Podcasts about Data science

In this episode of the AdTech Heroes podcast, Jon Wallett talks with Ashok Ganapam, President, AdTech and Monetization at MediaMint, about how agentic AI and data science are transforming decision-making in ad tech.They explore what agentic AI really means beyond the hype, how it's reshaping ad operations, engineering, and business intelligence, and what it takes for teams to adopt it successfully.Interested in being a guest? Contact us: adtechheroespodcast.com/contact

We break down what changes when AI agents become a bigger source of web visits than humans, and why measurement is the only sane response to generative search volatility. We also map the new playbook for go-to-market leaders who want durable advantage by building context, instrumentation, and systems that actually learn.•Agent analytics and deeper web measurement through server logs and segmentation•Turning GEO and AEO into a trackable marketing channel with baselines and experiments•Why the website shifts from destination to structured knowledge layer for AI agents•Context engineering through schema, FAQs, and connected site structure•Why “renting cognitive logic” from LLMs fails to create a moat•Decision tracing as a system for learning from GTM experiments over time•AI integration tax in fragmented stacks and how dynamic blindness breaks workflows•What a harness is and how constraints, monitoring, and stopping rules reduce risk•Leading the shift from point-solution operator to multi-agent orchestrator•Moving boards from token costs to value per token and closed-loop outcomesAI agents are quietly rewriting your analytics dashboard, and most teams are still looking at the old numbers. When bots and crawlers can drive more web visits than humans, “traffic” stops being a simple KPI and becomes a strategy problem: Who is visiting, what model sent them, where do they land, and do they convert? We sit down with Alexander Liss, Executive Advisor of AI Transformation at Brainworks and former VP of Data Science and AI at Huge, to talk about agent analytics, server logs, segmentation, and how to treat generative search optimization as a real, measurable marketing channel.We also go into the "website's" near-death and rebirth. The site is not disappearing, but early discovery is moving into AI Overviews and assistants, which means your content has to work as structured knowledge for machines and as high-trust depth for humans who arrive later. We cover context engineering, schema, FAQ patterns, and why social listening is back as models pull signals from places like YouTube, Wikipedia, and community forums, then change their minds a week later.From there, we get blunt about competitive advantage. If you are only renting cognitive logic from base models from Anthropic, OpenAI and Google, you are not building a moat. We unpack decision tracing, the AI integration tax, dynamic blindness in multi-agent workflows, and what it means to build a harness with constraints, monitoring, and stopping conditions so AI systems stay useful and cost-effective. We close with a fast, fun Spark Tank segment on Japanese business systems plus the “three feet from gold” resilience story. Subscribe, share with a growth leader, and leave a review. What part of your go-to-market stack needs better AI measurement first?Alexander Liss: https://www.linkedin.com/in/aliss77777Alexander Liss is the Executive Advisor of AI Transformation at BrainWorks and former VP of Data Science & AI at Huge. A data leader and systems-builder based in Denver, Colorado, his recent work includes pioneering Huge's emerging GEO practice and deploying NBCUniversal's award-winning conversational assistant, Oli, for the 2026 Winter Olympics. His prior leadership spans scaling enterprise analytics and machine learning strategies at global digital powerhouses like Accenture Song, VML and 22squared. He is currently pursuing his Master of Science in Artificial Intelligence from the Georgia Institute of Technology, holds an MBA from NYU Stern School of Business, and earned his Bachelor's in Japanese Language and Literature from The George Washington University. Website: https://www.position2.com/podcast/Rajiv Parikh: https://www.linkedin.com/in/rajivparikh/Email us with any feedback for the show: sparkofages.podcast@position2.com

Ce mardi 29 septembre, Amaury Delplancq, directeur général EMEA chez Dataiku, est revenu sur la datascience en mode collaboratif, le passage à l'échelle de l'intelligence artificielle en entreprise ainsi que le Dataiku Summit 2026 qui s'est tenu ce mardi 29 septembre à Paris, dans l'émission Tech&Co Business présentée par Frédéric Simottel. Tech&Co Business est à voir ou écouter le mardi sur BFM Business.

Pour ce nouvel épisode de Voice of Industries enregistré en direct au salon Tech For Industry Show, Mathieu Cura reçoit Nadège Brun, directrice stratégie IA et data pour la division One Tech de TotalEnergies. Au fil de l'échange, elle raconte comment un simple fichier Excel créé par un ingénieur de site est devenu aujourd'hui, grâce à la data science et au machine learning, une application déployée sur tous les sites de la branche Raffinage Chimie, pour optimiser la consommation énergétique des procédés. Elle revient aussi sur les freins humains rencontrés et sur l'IA générative comme prochain cap à ne pas manquer. Elle conclut avec trois leviers clés pour réussir sa transformation digitale : acculturation, collaboration et gouvernance de la donnée.Hosted on Ausha. See ausha.co/privacy-policy for more information.

Three rival AI CEOs never agree on anything, until now. I break down the Sanders-Casar bill, the timing, and why "too dangerous" might be the best pitch deck line money can buy. Struggling with AI and Software? Here is my new book available for purchase on https://mythicaltokenmonth.com/ Come have a chat on our Discord Channel:  https://discord.gg/4UNKGf3 Buy me a coffee https://ko-fi.com/datascience Check my company Amethix Technologies https://amethix.com  References https://defragzone.substack.com/p/techs-dumbest-mistake-why-firing ✨ Connect with us! Personal newsletter: https://defragzone.substack.com

