Podcast anything.

Any source. One RSS feed.

Blogs

X accounts

Hacker News

Podcasts

Any website

The Daily FMSummarizes & narrates

One RSS feed

Subscribe to ready-made pods

Grab any pod from the library with its RSS feed (no account needed!)

Combine pods into one feed

Merge many podcasts into your own custom pod to get them all in a single feed

Customize your own pod

Create a custom summary pod with your own sources

Sources can be the things you already follow:

Websites: any article, blog, or newsletter link
X: the latest posts from your favorite @handles
Hacker News: top stories and their comment threads
Podcasts: fresh-episode recaps of any show
State legislatures: weekly bill activity

Connect your AI agent

Connect it over MCP and it can build and run pods for you

Featured Pods

AI Daily

anthropicopenaideepmindblognewssimonwillison@karpathy@sama@AnthropicAI@swyx@levelsio@yoheinakajima@dair_ai@rowancheung@rasbt@bcherny@steipete

Combine these episodes into your pod

All episodes from this podcast will be fed into your own.

AI Daily October 6: OpenAI Eyes Ads as Anthropic Moves Claude Cowork to CloudHere is today's AI Daily for Tuesday October 6th. Yesterday, OpenAI published a new product post titled “Building advertising for the way people use AI.” The source does not yet provide operational details, but the direction is significant: advertising is likely becoming a more explicit part of the business model for consumer AI products. The key question will be whether ads can be useful, clearly labeled, and separated from an assistant’s recommendations. As AI systems become more personal and more embedded in everyday workflows, preserving user trust will matter as much as monetization. Also yesterday, Anthropic shifted Claude Cowork’s computing environment into the cloud. Previously, Cowork used an Anthropic-provided virtual machine running on the user’s computer for tool use. The new design runs both model inference and a separate sandboxed virtual machine in the cloud, while the desktop app handles requests involving local files. Anthropic says this should reduce battery and disk costs, keep tasks running when a laptop is closed, and enable use from a phone. The trade-off is familiar: cloud execution improves convenience and persistence, but makes clear permissions and data boundaries even more important. Cloudflare yesterday wrapped up its Birthday Week with 46 announcements spanning AI agents, developer infrastructure, open source, and post-quantum security. The AI through-line is that agents are becoming an infrastructure problem, not just a model problem: they need sandboxes, web search, routing, access controls, usage measurement, and payment systems. That is a crowded but consequential layer of the market. Finally, fresh research highlighted today focuses on long-running agents’ weak spot: context compression. A paper described by DAIR.AI finds that compression can damage agent performance at particular points in a task. Its proposed PAIR method compares an agent replayed from the same state with and without a compression step, helping identify when summaries have discarded something essential. The broader lesson is that persistent AI agents will succeed or fail on operational details—memory, compute location, permissions, cost controls, and trust—not just raw model capability. Thank you for listening to AI Daily from The Daily FM. See you tomorrow!

Financial Markets

bloombergcnbcfinanceft@WSJmarkets@markets@business@CNBCnow@ReutersBiz

Combine these episodes into your pod

All episodes from this podcast will be fed into your own.

Financial Markets October 6: Oil Below $100 Lifts Stocks as Gulf Flows RecoverHere is today's Financial Markets for Tuesday October 6th. U.S. stock futures are nudging higher as oil falls back below $100 a barrel, offering some relief after an extended period of energy-driven inflation concerns. Yesterday, U.S. stocks rose and the Nasdaq reached a record high, with optimism around corporate earnings and AI investment outweighing anxiety over elevated Treasury yields. The early message today is that lower oil is helping risk appetite, but markets remain sensitive to whether the decline reflects durable supply recovery or merely a pause in a volatile energy market. Oil flows from the Gulf reached more than 80% of pre-war levels in September, Reuters reports, easing pressure on global crude prices. Higher freight rates have encouraged more shipowners to return to the region. Still, the disruption is not over: diesel shortages are worsening, and shipping risks around the Strait of Hormuz remain. China’s independent refiners are also shifting toward Iraqi crude after U.S. restrictions constrained Iranian supply. Energy markets may be less stressed than they were, but regional supply routes and refined-product availability remain major inflation risks. Bond markets are finding some support. U.S. yields have fallen from their highest levels since 2002 as oil retreated, while Treasury Secretary Scott Bessent said the government’s debt burden can be brought under control. But long-end yields remain elevated ahead of a three-year Treasury auction. Investors will be watching demand closely, because weak auctions could quickly renew pressure on borrowing costs and equity valuations. In Europe, energy concerns are broadening beyond oil. European gas prices are rising as low storage levels leave the market vulnerable, and the European Union plans to seek an operator for joint energy purchases. Meanwhile, Bank of England policymaker Catherine Mann says high inflation has become embedded in Britain, while ECB officials argue high yields may limit the broader inflation pass-through from energy. The broader trend is a fragile improvement in market sentiment: cheaper crude supports stocks and bonds, but high yields, diesel scarcity, and Europe’s renewed energy concerns mean volatility has not gone away. Thank you for listening to Financial Markets from The Daily FM. See you tomorrow!

