement learning initiatives.

Supply chain disruptions cost businesses $184 billion annually. While AI excels at detecting problems, swift action remains a bottleneck due to reliance on manual intervention and “ticket” systems. The next frontier is “bounded action,” where pre-authorized, rule-based decisions allow automated agents to execute critical responses immediately. This requires codifying decisions as policies, enabling execution systems for machine-initiated transactions, and evolving accountability. Companies empowering bounded action will gain a significant competitive advantage by drastically shortening response times.

JD.com is launching a “Physical AI Acceleration Plan” to hyper-automate its logistics with 3 million robots, 1 million autonomous vehicles, and 100,000 drones over five years. This builds on its existing automation, integrating hardware with the advanced “Meta Brain” AI system. The company is significantly increasing R&D and capital expenditures to enhance its global logistics network and adapt its workforce to automation.

CloudNC, a UK-based AI precision machining innovator, has secured $20 million in funding led by Nimble Ventures. The investment will accelerate the global rollout of its AI technology, which automates CNC programming via its CAM Assist software. This innovation streamlines manufacturing workflows, reduces lead times, and enhances productivity. CloudNC also plans to launch Quote Agent, an AI-powered quoting tool, and is pursuing FedRAMP authorization for government deployment.

Samsung is partnering with Mistral AI to integrate on-premises artificial intelligence into its semiconductor manufacturing. This collaboration, announced at a state summit, will leverage Mistral’s software, including the Mistral Large model, to create specialized AI for intelligent factory operations. The aim is to enhance efficiency, precision in defect detection, and accelerate development cycles, ultimately stabilizing production yields. Samsung also invested in Mistral AI’s funding round, solidifying a commitment to co-innovation.

Arm is standardizing physical AI and robotics with “Arm Total Design.” This initiative unites over 80 partners to overcome fragmentation in industries like mining and logistics, aiming for a $200 billion compute market. It introduces a Robotics Capability Framework, inspired by SAE driving levels, to categorize robot sophistication. Virtual platforms accelerate pre-silicon development, as demonstrated in automotive, enabling earlier integration and validation of AI models, software, and hardware for smarter, more capable physical systems.

Coca-Cola is using AI in Malaysia via its “Coke Buddy” platform and “Perfect Basket” feature to optimize retailer orders. This AI analyzes historical data, seasonality, and weather to recommend optimal product stocking and quantities. The system empowers thousands of retailers, offering personalized suggestions while retaining their final decision-making authority. This initiative aims to boost efficiency and sales across Malaysia and is being expanded globally.

Google introduces WeatherNext 3, an advanced AI model offering hyper-accurate, hourly weather predictions crucial for the energy sector. This system challenges existing providers by delivering granular data on wind speed, cloud cover, and solar irradiance, essential for optimizing renewable energy generation and managing increasing AI-driven electricity demand. Accessible via Google Cloud and integrated into various Google services, WeatherNext 3 aims to provide a competitive edge in the growing energy forecasting market.

MG Ship has launched an AI-driven module optimizing shipping routes and carrier selection for global retailers. Early deployments show significant cost reductions and transit time improvements, with rapid ROI in dynamic route planning, demand forecasting, and documentation processing. The platform integrates real-time data and predictive analytics, offering sophisticated carrier evaluation beyond price to enhance OTIF rates and profitability.

M&T Bank is aggressively integrating AI, deploying AI copilots to over 15,000 employees to boost efficiency and customer service across operations, development, and risk management. The bank has built a strong technology and data foundation since 2018, significantly reducing outages and increasing system upgrades. AI applications range from content generation and summarization to identifying customer needs and flagging portfolio risks, mirroring industry-wide adoption of AI in finance. M&T emphasizes human oversight and data governance in its AI strategy.