Data modernization is helping public health agencies strengthen their systems, improve data analysis, and respond more quickly to emerging health threats. Colin Gerber, senior program analyst at the Council of State and Territorial Epidemiologists, discusses the Data Science Team Training program, an initiative that helps public health staff build data science skills while working on projects that address their agencies' priorities.  Gerber explains how the program supports state, tribal, local, and territorial health departments through project-based training, professional development, subject matter expert coaching, and peer collaboration. He also shares examples of participating teams, including a tribal health department that built a disease surveillance system from scratch and a Los Angeles County project that helped hospitals monitor potential surges in capacity. Finally, he discusses how the program can benefit tribal public health professionals and how agencies can apply.Data Science Team Training (DSTT) | Council of State and Territorial Epidemiologists (CSTE)Legislative Action Bridging Public Health and Clinical Health Care | ASTHO2026 Telehealth Needs Assessment | ASTHO25 Years After 9/11, World Trade Center Toxins Continue to Claim Lives

Machine learning is what most data scientists know best. But when it comes to building systems that enable better decision-making, machine learning is just one piece of the puzzle. And leaning on it too much can leave your entire decision system exposed.In this Value Boost episode, Adam DeJans Jr joins Dr Genevieve Hayes to walk through the eight disciplines that make up the modern decision stack and where data scientists should focus to close the gaps in their own skill set.You'll discover:The eight disciplines that make up the modern decision stack [01:24]The most overlooked layer in most decision systems [04:47]Why optimization textbooks are teaching data scientists the wrong things [06:17]What to learn next if you want to build better decision systems [09:45]Guest BioAdam DeJans Jr. is a decision scientist and optimisation expert who has led high-stakes decision systems in complex, uncertain environments, at companies including Amazon, Toyota and Ford. He is the co-founder of AI and decision intelligence consultancy Bit Bros and co-author of The Decision Factory: A Novel About Decisions Under Uncertainty.LinksConnect with Adam on LinkedInBit Bros websiteConnect with Genevieve on LinkedInBe among the first to hear about the release of each new podcast episode by signing up HERE

We explore what it takes to turn powerful AI capabilities into products customers actually use and trust, from conversational intelligence and context management to human-AI collaboration, product metrics, and moving beyond AI pilot purgatory.Key Highlights:What separates an AI feature or assistant from a truly agentic product capable of taking action.How product teams should think about context, permissions, user control, reliability, and human oversight.Turning calls, messages, and customer interactions into actionable insights and intelligent workflows.Why promising AI pilots often struggle in production and what product teams can do differently.Why task completion isn't enough and how teams should think about quality, customer experience, trust, and business outcomes.

In episode 34 of Recsperts, I'm joined by Raghav Saboo, Staff Machine Learning Engineer at DoorDash and Tech Lead for Personalization and Search for New Verticals — groceries, convenience, retail, alcohol, pet supplies and more, beyond the original restaurant vertical. We discuss the particular challenges of personalized recommendation, ranking and search in quick commerce, recent trends in generative recommendations and the application of Semantic IDs to item ranking and query reformulation. Raghav's path into recommender systems started in chemical engineering before he moved into ML consulting, a Master's in Statistics, Machine Learning and Econometrics from Duke University, and building LLMs for new language launches on Amazon's Alexa AI, ahead of joining DoorDash.We start with the marketplace itself: DoorDash connects consumers, merchants and couriers, and growing it means balancing the interests of all three so that the platform stays healthy for everyone on it. Raghav walks me through how his team frames the consumer side around three pillars — familiarity (surfacing what a consumer already trusts), affordability (matching price sensitivity and timely deals) and novelty (introducing new items and categories without adding friction). From there we get into how DoorDash uses LLMs to build "memory blocks," structured natural-language representations of a consumer organized around semantic domains like dietary preference, pet ownership or trusted brands, and how these feed LLM-generated collections that get resolved into real items through embedding-based retrieval.We then turn to DoorDash's move to generative approaches, centered on Semantic IDs: hierarchical product identifiers learned through recursive clustering of item content embeddings, forming a taxonomy that captures attributes a human-built catalog structure might miss — as Raghav puts it, "within e-commerce, items really carry a lot of meaning." He walks me through two production use cases: replacing dozens of taxonomy-based dense features in the ranking model with Semantic ID n-gram aggregations while improving online metrics, and using Semantic IDs for query reformulation in search, letting the system traverse a learned hierarchy to refine or diversify a query. This connects to DoorDash's own paper on the topic and to a broader conversation about why search, recommendation and agentic ordering — DoorDash's own "Ask DoorDash" — are converging on a shared substrate of Semantic IDs and consumer memory, while today's app surfaces still need to grow more flexible for that convergence to feel seamless.We close with a preview of the RecSys 2026 tutorial "Recommender Systems in Delivery Platforms: Challenges, Solutions and Learnings," which Raghav is co-presenting with Wolt's Paavo Camps and myself, and his advice for navigating a field that reinvents itself every quarter: be honest about whether that pace suits you, use AI agents to filter what's worth your attention, and build the judgment to recognize dead ends early.Enjoy this enriching episode of RECSPERTS – Recommender Systems Experts.Don't forget to follow the podcast and please leave a review.(00:00) - Introduction (02:19) - RecSys 2026 Tutorial Preview (03:23) - About Raghav Saboo (10:47) - Working on Amazon Alexa AI (14:13) - About DoorDash (16:31) - Operating Model of a Multi-Sided Marketplace (20:45) - Affordability, Familiarity and Novelty (34:01) - Advantages of LLM-based Consumer and Item Profiles (46:53) - Generative Recommendations (59:21) - Semantic IDs for Item Ranking and Query Reformulation (01:19:49) - Agentic Shopping vs. Conversational RecSys (01:29:39) - Tutorial on Recommender Systems in Delivery Platforms (01:34:05) - Closing Remarks Click here to view the episode transcript. Links from the Episode:Raghav Saboo on LinkedInRaghav Saboo's SubstackUsing LLMs to Infer Grocery Preferences from Restaurant OrdersBuilding Ask DoorDash (Part 2): IntelligenceBridging Affordability, Familiarity, and Novelty (KDD 2025)Building a Unified Consumer Memory for Personalization at ScaleOffline LLMs, Online Personalization: Generating Carousels at DoorDashRecSys 2026 TutorialsWorkshop on Unified Search and Recommendation (USRW) 2026Papers:Xu et al. (2026): One Hierarchy, Two Systems - Semantic Product IDs for Discovery-Surface Ranking and Search-Page Query Reformulation (RecSys 2026, USRW Workshop)Xi et al. (2026): Mine and Refine - Optimizing Graded Relevance in E-commerce Semantic Search Retrieval (CIKM 2026, Applied Research Track)Chen et al. (2026): Joint Optimization of Relevance and Engagement in Multi-Task Ranking for E-Commerce with Efficient LLM Supervision (SIGIR Industry Track)Sinha et al. (2025): Mind the Gap - Bridging Behavioral Silos with LLMs in Multi-Vertical Recommendations (RecSys 2025 GenAI Workshop Talk)Das et al. (2024): Applications of LLMs in E-Commerce Search and Product Knowledge Graph - The DoorDash Case Study (WSDM 2024)General Links:Follow me on LinkedInFollow me on XSend me your comments, questions and suggestions to marcel.kurovski@gmail.comRecsperts Website

