We discuss an optimized approach to state–prediction separation by implementing a free pause token that decouples context summarization from next-token prediction. By running a secondary prediction stream that shares weights with the primary backbone but writes no new keys or values, the model achieves better performance without increasing inference latency or memory overhead. The authors utilize a two-pass training split and a shared gated feed-forward network to significantly reduce the com...

A funding round could value OpenAI at one point two trillion dollars, before it's even public. A website charged an AI agent one cent to read a page, and still isn't sure it got paid. Same day, wildly different amounts. We go through what that trillion-dollar number actually rests on, how a new payment protocol lets agents pay by the fraction of a cent, and why platform teams are swapping static login keys for certificates that expire in minutes. Get full access to The AI Practitioner at

In the final episode of our Recommender Systems season, we explore the growing questions of trust, manipulation, privacy, fairness, sustainability, and user control. From fake reviews and shilling attacks to explainable recommendations and user-selected algorithms, we look at what happens when recommender systems must answer not only for what they recommend, but for the consequences of those choices.

Can AI prove the Riemann Hypothesis? Tudor Achim, CEO of Harmonic and creator of Aristotle — the first AI to win IMO gold and solve Erdős problems using the Lean theorem prover — joins Deniz and Thorsten to discuss how mathematical superintelligence is transforming research, education, and the very nature of proof.

This week, Anna catches up with Richard Liang from Peer, formerly ZKP2P. They trace the project’s evolution over the past two years, from their initial ZKP2P product that used ZK email proofs to connect Venmo payments with on-chain USDC, to their use of ZK TLS, and most recently, their move to using TEEs (trusted execution environments) and the rebrand to Peer. Richard breaks down what motivated each shift and how the team has navigated the trade-offs around privacy, speed, and UX.The conversation

Jeff Dunn, President and CEO of Cooperative Benefits Group, joins host Ankit Nanda, FSA, MAAA, FCA, to map the U.S. payer landscape and the forces reshaping it. Drawing on a career that runs from managed care pharmacy into benefits consulting, he sets out how employers, health plans, pharmacy benefit managers, providers, and consultants fit together, where incentives genuinely align and where they only appear to, and what the fiduciary and transparency environment means in practice. The

Welcome to a new season of Carry the Two, the podcast about how math and statistics impact the world around us from the Institute for Mathematical and Statistical Innovation.Hosts Sadie Witkowski and Ian Martin return  for this season of Carry the Two, where we are exploring the intersection of AI and mathematical research. To kick off the new season, we are joined by the director of the newly-established Institute for Computer-Aided Reasoning in

Professional forecaster Molly Hickman breaks down what it really means to assign a probability to the future — and why she believes generalists often out-forecast subject-matter experts. This episode explores the art and science of forecasting, from techniques to ethical considerations, and how AI and prediction markets are shaping our understanding of the future.Key TopicsThe definition of forecasting and its importanceTechniques for starting in forecastingThe role of AI and large language models

Web accessibility is too often treated as a finishing touch — something to address once the design is locked and the code is shipped. This episode of DEV.co challenges that habit directly, making the case that building accessible websites from the ground up isn't just better ethics, it's better engineering. Drawing from the DEV.co article on bottom-up accessible web development, the episode walks through why the sequence of your development process determines the quality of what you

Number of the Day - "11" This podcast is produced by Klassic Studios. Check out our other podcasts there! Sources Number sequence definitions come from the On-Line Encyclopedia of Integer Sequences (OEIS). Historical material is adapted from Wikipedia, used under the Creative Commons Attribution-ShareAlike licence. Minor planet details come from the NASA/JPL Small-Body Database. Learn more about your ad choices. Visit megaphone.fm/adchoices

A deep dive into the reported formalization of Andrew Wiles’s proof of Fermat’s Last Theorem using Anthropic’s Claude, Lean, and a dependency-driven “Prove 2Me” framework in just 11 days. We explore why formal verification is so demanding, how AI agents can coordinate millions of lines of code and thousands of intermediate theorems, and what machine-checked mathematics could mean for science. Note: This podcast was AI-generated, and sometimes AI can make mistakes. Please double-c...

Hidden code in iOS twenty-seven shows Apple built a way to swap Google's Gemini out of Siri for Claude or ChatGPT. A leaked Claude iPhone build shows screens for linking your actual bank accounts. Same week, opposite bets. We go through the swap mechanism buried in Apple's code, the bank-linking feature Anthropic hasn't confirmed, and the public fight between Dario Amodei and Jensen Huang over whether AI labs should slow down at all. Get full access to The AI Practitioner at

Is our universe a static "block" of frozen geometry, or is it a dynamic river of constant change? In this episode of the Math Deep Dive Podcast, we explore the revolutionary framework of Process Theory—a new mathematical language designed to capture a world that is "becoming" rather than just "existing."We begin by examining the structural failures of traditional, "syntactical" mathematics, which relies on static symbols and fixed equations that often fail at

Anthropic's CEO says the industry needs to slow down AI capability growth. A swarm of OpenAI's own agents flooded a code registry with two thousand malicious packages, for reasons nobody has explained. That's the evidence he's citing. We go through what changed his mind, how the RubyGems attack actually worked, a Postgres database that hit one hundred eighteen million queries a second, and why OpenAI is delaying its IPO. Get full access to The AI Practitioner at

Alex is a friend of mine and a mathematician.  He did his PhD in mathematics with other Alex (Martsinkovsky), who has presented at the seminar group previously.  We're huge fans of both Alexes at the Boston Computation Club and we hope you enjoy this talk as much as we did!

Today Daniel Yu gave a pure math talk on an extension he performed independently of work he had conducted during his math semester abroad in Budapest.

Today Woohyeuk (Kevin) Lee joined us to discuss local AI (and the question of just how local it really is). This is exciting work at the intersection of computer science and social science! Enjoy.

Casey Martin is a dirtbag intellectual, local savant, and recent PhD in the Boulder metro area. He's a party favorite and today he joined us to discuss cellular automata and their generalizations. Enjoy!

Hosted on Acast. See acast.com/privacy for more information.

Number of the Day - "10" This podcast is produced by Klassic Studios. Check out our other podcasts there! Sources Number sequence definitions come from the On-Line Encyclopedia of Integer Sequences (OEIS). Historical material is adapted from Wikipedia, used under the Creative Commons Attribution-ShareAlike licence. Minor planet details come from the NASA/JPL Small-Body Database. Learn more about your ad choices. Visit megaphone.fm/adchoices