This week we feature articles from Jack Lindsey, et al, and Eliza Strickland.
Additional reading comes from Omer Goldman & Uri Shaham, Steven Sinofsky, Volodymyr Pavlyshyn, and Adam Bignell.
News comes from MindsDB, Access Innovations, Perplexity & firmly.ai, and Contentful.
Our next issue arrives April 23.
All previous issues are available at https://gilbane.com/gilbane-advisor-index
Opinion / Analysis
On the biology of a large language model
Jack Lindsey and his colleagues at Anthropic…
“The black-box nature of models is increasingly unsatisfactory as they advance in intelligence and are deployed in a growing number of applications. Our goal is to reverse engineer how these models work on the inside, so we may better understand them and assess their fitness for purpose.”
Anthropic’s new research into how LLMs “think” is fascinating. Don’t be put off by the length of their paper — you can learn a lot by reading some of the introductory material, the discussion at the end, and then surfing the case studies and topics in-between. (1 hr 49 min)
https://transformer-circuits.pub/2025/attribution-graphs/biology.html
12 graphs that explain the state of AI in 2025
The always useful and free AI Index from Stanford University’s Institute for Human-Centered Artificial Intelligence (HAI) is available. The 455 page 2025 issue is “stuffed with graphs and data on the topics of R&D, technical performance, responsible AI, economic impacts, science and medicine, policy, education, and public opinion.” IEEE Spectrum’s Eliza Strickland provides highlights and links to the full report. (7 min)
https://spectrum.ieee.org/ai-index-2025
More Reading
- ECLeKTic: A novel benchmark for evaluating cross-lingual knowledge transfer in LLMs via Google Research Blog
- MCP – It’s hot, but will it win? Looking at middleware history. via Hardcore Software
- Metagraphs as homoiconic structures: Revolutionizing knowledge representation via Artificial Intelligence in Plain English
- New in NotebookLM: Discover sources from around the web via Google Labs blog
Content technology news
MindsDB brings federated data access to Model Context Protocol
MindsDB & MCP enables AI applications and agents to run federated queries over data stored in different databases and applications as if they were a single database.
https://mindsdb.com
Access Innovations releases Data Harmony 3.17
The new version includes knowledge maps that illustrate the linkages (with color-coded legends) between categories and terms in a taxonomy.
https://www.accessinn.com/data-harmony-products/
Perplexity integrates firmly.ai ecommerce technology
By embedding firmly.ai’s commerce platform, Perplexity users can now complete their entire shopping journey, from product discovery to secure transaction, without leaving the platform.
https://www.firmly.ai ■ https://www.perplexity.ai
Contentful launches new CMS capabilities and Shopify partnership
New AI and personalization capabilities for technical teams build on the company’s CMS, and Contentful Studio, and a low-code product for marketing teams.
https://www.contentful.com/newsroom/contentful-reveals-next-phase-of-growth-with-modern-digital-experience
The Gilbane Advisor is authored by Frank Gilbane and is ad-free, cost-free, and curated for content, computing, data, web, and digital experience technology and information professionals. We publish recommended articles and content technology news most Wednesdays. We do not sell or share personal data.
Subscribe | View online | Editorial policy | Privacy policy | Contact