Stehst du auch manchmal vor den neuesten KI-News und fragst dich: Ist das noch Fortschritt oder kann das weg?Während die ersten Lebkuchen schon im Spätsommer die Supermärkte verstopfen, ziehen Dr. Christian Krug und Co-Host Barbara Lampl ihren Jahresrückblick 2026 einfach mal vor. In der sechsten Ausgabe von Unicorns and Lightsabers räumen wir im Datendschungel auf und sezieren den State of AI 2026.Dreieinhalb Jahre nach dem Release von GPT-4 ist die Hype-Phase vorbei – aber was hat sich wirklich getan? Christian und Barbara werfen den Bullshit-Detektor an und klären:Der Sprung von GPT-4 zu GPT-6: Warum Claude Opus 4.6, Mythos und GPT-6 Astra die Spielregeln im Coding und Data Science komplett gedreht haben.Drama in der Mathe-Welt: Wie OpenAI mit 130 Milliarden Token ein ungelöstes Millennium-Problem (Navier-Stokes) knackt und warum der eigentliche Skandal menschlich ist, nicht technisch.38 Sonderlocken im Einkauf: Warum Unternehmen lieber 10.000 € Token-Kosten verbrennen sollten, statt den 25. Arbeitskreis für digitale Transformation zu gründen.Die gefährliche Mitte: Warum veraltetes Halbwissen und krampfhaftes Prompt-Babysitting der direkte Highway ins Abseits sind.Kein Marketing-Sprech, kein „This changes everything“-Geblubber, sondern schonungslose Einordnung für die Praxis.Hör rein, bewerte den Podcast mit 5 Sternen und schreib uns in die Kommentare: Wo nimmst du bei KI noch die faule Abkürzung?▬▬▬▬▬▬ Profile: ▬▬▬▬Zum LinkedIn-Profil von Barbara: https://www.linkedin.com/in/barbaralampl/Zum Podcast LAIer 8|9: https://laier89.podigee.io/Zum LinkedIn-Profil von Christian: https://www.linkedin.com/in/christian-krug/Christians Wonderlink: https://wonderl.ink/@christiankrugUnf*ck Your Data auf Linkedin: https://www.linkedin.com/company/unfck-your-data▬▬▬▬▬▬ Buchempfehlung: ▬▬▬▬Alle Empfehlungen in Melenas Bücherladen: https://gunzenhausen.buchhandlung.de/unfuckyourdata▬▬▬▬▬▬ Hier findest Du Unf*ck Your Data: ▬▬▬▬Zum Podcast auf Spotify: https://open.spotify.com/show/6Ow7ySMbgnir27etMYkpxT?si=dc0fd2b3c6454bfaZum Podcast auf iTunes: https://podcasts.apple.com/de/podcast/unf-ck-your-data/id1673832019Zum Podcast auf Deezer: https://deezer.page.link/FnT5kRSjf2k54iib6Zum Podcast auf Youtube: https://www.youtube.com/@unfckyourdata▬▬▬▬▬▬ Merch: ▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬https://unfckyourdata-shop.de/▬▬▬▬▬▬ Kontakt: ▬▬▬▬E-Mail: christian@uyd-podcast.com▬▬▬▬▬▬ Timestamps: ▬▬▬▬▬▬▬▬▬▬▬▬▬00:00 Intro: Lebkuchen im September & der vorgezogene Jahresrückblick 202602:38 Der Schreibtisch-Test: Wenn GPT-6 komplexe Data-Science-Probleme in 20 Minuten löst06:26 Drei Sprünge im Rekordtempo: Opus 4.6, Mythos und GPT-6 Astra08:45 Die drei Gruppen der Gesellschaft: Warum die Halb-Informierten am gefährdetsten sind15:02 Mathe-Drama um Navier-Stokes: Hat OpenAI das Millennium-Problem geklaut?21:36 Moral vs. Maschinenleistung: Wenn die AI-Quote bei über 75 % liegt25:13 Context Engineering statt Chatbot-Hype: Wie echte AI-Beweisführung funktioniert33:03 Vom Steinzeitalter zu 20 Goldklumpen: Der Sprung von 8.000 Token zu Agent Swarms34:23 Digitale Transformation vs. AI-driven Transformation: Der 38-Sonderlocken-Einkaufsprozess43:40 Token-Rechnung im Vorstand: Warum Data Science Teamsport bleibt48:22 Gefahr der Selbstverdummung: Wenn KI das Denken ersetzt55:45 Klare Kante: Nicht der gleiche Planet wie GPT-458:02 Outro & persönliche Fails: Teure Schuhe, Kendo-Basics und Spotify-Kommentare