Latent Space in 3 minutes

feed:api.substack.com

Combine these episodes into your pod

All episodes from this podcast will be fed into your own.

Latent Space in 3 minutes: Academia is for Ambition — Alex Zhang, MITHere is The Daily FM summary of the Latent Space that aired on Thursday October 1st. MIT PhD student Alex Zhang joined Swyx and Vibhu for a wide-ranging conversation about a central idea in modern AI: today’s frontier models may be far more capable than their interfaces, harnesses, and agent systems allow us to see. Zhang began with GPU Mode and KernelBench, communities and benchmarks built around writing high-performance GPU kernels. While AI now generates many leaderboard-winning kernels, he argued that human expertise still matters enormously. The hard part is not only proposing fast code, but verifying that it is stable and reliable in real systems rather than gaming a benchmark. A knowledgeable engineer can steer a model toward the right insight, potentially replacing billions of tokens’ worth of blind search. The same lesson, Zhang said, applies to math, coding, and agents: experts are not obsolete; they become powerful verifiers and directors. That theme carried into his advice for researchers. Academia’s advantage, he argued, is not competing directly with frontier labs on compute. It is taking weird, ambitious bets that industry may initially dismiss as trivial or useless. He cited ideas such as SWE-bench, ReAct, Quiet-STaR, and his own Recursive Language Models, or RLMs, which were initially met with skepticism before shaping how the field thinks about agents. RLMs are not a new base-model architecture. They are a harness design: the model stores context outside its prompt, in a filesystem or code environment, and uses code to call tools and recursively spawn subagents. Instead of cramming an entire long task into one ever-growing conversation, the system breaks work into smaller, locally manageable problems. Zhang’s claim is that this produces compositional generalization: a strategy learned on short tasks can extend to much longer tasks, or even transfer across apparently unrelated domains, because the high-level program stays similar. He thinks many popular coding-agent harnesses—Claude Code, Codex, and Pi—are structurally more alike than they appear. Their differences may matter less than finding genuinely new harness abstractions. Prime Agent, a project built with Prime Intellect, pushes the RLM approach by making IPython its central tool, externalizing memory to disk, allowing agent-to-agent communication, and supporting persistent subagents that survive beyond a single interaction. The conversation then turned to swarms. Zhang believes the future “language model” users query may secretly be a large scaffold or swarm of agents, while the user sees only a clean interface. OpenAI’s reported experiment—10,000 agents running for 88 hours, producing 130 billion output tokens and solving difficult mathematical problems—was impressive precisely because it showed that enough search plus shared context can uncover solutions. But Zhang cautioned that most swarm activity may be wasted exploration. The unsolved challenge is getting agents to converge efficiently rather than simply burning tokens. A striking related point involved newer fast models such as GEV, which challenge the assumption that every language model must be an autoregressive text generator. If a task only needs a quick binary classification or calibrated decision, generating hundreds of tokens is wasteful. Zhang expects more models optimized for distinct output spaces, latency requirements, and roles inside larger agent systems. His final thesis was “capability overhang.” Frontier models may already be smart enough to automate many long-running but ordinary knowledge-work tasks; the missing ingredient is a reliable general-purpose harness that lets them learn, use tools, maintain context, and persist over time. For Zhang, that is both a research opportunity and a call for more unusual, high-conviction experiments. Thank you for listening to Latent Space in 3 minutes from The Daily FM. See you next time!

View the whole library →

Subscribe to one of these pods with the RSS feed, or create your own →