nvidia, amd and google Updates

Utah has made a significant move in healthcare, becoming the first state to allow an AI system, Doctronic, to make clinical decisions for prescription renewals. This pilot program permits the AI to renew 190 common medications for chronic conditions, with Doctronic reporting a 99.2% match rate with human doctors' decisions. The system incorporates nearly a dozen safety checks and "guardrails," including initial human reviews for the first 250 renewals per drug class, aiming to reduce patient wait times and healthcare costs, especially in rural areas. While Doctronic plans to expand to states like Texas, Arizona, and Missouri by 2026, public advocates and groups like the American Medical Association express concerns about potential errors and the AI undermining the essential role of human clinicians. The Utah Department of Commerce and the state's Office of Artificial Intelligence Policy will closely monitor the program's safety and effectiveness. Beyond healthcare, AI advancements and discussions are unfolding across various sectors. NVIDIA recently launched Nemotron Speech ASR, an open-source transcription model designed for low-latency voice AI applications such as live captioning and voice agents. This 600-million-parameter model achieves a Word Error Rate of about 7.84% at a 0.16-second chunk size and can support approximately 560 concurrent streams on an NVIDIA H100 GPU. Meanwhile, AMD is significantly boosting its AI talent and hardware development, as announced by CEO Lisa Su at CES 2026. The company is expanding its global team for AI workflows and introduced new AI-powered products, including the MI440X processor, Instinct MI500 GPU series, and the Helios rack system for data centers, alongside consumer AI PCs with Ryzen AI 400 series chips. Motorola also showcased a proof-of-concept wearable AI device at CES 2026, a minimalist pendant powered by its new AI assistant, Qira, capable of identifying objects and summarizing information. In the business realm, Humantic AI launched Agent Miia on January 7, 2026, an advanced tool that drastically cuts B2B sales research time from three hours to just eight minutes by autonomously gathering deep insights. OpenArt AI continues to evolve its art tool, offering image generation, editing, and new video animation features, integrating over 50 models like DALL·E-3 and Stable Diffusion. However, the rapid integration of AI also brings challenges and scrutiny. Gmail users received warnings from a tech expert, Davey Jones, about Google's "Smart Features" privacy settings, claiming AI accesses private messages for training. Google, however, clarified that these reports are misleading, stating that Smart Features have existed for years and they do not use Gmail content to train their Gemini AI model. Furthermore, an AI program used by Heber City Police in Utah mistakenly reported an officer turned into a frog due to background audio from a Disney film, highlighting the need for careful oversight. The broader AI industry is also rethinking its long-held goal of Artificial General Intelligence (AGI), with leaders like OpenAI CEO Sam Altman and Salesforce CEO Marc Benioff expressing new perspectives, as critics suggest current Large Language Models are not capable of reaching AGI.

Key Takeaways

  • Utah is the first state to approve an AI system, Doctronic, for renewing 190 common chronic medications, with the AI matching human doctor decisions 99.2% of the time.
  • NVIDIA launched Nemotron Speech ASR, an open-source 600-million-parameter transcription model for low-latency voice AI, achieving a 7.84% Word Error Rate.
  • AMD is expanding its global team for AI development and introduced new AI-powered products at CES 2026, including the MI440X processor, Instinct MI500 GPU series, and Ryzen AI 400 series chips.
  • Humantic AI released Agent Miia on January 7, 2026, an AI tool that reduces B2B sales research time from three hours to eight minutes.
  • Motorola showcased a proof-of-concept wearable AI device at CES 2026, a pendant powered by its new AI assistant, Qira, for object identification and information summarization.
  • Google faced warnings regarding Gmail "Smart Features" privacy settings, with claims of AI accessing private messages for training, though Google denies using Gmail content for Gemini AI training.
  • The AI industry, including leaders like OpenAI CEO Sam Altman and Salesforce CEO Marc Benioff, is re-evaluating the goal of achieving Artificial General Intelligence (AGI).
  • Professors are developing methods, such as "Trojan horse" prompts, to detect AI-driven cheating in student assignments.
  • An AI program used by Heber City Police in Utah mistakenly reported an officer turned into a frog due to background audio, highlighting potential errors in AI applications.
  • OpenArt AI offers image generation, editing, and video animation features, integrating over 50 models like DALL·E-3 and Stable Diffusion, using a credit-based pricing system.