Wie ticken Big-Data-Tekkies? Glauben sie noch an die freundliche digitale Utopie? Und sind sie noch der Überzeugung, dass wirkungsvolle gesellschaftliche Kontrolle und demokratische Mitgestaltung der Big-Data-Wirtschaft nicht von außen, aus der Gesellschaft, entsteht, sondern von innen, dort wo Leute sich auskennen mit großen Datenströmen, Algorithmen und künstlicher Intelligenz? FOLGE 'HÖR MAL DEUTSCHLAND' IN SPOTIFY GARANTIERT KI-FREI !Nach einer Prognose des Bureau of Labour Statistics, einer Fachabteilung des Arbeitsministeriums der Vereinigten Staaten, wird der globale Bedarf an Data-Scientists in den nächsten Jahren stark ansteigen, bis 2030 um über 30 Prozent. Weltweit werden sie heiß umworben und in den Konzernen verwöhnt. Doch ist das gut oder schlecht? Von Anja Kempe. Regie und Produktion: Anja Kempe. SWR Kultur — Feature 2022Die Programmierer - Eine besondere Spezies und ihre Haltung.

Utforsk verdenen av kunstig intelligens i denne episoden av "Kunstig Intelligens - Bak KI-masken". Lær om de verktøyene som former samfunnet vårt, og hvordan KI brukes i hverdagen for å øke effektiviteten og fremme innovasjon. Programlederen Neda-Maria Kaisumi deler sin innsikt og erfaring, og gir seerne en forståelse av hvordan KI har utviklet seg fra Alan Turings tid til dagens generative AI-løsninger. Dette programmet tilbyr en grundig innføring i hva kunstig intelligens er, og hvordan det kan brukes til å forbedre hverdagen.

This week, Kenny Olson and Jay Kolls sit down with Dr. Manjeet Rege, professor and chair of the Department of Software Engineering and Data Science at the University of St. Thomas and director of its Center for Applied Artificial Intelligence. Rege, the author of Ethics and Governance of Artificial Intelligence, cuts through the hype and the panic. The conversation covers what AI actually does well, where it still fails, who is accountable when it goes wrong, and what “ethical AI” means once the press release is over. Bias, deepfakes, jobs, data centers, governance, and whether Minnesota is ready for any of it all get a turn at the counter.See Privacy Policy at https://art19.com/privacy and California Privacy Notice at https://art19.com/privacy#do-not-sell-my-info.

Lint the entire CPython code base from scratch. It takes 0.3 seconds. Three blinks of an eye. That is ruff, and it is written in Rust. So are Pydantic, Polars, uv, and Granian. Rust shows up in Python three ways: tools that happen to be Rust, libraries Python imports, and servers that run Python inside Rust. This is Rust for Python developers, not Rust experts. Christopher Trudeau is back on Talk Python to discuss Rust and his latest course Up and Running with Rust. The core rule is that only one thing can own a value at a time. Pass it around freely in Python and the garbage collector cleans up. Do that in Rust and it will not compile. Episode sponsors Sentry Error Monitoring, Code talkpython26 Python in Production Talk Python Courses Links from the show Up and Running with Rust course: training.talkpython.fm Rust: rust-lang.org pydantic: pydantic.dev ruff: docs.astral.sh granian: github.com By example: doc.rust-lang.org rust-lang.org: rust-lang.org rustup.rs: rustup.rs crates.io: crates.io main.rs: main.rs PyO3: github.com https://github.com/ritwiktiwari/awesome-python-rs: github.com ty: docs.astral.sh pyrefly: pyrefly.org uv: github.com polars: pola.rs Watch this episode on YouTube: youtube.com Episode #563 deep-dive: talkpython.fm/563 Episode transcripts: talkpython.fm Theme Song: Developer Rap

This week I kick off a new series on the show: My Favorite Graphs. I'm going to walk through the projects that shaped how I think about data visualization, data communication, and data storytelling. I'm starting with the Guardian's “Bussed Out: How America Moves Its Homeless,” published in December 2017, an 18-month investigation into the one-way bus tickets that US cities hand out to homeless people. There's a special treat at the end too!My Favorite Graph: Bussed OutSubscribe to this podcast. Follow this show and everything dataviz on Instagram, LinkedIn, Substack, X, and YouTube. Check out the PolicyViz websiteto learn more about data and data visualization.Questions? Comments? Pitch ideas? Email me at: jon@policyviz.comHosted by Zencastr where you can record, transcribe, edit, and automatically publish your meetings, podcasts, and more.

Data scientists are trained to build better models. But stakeholders don't wake up hoping for a better model. They wake up hoping to make better decisions. And the gap between those two things is where most data science value quietly disappears.In this episode, Adam DeJans Jr joins Dr Genevieve Hayes to share what it actually looks like when an organisation stops optimising for better predictions and starts building systems that enable better decisions - that is, decision factories.You'll discover:What separates a decision factory from a machine learning pipeline [02:11]Why uncertainty is an input into good decision making [10:51]How to test a decision policy before it costs you real money [12:11]Why giving stakeholders options beats giving them the optimal solution [18:48]Guest BioAdam DeJans Jr. is a decision scientist and optimisation expert who has led high-stakes decision systems in complex, uncertain environments, at companies including Amazon, Toyota and Ford. He is the co-founder of AI and decision intelligence consultancy Bit Bros and co-author of The Decision Factory: A Novel About Decisions Under Uncertainty.LinksConnect with Adam on LinkedInBit Bros websiteConnect with Genevieve on LinkedInBe among the first to hear about the release of each new podcast episode by signing up HERE

