Schneider Electric, NVIDIA, and HPE have launched Industrial AI Agents designed for deployment at the edge. The solution bundles HPE Private Cloud AI with NVIDIA technology and Schneider's EcoStruxure platform, targeting operational technology monitoring, cybersecurity, and energy optimization. CMPC Chile in Latin America is already using the underlying technology, showing early real-world traction for keeping sensitive industrial data secure while cutting costs.
Meanwhile, New Brunswick, New Jersey is putting AI to work in a very different setting. The city now uses an on-premise tool called Isermo to provide real-time transcription and translation during City Council meetings. The system operates without cloud or data centers and achieves roughly 98% accuracy across about 100 languages. It also identifies which council member is speaking, helping break down language barriers and improving access for deaf or hard-of-hearing residents.
On the research front, a team from Cornell has raised concerns about how AI companies measure harm. They argue that the industry focuses too heavily on short-term risks while overlooking long-term effects, particularly the possibility that years of AI chatbot use could gradually replace human relationships. Current safeguards like usage limits only work per message or session, leaving broader social patterns unaddressed. The researchers are calling for better measurement tools and data access to track whether AI is contributing to social isolation at scale.
In market news, China's AI and semiconductor sector has experienced a sharp downturn. The mainland STAR50 index dropped 40% in less than four months, wiping out approximately $480 billion in market value. Concerns over AI demand and export restrictions fueled the decline, stalling China's two-year rally and its broader push to stabilize its economy. The crash raises questions about global market ripple effects and China's plans to draw more savers into equities.
Key Takeaways
- Schneider Electric, NVIDIA, and HPE launched Industrial AI Agents for edge deployment, targeting OT monitoring, cybersecurity, and energy optimization.
- CMPC Chile in Latin America is already using the foundation technology behind the new Industrial AI Agents.
- New Brunswick, New Jersey uses on-premise AI tool Isermo for real-time council meeting translation at 98% accuracy across about 100 languages.
- Cornell researchers urge AI companies to better measure long-term social isolation risks from extended chatbot use.
- China's STAR50 index fell 40% in under four months, erasing roughly $480 billion in market value amid AI and export concerns.
- Mark Weatherford from Nvidia says CISOs should embrace AI, accept more risk, and hire young AI talent through paid internships.
- A study of 1,222 participants found that relying on AI for simple tasks can reduce motivation and persistence when the tool is removed.
- LeafSkill, a new action grounding layer, improved LLM agent task completion on 200 Robotouille tasks and cut token use by 14.42%.
- Nearly 1,100 participants in a study rated responsive AI as less conscious than even the least responsive human.
- University of Nottingham researchers built a self-driving lab using AI and robotics to automate discovery of heat-resistant coatings.
Schneider Electric NVIDIA HPE launch edge AI agents for industry
Schneider Electric, NVIDIA and HPE introduced Industrial AI Agents at the Edge. The solution combines HPE Private Cloud AI with NVIDIA technology and Schneider Electric's EcoStruxure platform. It targets OT monitoring, cybersecurity and energy optimization. CMPC Chile in Latin America already uses the foundation technology OTMS. The goal is to keep sensitive data secure at the industrial edge while cutting costs.
Schneider teams with HPE and Nvidia on industrial AI agents at the edge
Schneider Electric partnered with HPE and Nvidia to develop industrial AI agents for edge deployment. The initial applications focus on OT monitoring, cybersecurity and energy optimization. The companies aim to integrate AI into existing OT systems for real-time insights. This collaboration is a step toward a comprehensive AI platform for industrial automation.
LeafSkill improves LLM agents with reusable action grounding
LeafSkill is a new layer that helps recursive LLM agents reuse action grounding across tasks. It uses a lightweight 1.7B selector to choose bound actions or defer subtasks. On 200 Robotouille tasks, LeafSkill completed 137 tasks versus 121 for ReCAP. It also reduced Qwen3.5-27B tokens by 14.42% and calls by 14.08%. The approach improves task completion without replacing high-level planning.
