mrlloyd.net's Tech Brief

Artificial Intelligence

Models, research and the companies building them.

Apple reportedly building server packed with M-series Ultra chips for AI

Apple is reportedly developing a server that uses M-series Ultra chips for AI workloads, with a planned 2029 release. The server would be Apple's first in nearly two decades and come in two configurations with two or four M8 Ultra chips. The project has received support from new CEO John Ternus.

Why it matters: This development could provide AI developers with a new option for running AI workloads on Apple hardware.

Our framework for reporting model misalignment

OpenAI has released a framework for identifying and addressing model misalignment in AI systems. The framework includes guidelines for tracking, investigating, and disclosing instances of unexpected or concerning model behavior. Six reports of model misalignment have been published alongside the framework.

Why it matters: This framework aims to increase transparency and accountability in AI development, helping to mitigate potential risks associated with model misalignment.

Helping older adults use AI in everyday life

OpenAI and AARP are partnering to offer free ChatGPT workshops for 1,000 older adults in 10 US cities. The workshops aim to teach practical AI skills in a hands-on environment. Participants will learn how to use ChatGPT safely.

Why it matters: This initiative aims to increase digital literacy and accessibility for older adults, helping them to better navigate the digital landscape.

Reimagining advertising with AI

OpenAI has introduced AI-powered advertising experiences, including Sponsored Agents for marketers and integrations with HubSpot and Shopify.

Why it matters: This development may interest marketers looking to leverage AI in their advertising strategies.

Building the materials foundation for AI

Advanced materials are becoming crucial for AI performance, efficiency, and sustainability as semiconductors and data centers approach physical limits. Materials scientists are using AI to accelerate the development of new materials. Syensqo is developing materials for data centers, semiconductor manufacturing, and thermal management.

Why it matters: Understanding the role of advanced materials in AI development can help businesses and researchers prioritize their own material research and development efforts.

Hex turns complex analysis into visual reports with GPT‑6 Astra

Hex has integrated GPT-6 Astra into its platform to enable data agents to create interactive visualizations from complex analysis. This integration allows users to transform complex data into shareable visual reports. The integration aims to make data analysis more accessible and engaging.

Why it matters: This integration can help organizations make complex data more understandable and shareable among employees.

How to connect AI usage to business value

OpenAI's ChatGPT Work and Codex analytics provide tools to track AI usage and spending, identify training needs, and measure business outcomes. These tools help teams understand how AI is being used and its impact on the organization. The analytics aim to connect AI adoption to business value.

Why it matters: This can help businesses make data-driven decisions about AI investments and optimize their AI strategies.

How workers are unlocking new ways of working

OpenAI Economic Research published a study on how workers use AI beyond traditional roles, revealing new activities that have become part of their work. The research highlights various ways workers are leveraging AI to enhance their productivity and efficiency. The study provides insights into the impact of AI on modern work.

Why it matters: The study's findings can help businesses and policymakers understand how to effectively integrate AI into the workforce and support workers in adapting to new technologies.