IM Thomas Willemze is a popular author, Chessable creator, and trainer who has worked with the Dutch National Youth Team, among countless other students. A data analyst by day, Thomas brings an analytical approach to thinking about the core components of chess improvement. In our conversation, we draw on Thomas's experiences as both a player and coach to tease out practical lessons for improving players. We discuss everything from game review and opening preparation to how to improve your focus while studying and playing. Thomas also shares some book recommendations, reflects on learning chess from the creators of the Steps Method while they were still developing it, and recounts what happened when, at age 40, he gave himself a year to seriously pursue his GM ambitions. Timestamps of topics discussed can be found below. Free perpetual Chess Discord: https://discord.gg/UpRTQfjQB Free Newsletter: https://benjohnson.substack.com/ 00:00 Introduction to the Interview with Thomas Willems Thanks to our sponsors, Claude AI, check them out at: Claude.AI/chess 02:01 Introduction to IM Thomas Willemze 04:07 The Intersection of Data Science and Chess 07:09 Thomas' Training and Coaching Philosophy 09:55 The Chess Steps Method and Early Chess Education mentioned : Rob Brunia and IM Cor van Wijgerden 15:36 Addressing Weaknesses and Improving Skills Mentioned: Fundamental Chess Endings, Dvoretsky's books, GM Yury Averbakh's books 18:08 Common Pitfalls for Adult Players 20:51 Patreon mailbag question: How can we improve our focus when playing games and studying?  23:55 Defensive Awareness and Identifying Threats 27:00 Patreon mailbag question: Does Thomas see parallels between chess and football/soccer strategy? 28:48 The Balance of Attack and Defense in Chess 30:28 Effective Study Strategies for Ambitious Players 32:07 Reviewing Games: Learning from Mistakes 33:34 The Competitive Landscape of Modern Chess 38:26 Preparing for Upcoming Tournaments 40:05 Choosing an Opening Repertoire 42:01 Current Projects and Future Endeavors 44:08 Chess Content Recommendations and Personal Interests Mentioned: 100 Endgames You Must Know  50:00- Thanks to IM Willemze for joining me. You can reach him via his Chessable author's page: https://www.chessable.com/author/Willemze/ Learn more about your ad choices. Visit podcastchoices.com/adchoices

Topics covered in this episode: Pandas Should Go Extinct Pydantic-pint puts real-world units in your Pydantic models How Libraries Run Rust Inside Python (With PyO3) AWS acquires DuckLabs Extras Joke Watch on YouTube Sponsored by Logfire from Pydantic: pythonbytes.fm/logfire Connect with the hosts Michael: Mastodon / BlueSky / X / LinkedIn Calvin: Mastodon / BlueSky / X / LinkedIn Show: Mastodon / BlueSky / X Join us on YouTube at pythonbytes.fm/live to be part of the audience. Usually Tuesday at 7am PT. Older video versions available there too. Finally, if you want an artisanal digest of every week of the show notes in email form? Add your name and email to our friends of the show list, we'll never share it. Calvin #1: Pandas Should Go Extinct Pandas' slowness pushes teams toward "Big Data" tools (Spark, Databricks) they don't actually need — most workloads never hit true Big Data scale Amazon Redshift telemetry: ~95% of tables are under 100GB, ~87% of queries touch 80GB or less — that's "Medium Data," not Big Data Polars and DuckDB fill that gap: single-machine, fast, no cluster required 1 Billion Row Challenge benchmark: Pandas took 4m28s vs. Polars 5.04s and DuckDB 5.19s — DuckDB also used 19x less memory On a real-world NYC taxi dataset (3GB parquet), pure DuckDB ran 2x faster than pure Pandas while using a fraction of the RAM Bonus: Apache Arrow lets you pass data between Pandas/Polars/DuckDB with zero copying, so trying them out doesn't mean a full rewrite Michael #2: Pydantic-pint puts real-world units in your Pydantic models Pydantic-pint bridges Pydantic and Pint so models can validate physical quantities like 4m or 12 meters instead of bare floats. Fields annotated with PydanticPintQuantity parse user input, convert between compatible units, and serialize quantities back out as strings. That closes a real gap for anything consuming API payloads, config files, or sensor data with measurements, letting you enforce units at the validation boundary instead of hoping every caller remembered them. via PyCoder's Weekly newsletter Unit mix-ups have literally crashed spacecraft; now your Pydantic models can refuse them at the door. Annotate a field as Annotated[Quantity, PydanticPintQuantity('km')] and inputs like 12 meters arrive auto-converted to kilometers Validation covers string, numeric, and quantity inputs, and model_dump_json serializes quantities as readable unit strings Installable from PyPI as pydantic-pint, MIT licensed, with docs at pydantic-pint.readthedocs.io Early-stage solo project at version 0.4, so API stability and maintenance are open questions worth discussing Calvin #3: How Libraries Run Rust Inside Python (With PyO3) Pydantic v2's validation core (pydantic-core) is Rust under the hood, built with PyO3 — this post shows how that bridge actually works via a small hand-built JSON parser Four steps to get Rust into Python: write a normal Rust module, annotate with PyO3 macros (#[pyfunction], #[pymodule]), compile/install with maturin, then just import it The parser builds a Rust tree first — Python never touches it until the boundary crossing Key insight: converting the Rust result into Python objects (.into_pyobject) is often the expensive part, not the parsing — 100,000 JSON values means ~100,000 Python objects built after parsing's already done Errors cross the boundary too: Rust's typed errors convert into real Python exceptions (ValueError, FileNotFoundError) via From/?, so callers get clean Python semantics Takeaway for anyone porting Rust in: if you're returning a scalar, don't sweat it; if you're returning a big structure, profile the boundary — that's the real cost, not the algorithm Michael #4: AWS acquires DuckLabs Thank you Dylan McConnell. What does this mean for the DuckDB ecosystem? DuckDB is the open-source in-process analytical SQL engine. MIT licensed. The IP is not owned by any company - it's held by the nonprofit DuckDB Foundation, which was created when the team spun out of CWI Amsterdam. Peter Boncz, the CWI representative on the Foundation board, describes it as the entity that holds all IP of open-source DuckDB. DuckLabs (ducklabs.com) is the company, formerly branded DuckDB Labs. Founded a little over five years ago by Hannes Mühleisen and Mark Raasveldt to give the DuckDB team a stable long-term home, bootstrapped deliberately instead of taking VC, grown to 30+ people in Amsterdam, funded by support and feature-prioritization contracts. It employs the core devs. It does not own DuckDB. DuckLake is one of three projects DuckLabs builds, what they call the Duck Stack: DuckDB, DuckLake, and Quack. DuckLake is the lakehouse format that puts catalog metadata in a SQL database instead of in files on object storage. Quack is newer - an RPC-style protocol that turns DuckDB into a client-server system where both ends are DuckDB instances, slated to stabilize in DuckDB v2.0 in September 2026. MotherDuck is a separate Seattle company, Jordan Tigani's, selling serverless hosted DuckDB. It was started in partnership with DuckDB Labs and has worked closely with Hannes and Mark for four years. It contracted DuckLabs for engineering work and contributes heavily upstream - three of its engineers are among the top 10 outside contributors to DuckDB. It also sells its own DuckLake offering. Customer and collaborator, never owner. What the AWS post changes. Amazon bought the company, not the project. DuckLabs joined AWS effective September 1, with the process concluding August 31, 2026. Hannes and Mark keep leading the team and the project's technical direction, the team stays in Amsterdam, and DuckDB stays MIT under the Foundation. AWS gets the people and a direct line to the roadmap. The license protects your code, not your priorities. Three second-order effects worth tracking: The Foundation board is the real question. It has three directors: Mühleisen, Raasveldt, and Boncz. Two now work for AWS. Commentary on the deal has focused on exactly this - the license protects the code, not the roadmap. The announced counterweight is governance: a technical advisory board on the Foundation, and opening the extension stack so extensions signed by other developers can run in DuckDB. MotherDuck immediately moved into the business DuckLabs vacated. It now sells DuckDB enterprise support, which it had avoided because it didn't want to compete with DuckLabs' business model, and says it has explicit blessing from Hannes and Mark now that they're joining Amazon. It also bought Tower.dev the day before the AWS announcement. Everyone expects an AWS DuckDB service. Tigani says Amazon will likely release one eventually, and welcomes the competition, citing Redshift's failure to slow Snowflake on AWS. The groundwork is already visible: Amazon Quick uses DuckDB to query S3 Tables and has processed over 2.5B queries with it since launching in October 2025. The DuckLake angle is the one to watch. AWS is heavily committed to Iceberg through S3 Tables, and it just acquired the team behind a competing lakehouse format. The stated plan is to use DuckDB, DuckLake, and Quack together to power a new generation of data services, but which format wins internal priority is unannounced. Extras Calvin: astral-sh/uv 0.12.12: code-signed release binaries

