The robots are at it again, and this time they’re solving, and breaking(?), math and cryptography! Things have been happening in the lattice corner including new leapfrogging complexity results in the closest vector problem (CVP), AND a possible poly-time quantum attack against the dihedral coset problem (DCP) that made everyone freak out for about a week (UPDATE: looks like it’s busted: https://eprint.iacr.org/2026/1693). ALSO, there was an important distinguisher attack against Classic McEl...

Ted Dintersmith, author of Aftermath: The Life-Changing Math that Schools Won't Teach You, argues that much of what we learn in math can be outsourced to machines. That frees people to understand the mathematical ideas that underlie much of modern life. An example is estimation, such as political polls. Understanding the various metrics that fill the news requires creativity coupled with logic. Learning to ask the right questions is more important than, say, learning to do closed-form

OpenAI's next model can find and exploit zero-day flaws on its own. Cloudflare's new bot filter retrains itself off a trillion requests a day. Offense and defense, both automated. We go through the deal quietly funding the power behind new AI data centers, the Kubernetes update that finally automates encryption key rotation, Cognition's fast-rising valuation, and a coding benchmark claim worth doubting. This is a public episode. If you would like to discuss this with other subscribers or get access

Proxy infrastructure is often the silent cost center killing the economics of large-scale data operations. This episode of Development makes the case that shared datacenter proxies — frequently dismissed as a budget compromise — are actually a powerful, purpose-built tool when matched to the right workloads. Drawing on this deep-dive on scalable, affordable proxy infrastructure, the episode walks through the mechanics, the ideal use cases, and the limits of shared datacenter IPs in a modern

Homotopy Type Theory, Narya & the Future of Proof Assistants with Michael Shulman. How does homotopy type theory bridge the gap between abstract mathematics and computational proof systems? Mike Shulman (University of San Diego) joins Deniz and Thorsten to discuss his journey from topology to higher observational type theory, the development of the Narya proof assistant, and how these tools are reshaping the way we think about equality, equivalence, and computation in mathematics.

Even more infinity?...

In this episode, we unpack why sparse, final-only rewards hobble reinforcement learning in large language models and how dense rewards via a process reward model act like a patient teacher, giving praise for micro-steps along the way. We explore how fortifying these steps reshapes the model’s reasoning, why broad, inconsistent feedback can cause global unlearning, and what this means for building AI that can truly reason across domains. Note: This podcast was AI-generated, and sometime...

Google pulled uBlock Origin and every other Manifest V2 extension from the Chrome Web Store today. The FTC says Amazon secretly added a hidden price floor to its ad auctions, costing advertisers twenty billion dollars or more. Different companies, same instinct. We go through OpenAI's test of paying only when a task actually completes, Google's new forecasting model built for time series data, Tim Cook's move to executive chairman after fifteen years, and the EU's toughest rules landing on ChatGPT,

Closing a manufacturing acquisition is only the beginning. The real test arrives when new ownership steps into a facility where machines, people, and habits are all still running on the previous playbook. This episode of Development digs into the case for operational improvement as the primary value driver in manufacturing buyouts — and explains why the gap between what a deal looks like on paper and what it becomes in practice is almost always an operations problem.The episode walks

Ever feel like the universe is playing by a different set of rules? In this episode of the Math Deep Dive Podcast, we peel back the intimidating jargon of quantum physics to reveal the elegant—and deeply strange—mathematics running the show.We begin with a mind-bending paradox: how can two isolated boxes, separated by 230 million miles, "cheat" classical logic to achieve a failure rate that should be mathematically impossible?. We move from the "ultraviolet catastrophe" that

Kubernetes just made bearer tokens replaceable with proof of possession, so a leaked token no longer means a stolen identity. OpenAI is cutting Cursor off from its models this November, after SpaceX bought the company for sixty billion dollars. Same week, control over AI access got shakier. We go through Datadog's claimed million-dollar-a-month AI savings, AI scrapers eating a fifth of the Linux kernel's servers, a Postgres typecast that made one query a thousand times faster, and Musk's warning

“He asked me this question because I'm his sister and he loves me.” Mathematics can often be seen as a solitary endeavour, but as this quote from one of our latest podcast guests show, that is far from the truth. In this episode, we learn how a new proof has developed from one young mathematician talking to his sister – and then collaborating with their father and close family friend. We spoke to Roman Sauer and Uri Bader – and Uri’s two children Saar and Shaked – after they had attended a

Explore WikiSkill, an AI framework where agents permanently log every action, a wiki maintainer extracts root causes from failures, and a skill proposer updates actionable instructions. Through a three-layer mind—raw, wiki, and skill—the system records lessons and gates improvements, letting knowledge accumulate without erasing context. In tests across math reasoning, document analysis, and spreadsheets, a 9‑billion-parameter model with WikiSkill outperforms a 27‑billion-parameter model witho...

Sarah Lotspeich and Lucy D'Agostino McGowan introduce Regression to the Mean Girls, a podcast about data-driven side quests in pop culture. In this episode, Sarah analyzes 45 recent romance novels to find out how different romance tropes affect the timing of the couple’s first kiss.Check out the full analysis here: https://regressiontothemeangirls.com/Links to content we chat about on the episode:Edward L. Kaplan and the Kaplan-Meier Survival CurveNonparametric Estimation from Incomplete

Project schedules fail for a reason that rarely gets named directly: the estimates feeding them confuse effort with duration. This episode of Development tackles that specific gap — not as an abstract concept, but as a concrete planning problem with a concrete solution. If your projects consistently run late despite solid teams and genuine commitment, the culprit is almost certainly hiding in how estimates are collected and scheduled from day one.The episode walks through why the

KinMarq provides a comprehensive digital identity platform designed to ensure the lifelong safety and traceability of pets. The service utilizes a tamper-evident history that combines traditional data, such as microchip numbers and medical records, with innovative biometric snout recognition to create an unchangeable profile. Pet owners can manage ownership transfers, share access with co-guardians, and activate a "Lost Mode" that utilizes QR code tags and AI-assisted matching to facilitate q...

Shipping an internal AI tool is the easy part. Knowing whether it's still working six months later — that's where most teams go silent. This episode of Development tackles the "eval gap": the absence of any repeatable system to detect when an AI tool's output quality has silently drifted, and what it actually takes to close it.The episode walks through a concrete, step-by-step framework for building an evaluation practice around custom internal tools — from structured data extraction to

Can maths help you win at games?In this AfterMaths Summer Special, Jon and Becky explore the line between luck and strategy — and whether understanding the maths behind a game can actually give you an advantage.They take in Rock Paper Scissors, Noughts & Crosses, chess, a deceptively simple counting game, the chessboard-and-rice problem, expected value, roulette and the mathematics hiding inside Monopoly.Along the way: why some apparently random games aren’t quite random, why exponential growth

Harvard Business School has launched HBS Foundry, an affordable online accelerator that utilizes synthetic AI avatars to provide scalable mentorship to entrepreneurs. These digital clones, modeled after actual faculty and venture capitalists, offer real-time feedback during simulated investor pitches and board meetings. While the program increases accessibility and allows for high-frequency practice, it also introduces concerns regarding intellectual property rights and the visual "uncanny va...

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