Listen to today’s podcast: https://www.youtube.com/channel/UC-nqwUyvLDEvs7bV985k-gQ
AI Daily Podcast
Today's podcast episode was created from the following stories:
Google is making it easier to use the Gemini API in multi-agent workflows
Google added structured outputs with full JSON Schema support to the Gemini API, making it simpler to build reliable, multi-agent workflows and integrate with validation tools like Pydantic and Zod. The update, available across supported Gemini models (including 2.5), improves consistency between agents and LLMs and even preserves schema key order. For developers, this means fewer translation layers and more trustworthy data extraction, moderation, and database tasks.
Google Finance with Deep Search AI lets users watch the market like a hawk
Google Finance is rolling out Gemini-powered Deep Search in the U.S., letting users ask complex market questions and see a transparent research plan with cited sources. It also introduces market probability features using data from Kalshi and Polymarket, with higher limits for AI Pro and Ultra subscribers. The revamped Finance experience expands to India next, signaling a bigger push to make AI-driven financial research mainstream and verifiable.
NotebookLM is becoming a better Android study tool with flashcards and quizzes
NotebookLM’s mobile apps now generate customizable flashcards and quizzes from your sources, bringing key web features to iOS and Android. Google says chatbot quality improves by 50%, with a 4x larger context window and 6x longer conversation memory thanks to the latest Gemini models. Students and researchers can tailor difficulty, topics, and sources—turning documents and links into focused study sessions on the go.
Microsoft’s vision for humanist superintelligence: a future for humanity, not AI domination
Microsoft AI CEO Mustafa Suleyman outlined a push toward “Humanist Superintelligence,” aiming for systems that surpass human cognition while remaining controllable and explicitly serving humanity. A new OpenAI agreement extends Microsoft’s model and product IP rights through 2032 and allows independent AGI pursuits, with claims of AGI subject to verification by an external panel. The emphasis is on perpetual containment and alignment—a call for industry-wide collaboration to manage risks as capabilities accelerate.
One of legal's hottest AI startups says it's cracked a way for law firms to get richer
Stockholm-based Legora, fresh off a $150 million round, is challenging Harvey with a client-facing Portal that lets firms monetize their playbooks via secure, branded workspaces. The pitch: centralize collaboration, keep data out of model training by default, and ship agentic workflows clients can run—potentially boosting margins as corporate legal teams automate more in-house. With design partners across Big Law and the Big Four, general availability is slated for early 2026, though firms will scrutinize conflict handling and data governance.
4 AI-powered consulting startups to watch into 2026
Investors are betting on a new wave of consulting-tech: PromptQL builds enterprise “AI analysts” to surface insights without heavy data prep; Aily Labs embeds a decision-intelligence layer for faster, unified executive decisions. Profound leads the nascent “generative engine optimization” market to influence how brands surface in AI answers, while Dialogue compresses end-to-end market research from weeks to a day. Together, they hint at consulting models that are always-on, data-native, and increasingly automated.
Googleマップがしゃべる相棒に。運転中もAIが質問に答えてくれる
Google Maps is integrating Gemini to act as a conversational copilot while driving—answering local queries, sharing ETAs, finding EV chargers, and even adding calendar events hands-free. Landmark-based directions (e.g., “turn after the restaurant”) and quick Lens queries bring more intuitive, visual navigation. The rollout hits iOS and Android in the coming weeks, with Android Auto to follow, aiming to make on-the-road assistance smarter and safer.