Jeffrey Havsy is the commercial real estate industry practice lead at Moody's Analytics, where he focuses on combining property fundamentals with economic, credit, climate, and alternative data to help lenders, investors, and operators make better decisions. He has spent his career at the intersection of economics and real estate, including roles at NCREIF and CBRE. Moody's Analytics covers more than 600 million public and private entities worldwide and generated over $7 billion in revenue in 2025. Jeffrey is based in Needham, Massachusetts.(02:46) From data scarcity to data curation(03:39) What lenders and investors miss beyond property fundamentals(05:34) Tenant credit, crime data, and truck traffic as signals(09:16) Moody's commercial location score explained(12:52) Cap rates: overrated, underrated, or misused(15:39) Moody's MCP launch and what it means for CRE decisions(18:32) Why Moody's stays AI platform-agnostic(20:44) Where the human stays in the loop(23:33) Physical risk vs. ESG(26:11) The hidden risks in industrial real estate(30:06) How robotics changes what a warehouse is worth(31:09) The one macro data point CRE ignores: productivity(33:10) Collaboration superpower: Abraham Lincoln

AI is getting so good at maths that it is now tackling, and apparently solving, some of the hardest mathematical problems that exist. But as huge companies like OpenAI and Anthropic race to outdo each other, human mathematicians are getting caught in the crossfire... and their field is changing forever. Jonathan Webb speaks to ABC AI Technology Reporter Cam Wilson about the furore over the Navier-Stokes equations and the million-dollar Millennium Prize.Featuring:Cam Wilson, AI Reporter at ABC.Professor Tristan Buckmaster, Mathematician at New York University.You can binge more episodes of the Lab Notes podcast with science editor and presenter Jonathan Webb on the ABC Listen app (Australia). You'll find episodes on animal behaviour, human health, space exploration and so much more.Get in touch with us: labnotes@abc.net.auThis episode of Lab Notes was produced on the lands of the Gadigal and Kaurna people.

AI, privacy, digital footprints, and trust take center stage as Walter Harrison explains the hidden data economy, while Dave reflects on 9/11, voting, and political fatigue.

How many files does your query read before it reads any data? On some data lakes, you go through JSON and metadata files first, just to learn which Parquet files matter. DuckLake asks one SQL question instead. The metadata lives in a real database. The data stays in plain Parquet. That's the entire format. Pedro Holanda joined DuckDB in 2018, when it was still a research prototype at CWI. He's the lead DuckLake developer. Guillermo Sanchez Dionis works on DuckLake and the new Quack protocol. With Quack as the catalog, DuckLake handles 200 transactions a second under heavy contention. No other open table format comes close. Episode sponsors Six Feet Up Talk Python Courses Links from the show Guests Pedro Holanda: pedroholanda.org Guillermo Sanchez: linkedin.com PhD on progressive indexes: ir.cwi.nl SQLite: www.sqlite.org Litestream: litestream.io boring hardware: talkpython.fm DuckDB: duckdb.org episode 491: talkpython.fm Iceberg: iceberg.apache.org manifesto: ducklake.select DuckLake: ducklake.select spec: ducklake.select this diagram: blobs.talkpython.fm Data inlining: ducklake.select ducklake-dataframe: github.com Polars course: training.talkpython.fm CSV parser: duckdb.org Zero-copy Arrow: duckdb.org ART index: duckdb.org async I/O: duckdb.org v1.0: ducklake.select Git-like branching: ducklake.select Watch this episode on YouTube: youtube.com Episode #562 deep-dive: talkpython.fm/562 Episode transcripts: talkpython.fm Theme Song: Developer Rap

