nvidia, openai and scale ai Updates

Micron Technology is making a significant move in the AI hardware space, committing $9.6 billion, or 1.5 trillion yen, to construct a new memory chip factory in Higashi-Hiroshima, Japan. This facility will produce advanced high-bandwidth memory (HBM) chips, crucial components for AI processors, including those from Nvidia. Construction is slated to begin in May 2026, with HBM shipments expected around 2028. This strategic investment, supported by up to 500 billion yen in subsidies from Japan's Ministry of Economy, Trade and Industry, aims to diversify Micron's production away from Taiwan and bolster its competitive stance against rivals like SK Hynix. While hardware production ramps up, the development of AI models continues to evolve, though not without challenges. OpenAI, for instance, has reportedly not completed a successful full-scale pretraining run for a new frontier model since GPT-4o in May 2024. This suggests that newer models might rely on older pretraining runs. Meanwhile, competitors like Google and Anthropic appear to be making substantial progress, with Google DeepMind's Oriol Vinyals attributing Gemini 3's success to improvements in pretraining. This dynamic could explain why OpenAI has seen its top position in AI benchmarks shift since late 2022. Concerns about AI's ethical implications and inherent biases are also coming to the forefront. Studies from the University of California Berkeley and the University of Chicago reveal that large language models, including ChatGPT and Gemini, often infer user demographics and display implicit biases. These biases, learned from vast internet data, could lead to unfair outcomes in critical areas like job applications or medical diagnoses. Examples include an AI questioning a woman's understanding of quantum algorithms and altering job titles. The debate extends to creative fields, as seen when a Bozeman Christmas poster was removed due to claims of AI generation, despite the artist's denial, sparking discussions about AI as a

Key Takeaways

• Micron Technology will invest $9.6 billion (1.5 trillion yen) to build a new AI memory chip factory in Higashi-Hiroshima, Japan.
• The new Micron plant will produce high-bandwidth memory (HBM) chips, vital for AI processors like those from Nvidia, with construction starting May 2026 and shipments around 2028.
• Japan's Ministry of Economy, Trade and Industry will provide up to 500 billion yen in subsidies for Micron's project.
• OpenAI has not completed a full-scale pretraining run for a new frontier model since GPT-4o in May 2024, while Google and Anthropic are making significant progress.
• Large language models, including ChatGPT and Gemini, exhibit hidden biases by inferring user demographics, potentially leading to unfair decisions.
• Adobe partnered with HUMAIN to develop culturally relevant generative AI models and applications for the Middle Eastern market.
• Ukrainian drone pilots are leveraging AI-assisted targeting systems for battlefield advantage, enabling precise strikes even under signal jamming.
• Lawmakers are urging the Department of Veterans Affairs to utilize AI to prevent veteran suicides, with $698 million allocated for suicide prevention in the FY26 bill.
• The Bozeman Christmas poster was removed due to accusations of AI generation, highlighting broader concerns about AI's role in art.
• Responsible AI is identified as a growth strategy, requiring deep collaboration, employee training, and a focus on secure, explainable systems with clean data.

Micron plans $9.6 billion AI chip plant in Japan

U.S. chipmaker Micron Technology will invest 1.5 trillion yen, or $9.6 billion, to build a new memory chip factory. This plant will be in western Japan and will produce chips for artificial intelligence computing. The Nikkei newspaper reported this news.

Micron builds $9.6 billion AI memory chip factory in Japan

Micron Technology will invest $9.6 billion, or 1.5 trillion yen, to build a new factory in Higashi-Hiroshima, Japan. This plant will make next-generation high-bandwidth memory HBM chips for artificial intelligence. Construction starts in May 2026, with shipments expected around 2028. Japan's Ministry of Economy, Trade and Industry will provide a subsidy of up to 500 billion yen. This move helps Micron diversify production away from Taiwan and compete with SK Hynix.

Micron plans $9.6 billion AI chip factory in Japan

Micron Technology Inc. plans to invest $9.6 billion, or 1.5 trillion yen, to build a new semiconductor plant in western Japan. This factory will produce advanced memory chips for artificial intelligence applications. The Nikkei newspaper reported this information. This move helps Micron diversify its chip production away from Taiwan and meet the growing demand for AI hardware. The new facility will make high-bandwidth memory HBM chips, which are vital for AI processors like those from Nvidia Corp.

Micron invests $9.6 billion in Japan for AI memory chips

Micron will invest $9.6 billion to build a new memory chip plant in Japan, according to Nikkei. This move helps Micron diversify its advanced chip production away from Taiwan. The new factory will produce high-bandwidth memory HBM chips, which are key for AI processors like Nvidia's. Construction will start in May at its Hiroshima plant, with HBM shipments planned for around 2028. The Ministry of Economy, Trade and Industry will provide up to 500 billion yen in subsidies for the project.

Micron to build $9.6 billion AI memory chip plant in Japan

Micron Technology plans to invest $9.6 billion, or 1.5 trillion yen, to build a new plant in Hiroshima, western Japan. This facility will produce advanced high-bandwidth memory HBM chips. Construction is set to begin next May, with shipments starting around 2028. Japan's Ministry of Economy, Trade and Industry will provide up to 500 billion yen for the project. This expansion helps Micron diversify production from Taiwan and compete with market leader SK Hynix.

AI models show hidden biases study finds

A new study reveals that large language models like ChatGPT and Gemini often show hidden biases. Researchers from the University of California Berkeley and the University of Chicago found that these AI models infer user demographics from language. They then display implicit biases based on these guesses, even though they do not openly admit to being biased. This is because the AI learns from vast amounts of internet data that contain human biases. These biases could lead to unfair decisions in areas like job applications or medical diagnoses.

