This Week in AI

This week, Matt Palmer, head of developer experience at Conductor, joined host and Intelligence Briefing founder Andreas Welsch to work through the week's biggest stories: what the export restrictions on Anthropic's Fable 5 and Mythos Preview mean for architecture decisions, why AI agents are making developers more exhausted rather than less, and what Sakana AI's new Fugu system offers as an alternative to single-vendor dependency.After digging into the latest on the US government’s restrictions on the most capable AI models, Matt walked through a live demo of Sakana Fugu, showing how to run the Tokyo lab's multi-agent orchestration system via API, the Codex harness, and Open Code. Along the way, Andreas and Matt also covered Qualcomm's $3.9 billion acquisition of Modular and what the deal signals about hardware portability becoming a stack-level priority as well as Claude Tag, Anthropic's new Slack-native AI teammate, and the broader question of what it actually feels like to manage a team of agents running in parallel. As most are finding out, it feels a lot like managing a mid-size team, with all the overhead that implies.

In this episode of This Week in AI, host Ksenia Se, founder of Turing Post, took us through three stories that may look unrelated but all point to the same shift: AI is moving out of conversation and into the operational infrastructure where real work happens.Ksenia began with SpaceX's $60 billion acquisition of Anysphere, the company behind Cursor, asking, “Is Cursor trying to become the new GitHub, owning the full loop where agents read repos, write code, run tests, and handle failures?” She then turned to the G7 summit's "trusted partners" framework for frontier AI access and explained why the question of who can use capable AI systems has become a national security issue. Ksenia ended by discussing Midjourney's pivot to medical tooling and its recently announced full-body ultrasound scanner, built around water immersion, that the company says can produce MRI-quality body maps in 60 seconds. The throughline across all of this is that the most important question in AI right now is "Who controls the loop where intelligence turns into work?"

This week John Lindquist, cofounder of egghead.io, joined host and CS Dojo founder YK Sugi to break down the week's biggest AI news and make the case for a smarter way to build with agents. The pair covered Claude Fable 5's brief but impressive run and the government-ordered shutdown that followed as well as Uber burning its entire 2026 AI budget by April, mostly on Claude Code and Cursor. Then John laid out his "Clone Wave" framework: Rather than prompting agents to build from scratch, use the GitHub CLI to find existing battle-tested open source code and feed it to your agents as ingredients. As John pointed out, "Ingredients beat inference." He also walked through how Deep Wiki lets agents explore repos without cloning them, how cmux enables autonomous multi-agent workspaces, and why every tool you build should expose endpoints and CLIs your agents can both control and debug. Watch now.

Recommendation systems quietly drive some of the most consequential numbers in tech—35% of Amazon's revenue, 75% of what Netflix surfaces, the entire logic of TikTok's feed. But as ex-Microsoft engineer and RecoMind founder Miguel Fierro explained to host Christina Stathopoulos on this week’s episode, most companies are nowhere near the state of the art, and the gap is widening. Miguel broke down the four trends separating leaders from laggards: sequential modeling that treats user behavior like next-token prediction, the convergence of search and retrieval into a single personalized system, the emergence of foundation models for recommendations (Netflix is the only shop known to have one), and the difference between a real sales agent and the conversational agents most companies employ today. As always Christina opened with a rapid-fire news round, covering Anthropic's valuation surge and quiet S-1 filing, recent pleas for responsible AI, Google I/O's multimodality push, and why enterprises are abandoning token leaderboards in favor of what some are calling valuemaxxing.

This week, host Andreas Welsch brought together Maya Mikhailov, cofounder and CEO of Savvi AI, and Doug Shannon, generative AI and intelligent automation leader, to cover four developments shaping how organizations build with and buy into AI: OpenAI’s push into personal finance, the role of metacognition in AI-assisted technical work, the growing backlash against token-based productivity metrics, and the new role of forward-deployed engineer.Maya reframed OpenAI's move into personal finance as an intent-harvesting play, explaining how transaction data combined with chat history gives AI companies a portrait of consumers that banks, advertisers, and anyone selling attention will pay dearly for. Doug made the case for metacognition as a professional skill: AI systems are designed to find the mean and that the human's job is to know when the mean is good enough and when it isn't. The panel then examined the limits of tokenmaxxing and discussed why the shift to usage-based pricing will force a reckoning that internal policy never quite managed. All this, plus a warning about intellectual surrender and IP, the problems with the forward-deployed engineer model, and why organizational knowledge is the key to successfully deploying AI. Watch now.

This week, host Eric Freeman and John Berryman, founder of Arcturus Labs, coauthor of Prompt Engineering for LLMs and an early production engineer on GitHub Copilot, cover the week's biggest AI developments: Anthropic's decision to restrict its Mythos model after it identified critical security flaws, the White House's possible pivot to FDA-style AI review, and the staggering compute deals reshaping the industry, including a 40,000-acre Utah data center planned for nine gigawatts of power. Berryman then takes you through four years of AI product development, from tiny 2,048-token context windows to today's agent harnesses, and shows why the gap between a bare model and a well-designed harness now drives more performance than any model benchmark. He also demos a personal agent that carries context from an Obsidian notebook into Wikipedia, giving a glimpse of how a future open agent protocol might work, and explains how he helped a client replace an entire bespoke application with a skills-driven agent that domain experts can read and fix themselves, in plain English, no developer required.If you build with AI or make decisions about AI tooling, this episode covers the infrastructure, policy, and architectural shifts you need to understand right now.