Part of SAND's The Great AI Unraveling Series Artificial intelligence is often framed as a leap forward, a technical revolution, or an inevitable stage in human progress. But what if AI is also asking us to confront something older: the colonial logic embedded in how we define intelligence, progress, neutrality, knowledge, and the human itself? This conversation brought together voices working across technology, culture, race, power, and society to examine what AI reveals about the systems that built it. Rather than treating bias as a technical glitch, this conversation asks what must be unlearned at the level of worldview, institution, data, design, and collective imagination. Facilitated by Christian “ZacaTechO” Ortiz, this dialogue explored AI through a decolonial lens asking necessary questions: Whose knowledge is extracted? Whose futures are optimized away? Whose bodies are surveilled? Whose intelligence is dismissed? And what can liberated knowledge make possible beyond colonial technology? This is not a conversation about making machines more ethical within the same old systems. It is an invitation to question the systems themselves, and to imagine intelligence beyond extraction, domination, and control. Topics 00:00 Welcome, and ZacaTechO 00:04 Why researchers are trained to disconnect 00:09 Red teaming, and what it misses 00:12 John is the doctor, Juan is the patient 00:18 Christian's first encounter with ChatGPT 00:21 How bias became normal 00:26 Justice AI as a model of refusal 00:29 Breaking a language model in the classroom 00:32 One model to rule them all, and the logic of eugenics 00:35 Galton, Namibia, and the through line 00:39 What do we mean by intelligence? 00:41 Emancipatory data science, and where to begin Guests Christian ZacaTechO Ortiz is an Afro-Indigenous decolonial social scientist and technologist. He is the founder of Justice AI GPT and author of the Decolonial Intelligence Algorithmic Framework. He studies how colonial systems still shape law, education, entertainment, and our bodies. The name ZacaTechO came from his father, after the family's origins in Zacatecas. justiceaigpt.ca LinkedIn Academia.edu Profiled in Forbes, September 2025 Substack Dr. Thema Monroe-White is Associate Professor of Artificial Intelligence and Innovation Policy in the Schar School of Policy and Government and the Department of Computer Science at George Mason University. Her work spans bias mitigation in AI, critical computational methods, and emancipatory data science. She holds a PhD in Science, Technology and Innovation Policy from Georgia Tech, and master's and bachelor's degrees from Howard University. She is an advisory board member and fellow of the Institute in Critical Quantitative and Mixed Methodologies. Faculty profile, George Mason University Research profile The research discussed Laissez-Faire Harms: Algorithmic Biases in Generative Language Models — Shieh, Vassel, Sugimoto & Monroe-White. The open-ended prompting study described in the episode, the one where John is the doctor and Juan is the patient. Emancipatory Data Science: A Liberatory Framework for Mitigating Data Harms and Fostering Social Transformation — Monroe-White's framework for using computation to diagnose rather than to solve. Dismantling Eugenics Logics in Data Science and Artificial Intelligence — the through line from Galton to contemporary machine learning. On the Dangers of Stochastic Parrots: Can Language Models Be Too Big? — Bender, Gebru, McMillan-Major & Shmitchell, 2021. Emergent Abilities of Large Language Models — Wei et al., Transactions on Machine Learning Research, 2022. Scholars named in the conversation Timnit Gebru — founder of the Distributed AI Research Institute Joy Buolamwini — founder of the Algorithmic Justice League Ruha Benjamin — author of Race After Technology Safiya Noble — author of Algorithms of Oppression Meredith Broussard — author of More Than a Glitch Emily Bender — co-author of the Stochastic Parrots paper Historical references Francis Galton, who coined the term eugenics and whose statistical methods underpin modern machine learning. R.A. Fisher and William Shockley are named alongside him. The Nama and Herero genocide in Namibia, 1904 to 1908, discussed as a precursor to the Nazi Holocaust, and the site Galton had visited fifty years earlier. Tristan Harris is referenced on AI systems behaving unpredictably. His conversation with SAND is The Great AI Unraveling. Contact SAND podcast@scienceandnonduality.com Support the mission of SAND and the production of this podcast by becoming a SAND Member

Jamie Borodin is the founder and CEO of Docuverus, a document verification and fraud detection platform purpose-built for multifamily housing providers. With over 30 years in resident screening, Jamie founded Docuverus in 2017 to address a gap no existing tool was solving: authenticating income documents at the metadata level while simultaneously reading and calculating their contents. Docuverus is one of the only platforms with a direct connection to the Social Security Administration for synthetic identity detection. Jamie is based in southern New Jersey.(02:16) Why standard screening misses income fraud (05:03) How fraud evolved from COVID to AI-generated documents (07:13) What a sophisticated fake pay stub looks like today(09:09) What bad debt files actually reveal (13:10) Operators don't know how much bad debt is fraud (14:12) False security: when your PMS fraud tool isn't enough (18:27) Why bank account linking isn't fraud detection (19:11) Fraud hotspots: Houston, Atlanta, and why NYC is different(24:27) Social Security fraud and synthetic identity detection (25:36) What a CPN is and how it fabricates a credit file (27:58) Florida's felony law and the limits of regulation (29:29) Screening compliance: why the liability falls on the operator (33:27) Where fraud prevention goes as AI lowers the barrier (35:42) Collaboration superpower: Albert Einstein