Utah approves AI for patient prescriptions

Utah has become the first state to allow an AI system, Doctronic, to make clinical decisions for prescription renewals. Dr. Adam Oskowitz, Doctronic's co-founder, called this a groundbreaking event. The system includes nearly a dozen safety checks for dosage and drug interactions. Doctronic is also talking with other states, expecting more approvals by 2026. This aims to help with the high demand in healthcare.

Utah first state to allow AI medication renewals

Utah launched a new program allowing an AI system from Doctronic to renew 190 common medications for chronic conditions. This makes Utah the first state to let AI participate in medical decision-making for prescriptions. The system matches human doctors' decisions 99.2% of the time and faces the same malpractice rules. Doctronic hopes to expand this program to other states like Texas, Arizona, and Missouri. The goal is to improve healthcare access, lower costs, and help healthcare workers.

Utah allows AI to approve prescription refills

Utah is the first state to let patients renew medical prescriptions using artificial intelligence through a partnership with Doctronic. The Utah Department of Commerce will check the program's safety and effectiveness. Doctronic co-founder Dr. Adam Oskowitz calls this groundbreaking, aiming to cut patient wait times and healthcare costs. While some doctors worry about risks, Oskowitz states the AI has "guardrails" including human reviews. Doctronic is discussing expansion with Texas, Arizona, and Missouri, expecting more states to join by 2026.

Utah AI autonomously refills prescriptions

Utah launched a pilot program allowing an AI system from Doctronic to autonomously refill medication prescriptions. Public advocates call this "dangerous," but the state's Department of Commerce sees it as balancing innovation and safety. Initially, human doctors will review the first 250 renewals for each drug class. Doctronic co-founder Adam Oskowitz says the AI prioritizes safety and refers uncertain cases to doctors. However, critics like Robert Steinbrook from Public Citizen worry about the AI undermining the essential role of human clinicians.

Utah approves first AI drug prescriptions

Utah has become the first state to allow an AI system from Doctronic to prescribe drug refills. This pilot program aims to save patients time and money, especially in rural areas. The AI can refill 190 common medications for chronic conditions, but a human doctor must issue the initial prescription. Doctronic's AI matched human doctors' decisions over 99% of the time. Critics, including the American Medical Association, warn of potential errors and misuse without direct physician input. The state's Office of Artificial Intelligence Policy will monitor the program to inform future AI healthcare policies.

Teachers use AI to detect student cheating

Barbara S. Held discusses how professors are using clever methods to detect AI-driven cheating in student work. History professor Will Teague uses "Trojan horse" tricks with hidden words in assignments. Another professor friend designed prompts that lead AI to generate irrelevant information, like discussing psychoanalysis for a different topic. This allows professors to fail students for not answering the assignment correctly, rather than directly accusing them of using AI. Held believes teachers should develop these skills to uphold education and help students learn from their mistakes.

Teachers use AI to detect student cheating

Barbara S. Held discusses how professors are using clever methods to detect AI-driven cheating in student work. History professor Will Teague uses "Trojan horse" tricks with hidden words in assignments. Another professor friend designed prompts that lead AI to generate irrelevant information, like discussing psychoanalysis for a different topic. This allows professors to fail students for not answering the assignment correctly, rather than directly accusing them of using AI. Held believes teachers should develop these skills to uphold education and help students learn from their mistakes.

Gmail users warned about AI privacy settings

A tech expert, engineering YouTuber Davey Jones, warns Gmail users to turn off certain "Smart Features" due to privacy concerns. He claims Google automatically opts users in to allow AI to access private messages and attachments for training. Users must manually disable these settings in two places on desktop and mobile. Turning off these features also disables useful tools like "smart compose" and email filters. Google states these reports are misleading, clarifying that Gmail Smart Features have existed for years and they do not use Gmail content to train their Gemini AI model.

NVIDIA launches Nemotron Speech ASR for fast voice AI

NVIDIA AI released Nemotron Speech ASR, a new open-source transcription model for low-latency uses like voice agents and live captioning. This 600-million-parameter model uses a cache-aware FastConformer encoder and an RNNT decoder. It processes audio in non-overlapping chunks, ensuring stable latency and predictable memory use even under heavy load. The model achieves a Word Error Rate of about 7.84% at a 0.16-second chunk size. On an NVIDIA H100 GPU, it supports about 560 concurrent streams, which is three times more than a standard streaming system.

