{"author":{"name":"Will Douglas Heaven","slug":"will-douglas-heaven","article_count":2,"latest_published_at":"2026-05-01T13:47:23.69+00:00","profile_url":"https://vff.ai/authors/will-douglas-heaven","api_url":"https://vff.ai/api/authors/will-douglas-heaven"},"articles":[{"slug":"this-startup-s-new-mechanistic-interpretability-tool-lets-you-debug-llms","title":"Goodfire's Silico Brings Mechanistic Interpretability to Model Development","url":"https://vff.ai/article/2026/05/01/this-startup-s-new-mechanistic-interpretability-tool-lets-you-debug-llms","content_type":"aggregated_news","summary":"Goodfire, a San Francisco startup, released Silico, a tool that lets developers inspect and adjust AI model parameters during training by mapping neurons and their connections. The tool automates mechanistic interpretability work previously done manually, aiming to make model development more precise and less trial-and-error. Silico works on open-source models where developers have access to internal parameters, though not on proprietary systems like ChatGPT or Gemini. The company claims this represents a shift from scaling-focused approaches toward understanding and controlling how models actually work.","published_at":"2026-05-01T13:47:23.69+00:00","updated_at":"2026-05-01T13:47:23.957103+00:00","source":{"url":"https://www.technologyreview.com/2026/04/30/1136721/this-startups-new-mechanistic-interpretability-tool-lets-you-debug-llms/","name":"MIT Technology Review"},"featured_image":{"url":"https://wp.technologyreview.com/wp-content/uploads/2026/04/maintenance-ai.jpg","alt":null},"categories":[{"name":"Research","slug":"research"},{"name":"AI Safety & Alignment","slug":"ai-safety-alignment"},{"name":"LLMs","slug":"llms"},{"name":"Generative AI","slug":"generative-ai"},{"name":"Open Source","slug":"open-source"},{"name":"Coding / Dev Tools","slug":"coding-dev-tools"}]},{"slug":"why-opinion-on-ai-is-so-divided","title":"The AI Perception Gap: Why Experts and the Public See Different Technologies","url":"https://vff.ai/article/2026/04/14/why-opinion-on-ai-is-so-divided","content_type":"aggregated_news","summary":"Stanford's 2026 AI Index reveals a stark divide in how experts and the general public perceive AI's impact, with 73% of US AI researchers optimistic about job effects versus only 23% of the public. The report documents major inconsistencies in AI capabilities, from models that win math olympiads but cannot read analog clocks, to a hardware supply chain concentrated in a single Taiwanese foundry. The gap appears rooted in divergent user experiences: technical professionals using AI for coding see transformative tools, while broader populations encounter more mixed results, creating fundamentally different assessments of the technology's trajectory.","published_at":"2026-04-14T11:54:44.45+00:00","updated_at":"2026-04-22T00:59:04.768177+00:00","source":{"url":"https://www.technologyreview.com/2026/04/13/1135720/why-opinion-on-ai-is-so-divided/","name":"MIT Technology Review"},"featured_image":{"url":"https://insidetelecom.com/wp-content/uploads/2024/10/conflict-resolution-in-AI-1.jpg","alt":null},"categories":[]}]}