AI assistance for simple tasks may erode motivation and persistence
A study with 1,222 participants found that relying on AI for simple problems can harm motivation. Researchers ran three experiments with math fractions and reading comprehension tasks. When AI was removed, participants who used it performed worse and skipped more problems. The findings suggest AI offloading reduces persistence across different domains. The study highlights concerns about long-term learning effects.
China's AI bubble has already burst as valuations collapse
China's AI and semiconductor stocks have crashed while global markets wobble. The mainland STAR50 index fell 40% in less than four months, wiping out about $480 billion in market value. Concerns over AI demand and export curbs drove the meltdown. This downturn stalls China's two-year rally and its push to break boom-bust economic cycles. The crash could also affect global markets and China's plans to attract savers to stocks.
Americans fear AI more than a loved one's serious illness
A survey found that Americans are more afraid of AI than a loved one getting cancer or seriously ill. The fear of AI ranks just below fears of corrupt government and a new world war. Researcher Christopher Bader says these concerns will shape national conversations for years. However, some experts say the rankings should be taken with caution since many other fears have decreased.
AI is changing the CISO role, says Nvidia's Mark Weatherford
Mark Weatherford from Nvidia says CISOs must help businesses use AI while also managing the risks it brings. He spoke at the ISC2 Strategic Congress 2026 in Denver about how fast technology changes. He advises CISOs to be curious, accept more risk, and learn new AI topics weekly. He also says companies should hire young AI talent through paid internships.
New Brunswick uses on-premise AI for council meeting translation
New Brunswick, New Jersey now uses an on-premise AI tool called Isermo for real-time transcription and translation at City Council meetings. The system works without the cloud or data centers and is about 98 percent accurate. It can translate into about 100 languages and identifies which council member is speaking. The tool helps break down language barriers and improves access for deaf or hard of hearing residents.
Researchers urge better data on long-term AI use and isolation
Cornell researchers say AI companies focus too much on short-term harms and not enough on long-term ones. They worry that months or years of AI chatbot use could slowly replace human relationships. Current safeguards like usage limits only work per message or session. The researchers call for better measurement tools and access to data to track whether AI is causing social isolation at a large scale.
Study finds people resist calling AI conscious despite human-like behavior
A study led by Ophelia Deroy and Louis Longin tested nearly 1,100 people on how they view AI and human awareness. Participants rated responsive AI as less conscious than even the least responsive human. However, both AI and humans received similar ratings for awareness. The researchers say the word choice matters because people are willing to say AI notices things but draw a firm line at calling it conscious.
Self-driving lab uses AI and robots to find heat-resistant coatings
Researchers at the University of Nottingham created a self-driving laboratory that uses AI and robotics to automate material discovery. The system uses two robotic arms, a liquid-dispensing tool, and custom software to handle tasks like mixing chemicals and running high-temperature tests. It aims to speed up the development of advanced coatings for aerospace, energy, and high-temperature engineering. This approach replaces slow, repeated manual experiments with a coordinated automated process.
Sources
- Schneider Electric, NVIDIA, HPE introduce edge AI agents for industrial monitoring, cybersecurity and energy optimization
- Schneider teams with HPE and Nvidia on industrial AI agents at the edge
- LeafSkill: Efficient Leaf-Level Action Grounding for Recursive LLM Agents
- 10 minutes is all it takes for AI assistance to erode your motivation, study shows
- Analysis-In China, the AI bubble has already burst
- Americans Are Officially More Scared of AI Than a Loved One Getting Cancer
- Risks & Rewards: How AI Is Redefining the CISO's Role
- AI Transcription and Translation, Minus the Cloud Connection
- Could long-term AI use lead to isolation? We need to find out — but researchers need better access to data
- Humans draw a line at AI consciousness
- AI-powered robotic laboratory paves the way for discovering next-generation heat-resistant coatings
Comments
Please log in to post a comment.