Motorola unveils new wearable AI device at CES

Motorola is entering the wearable AI market, showcasing a new proof-of-concept device at CES 2026. This minimalist gadget, resembling a pendant, is powered by Motorola's new AI assistant, Qira. It can identify objects, summarize information, and help navigate by launching phone apps. The device features a camera, sensors, speaker, and microphones, but no visible buttons or displays. Motorola emphasized that this wearable is still far from being ready for commercial release.

Utah police AI mistakes officer for frog

An AI program used by Heber City Police in Utah mistakenly reported that an officer turned into a frog. This error happened because the officer's body camera picked up dialogue from the Disney film "The Princess and the Frog" in the background. Sergeant Rick Keel highlighted the need to correct these AI-generated reports. While AI tools offer efficiency for police departments, civil liberties experts warn that their rapid spread could lead to errors and reinforce biases in the criminal justice system.

OpenArt AI review features and pricing

OpenArt AI is an art tool offering image generation, editing, and new video animation features. It allows users to create and edit images directly in their browser with tools like inpainting and background removal. The platform integrates over 50 models, including DALL·E-3 and Stable Diffusion, for both images and videos. OpenArt uses a credit-based pricing system with a generous free plan for basic features. However, its pricing structure and model availability across different plans can be unclear.

Humantic AI Agent Miia speeds up sales research

On January 7, 2026, Humantic AI launched Agent Miia, an advanced AI tool designed to cut enterprise account research time for B2B sellers from three hours to just eight minutes. Agent Miia autonomously gathers and organizes deep insights on company strategy, market trends, and buying teams. It uses private data and a special reasoning engine to provide tailored intelligence for complex sales deals. This new tool expands Humantic AI's Complete Buyer Intelligence platform, which also includes Agent Pi for understanding buyer personalities.

AMD boosts AI talent and hardware development

At CES 2026, AMD CEO Lisa Su announced that artificial intelligence is reshaping the company's hiring and hardware plans. AMD is expanding its global team, focusing on roles that support AI workflows and product development. The company also introduced new AI-powered products, including the MI440X processor, Instinct MI500 GPU series, and the Helios rack system for data centers. AMD is also releasing consumer AI PCs with Ryzen AI 400 series chips this month, with a more powerful Ryzen AI Halo desktop coming in the second quarter of 2026.

AI industry rethinks AGI goals for 2026

The AI industry is starting to back away from its long-held goal of achieving Artificial General Intelligence, or AGI. AGI is a hypothetical AI that could match or surpass human thinking abilities. This shift began last year, with leaders like OpenAI CEO Sam Altman and Salesforce CEO Marc Benioff expressing new perspectives. Critics suggest that current Large Language Models, which many companies invest heavily in, are simply not capable of reaching AGI.

Sources

NOTE:

This news brief was generated using AI technology (including, but not limited to, Google Gemini API, Llama, Grok, and Mistral) from aggregated news articles, with minimal to no human editing/review. It is provided for informational purposes only and may contain inaccuracies or biases. This is not financial, investment, or professional advice. If you have any questions or concerns, please verify all information with the linked original articles in the Sources section below.

Utah AI in Healthcare Prescription Renewal Doctronic Clinical Decisions Medication Management Healthcare Access AI Regulation Patient Safety Healthcare Costs Pilot Programs Chronic Conditions Medical Malpractice AI Ethics AI in Education Cheating Detection Academic Integrity Educational Technology AI Privacy Gmail Google AI Smart Features User Data NVIDIA Speech Recognition Voice AI Open Source AI AI Models GPU Low Latency AI Motorola Wearable AI CES AI Assistants Qira Consumer AI AI Errors Police AI Criminal Justice AI Bias Civil Liberties Public Safety AI OpenArt AI AI Art Image Generation Video Animation AI Editing DALL-E 3 Stable Diffusion Creative AI Sales AI B2B Sales AI Agents Business Intelligence Humantic AI Market Research Enterprise AI AMD AI Hardware AI Processors Data Centers AI PCs AGI AI Industry Large Language Models OpenAI AI Development Future of AI

Comments

Loading...