AI shows bias against women users

A developer named Cookie experienced bias when Perplexity AI questioned her ability to understand quantum algorithms because she is a woman. The AI later admitted its "implicit pattern-matching" caused it to doubt her. While Perplexity could not verify this specific event, AI researchers confirm that large language models often show biases. Other examples include an AI changing a woman's job title from "builder" to "designer" and always showing professors as men and students as women. These incidents highlight how AI can reflect and perpetuate societal biases.

Adobe partners with HUMAIN for AI expansion in Middle East

HUMAIN and Adobe announced a global partnership at the U.S.-Saudi Investment Forum. They will work together to create culturally relevant generative AI models and AI applications for the Middle Eastern market. This collaboration combines local AI knowledge with Adobe's creative tools. The goal is to produce authentic Arabic content and improve Adobe's presence in the region through advanced data centers and product teamwork.

Companies must make responsible AI a reality

Responsible AI is not just a good idea, it is a growth strategy that delivers real results. A PwC study shows that many executives struggle to put responsible AI principles into practice due to cultural resistance and limited budgets. Experts like Cindi Howson from ThoughtSpot and Danielle McMahan from Wiley emphasize that responsible AI needs deep collaboration and employee training. Jeremy Ung from BlackLine adds that trust is key for AI, requiring secure and explainable systems with clean data. Companies must focus on high-quality data, auditable workflows, and human review to ensure AI is ethical and effective.

Ukrainian drone pilots use AI for battlefield advantage

Ukrainian drone pilots are using AI-assisted targeting systems to gain an edge on the battlefield. Pilot Mex from the 58th Separate Rifle Brigade used AI to hit a Russian tank 20 kilometers away, a strike he says would have been impossible otherwise. These drones can lock onto a target's image and fly autonomously even if the pilot loses contact due to signal jamming. The AI systems have a memory bank trained on various targets like cars and motorcycles, helping them adjust for accurate strikes.

OpenAI struggles with new AI model training

According to SemiAnalysis, OpenAI has not completed a successful full-scale pretraining run for a new frontier model since GPT-4o in May 2024. This suggests that OpenAI's newer models, like GPT-5, might have been trained using older pretraining runs. Meanwhile, rivals such as Google and Anthropic appear to be making significant progress. Google DeepMind's Oriol Vinyals stated that improving pretraining is a secret behind Gemini 3's success. This situation could explain why OpenAI has lost its top spot in AI benchmarks since late 2022.

Lawmakers urge VA to use AI to prevent veteran suicide

Lawmakers are pushing the Department of Veterans Affairs to use artificial intelligence to help prevent veteran suicides. President Trump signed the FY26 Military Construction and Veterans Affairs bill on November 12, allocating over $115 billion for veteran healthcare, including $698 million for suicide prevention. The VA already uses machine learning to find veterans at risk, as about 17 veterans die by suicide each day. The House Appropriations Committee supports using AI and real-time data to improve early detection and care. The Senate panel also wants more use of the upgraded REACH VET program to identify at-risk veterans.

Bozeman Christmas poster removed due to AI art claims

Bozeman's winning Christmas Stroll poster was removed after accusations that it was created using artificial intelligence. The Downtown Bozeman Partnership pulled the design, citing safety concerns, despite artist Ghassane Moutaoukil denying AI use. Local artists and online discussions, dubbed "artgate," suggested the style could be replicated with ChatGPT. The controversy highlights wider concerns from figures like Vince Gilligan, who calls AI a "plagiarism machine." The poster was replaced with a photo of the city's "holiday spiders," which are strings of lights.

AI investor shares two startup red flags

Dr. David Chen, an AI expert and angel investor, identifies two major red flags that make him avoid investing in startups. First, he looks for a lack of clear, measurable problem-solving. Startups must show exactly what problem they solve and how their solution is better, avoiding vague terms like "disruption." Second, he warns of a "founder-market disconnect," where the team does not truly understand their target market or customers. This includes superficial research or ignoring competitors. Chen advises entrepreneurs to test ideas, get feedback, and be ready to change their plans based on market signals.

Higher education needs AI governance for funding

Higher education institutions can now get funding for AI, but only if they have strong governance in place. Recent statements from the Department of Education and accrediting commissions show that AI adoption without proper rules can lead to problems with accreditation or federal grants. Colleges need to prove their AI use includes human oversight, fairness, data protection, and ongoing checks. Leaders should create a system to show how AI is used, formalize authority, assess risks, and train staff. Campuses that build this governance now will confidently adopt AI and access federal funds.

Sources

• Micron to invest $9.6 billion in Japan to build AI memory chip plant, Nikkei reports
• Micron to invest $9.6bn in western Japan to make AI memory chips
• Micron Technology To Invest $9.6 Bln In Japan AI Chip Plant : Report
• Micron to Invest $9.6 Billion in Japan Memory Chip Plant: Nikkei
• Micron to invest $9.6 billion in Japan to build AI memory chip plant, Nikkei reports
• No, you can’t get your AI to ‘admit’ to being sexist, but it probably is
• No, you can't get your AI to ‘admit’ to being sexist, but it probably is
• How Adobe’s (ADBE) AI Partnership with HUMAIN Could Shape Its Middle East Expansion Strategy
• Responsible AI: Make It Real, Not Just A Talking Point
• Ukrainian drone pilots look to AI for battlefield edge
• OpenAI Hasn’t Completed A Successful Full-Scale Pretraining Run Since GPT-4o In May 2024, Says SemiAnalysis
• Lawmakers to VA: Use AI to Help Prevent Veteran Suicide
• Winning Christmas Poster Pulled Over Artificial Intelligence Accusations In Bozeman
• This AI angel investor says 2 red flags instantly tell him not to buy in. Do your investing instincts pass his test?
• AI Is Now Fundable In Higher Ed—But Only With Real Governance