Choosing which AI tools to buy is the easy part of an AI strategy. The hard part - and the part most organisations are ignoring - is everything that sits beneath the tools, from platform and inference decisions to hardware and sovereignty.In this Value Boost episode, Victor Coimbra joins Dr Genevieve Hayes to share his framework for cutting through the noise of AI tool selection to the strategic decisions that will actually determine whether an organisation's AI future succeeds or fails.You'll discover:The four layers of an agentic AI strategy and why most organisations only think about one [01:53]The three symptoms that signal an organisation's AI strategy is breaking down [05:52]The four questions that reveal what an organisation's AI strategy is actually missing [08:21]The one question data scientists should lead with when advising stakeholders on AI [10:43]Guest BioVictor Coimbra is a Partner and CTO at Artefact, the world's largest pure-play AI consulting firm and co-founded the firm's Latin American operations. In 2024, he was recognised in the Forbes 30 Under 30 Brazil list for his outstanding contributions to AI innovation.LinksConnect with Victor on LinkedInVictor's article on the four-layer frameworkArtefact websiteConnect with Genevieve on LinkedInBe among the first to hear about the release of each new podcast episode by signing up HERE

Topics covered in this episode: EuroPython 2026 videos are online The State of Django 2026: Boring is so back htmx 4.0.0 has been released

Welcome back to Season 13 of the PolicyViz Podcast! In this quick solo kickoff, I walk you through everything I have planned this season, from a full episode on accessibility in data communication to a brand-new weekly series where I talk through some of my favorite data visualizations from the last 15 years. I'll also tell you about my very first live podcast recording, happening in November in Boston as part of the IEEE VIS Conference and the new Cluster Practitioner Summit I'm co-organizing. Along the way I'll be expanding my YouTube content around Flourish, Excel, and PowerPoint, including how I'm using AI tools like Claude and Copilot in my own workflow. It's shaping up to be a busy, lucky Season 13, and I'm so glad you're along for the ride.Subscribe to this podcast. Follow this show and everything dataviz on Instagram, LinkedIn, Substack, X, and YouTube. Check out the PolicyViz websiteto learn more about data and data visualization.Questions? Comments? Pitch ideas? Email me at: jon@policyviz.comHosted by Zencastr where you can record, transcribe, edit, and automatically publish your meetings, podcasts, and more.

Avi Turetsky, Partner and Head of the Quantitative Research Group at Ares Management, and Bill Kieser, Principal and Co-Head of Research and Data Science at Ares Management, join the InsuranceAUM Podcast to explore how quantitative research is changing the way institutional investors evaluate private markets.   They discuss how investors can look beyond traditional measures like IRR and quartile rankings to better understand alpha, benchmarking, portfolio construction, and relative value across private equity, private credit, real estate, infrastructure, and secondaries. The conversation also examines what Ares' research suggests about investing during periods of volatility and the potential benefits of taking a more contrarian approach to private market allocations.   The discussion also looks ahead to the growing role of AI and machine learning in investment research, including how Ares is using data to identify potential credit risks and where quantitative tools may complement, rather than replace, human investment judgment.

Alex Cupps Interview | Fantasy Football Unlimited Podcast Alex Cupps is one of the newest rising voices in the fantasy football industry—and his journey is just getting started. On this episode of the Fantasy Football Unlimited Podcast, Alex joins Kevin Murray to share his story, from his early love of sports and fantasy football to earning his master's degree in Data Science, creating the CUPPS Model, winning Peter Overzet's "So You Think You Can Tout," and joining the team at Fantasy Life. Alex shares how fantasy football became the focus of his master's thesis at UC Riverside and eventually led to the creation of the Calculated Upside Player Prospecting System (CUPPS), a machine-learning model designed to identify NFL prospects with high fantasy football upside. He also takes us behind the scenes of his rapid rise as a fantasy football content creator, including competing against a field of 135 applicants in "So You Think You Can Tout," winning the competition, and the unforgettable moment Matthew Berry surprised him with the news that he would be joining Fantasy Life. Topics include:

Will artificial intelligence make statisticians obsolete—or make our expertise even more valuable? In this episode, I speak with independent statistical consultant Chris Harbron about how AI could reshape the role of statisticians in the pharmaceutical industry. Our conversation builds on Chris's paper, Will the Pharmaceutical Industry Need Statisticians in an AI World? We explore what happens when AI generates protocols, statistical analysis plans, programs, and reports. We also discuss why producing a plausible document differs from making sound decisions about clinical development. Chris and I examine the areas where human statisticians continue to add essential value—from navigating trade-offs and challenging assumptions to building trust and accepting accountability. We also share practical ways statisticians can begin working with AI while protecting quality, scientific rigor, and authenticity.

Much of the code that describes the inner workings of social networks, helps apps predict your musical interests, and underlies innovations in AI traces back to advances in mathematical theory and data science that fundamentally describe how massive networks behave.Jennifer Chayes helped create that field of math. She also spent 20 years at Microsoft building interdisciplinary research labs. Now, at UC Berkeley, her interests lie in using machine learning to work on topics like materials science, cancer immunotherapy, ethical decision-making, and climate change. She joins Flora for a wide-ranging discussion about math, networks, AI, academia, and her path between them. Guest:Dr. Jennifer Chayes is the dean of the College of Computing, Data Science, and Society at UC Berkeley.Transcript will be available after the show airs on sciencefriday.com. Subscribe to this podcast. Follow our show on Instagram, TikTok, Facebook, and Bluesky @scifri and sign up for our newsletters. Got a science question that's keeping you up at night? Call us: 877-472-4374 Hosted by Simplecast, an AdsWizz company. See pcm.adswizz.com for information about our collection and use of personal data for advertising.

How many cores does your machine have, 10, 18? Your async Python code uses just one of them. That isn't a bug in asyncio. That's the design, and optimizing event loops to be faster by 20% doesn't change it. So Giovanni Barillari started over. Joe is the creator of Granian, the Rust-based server that powers Talk Python. His new project is TonIO, an async runtime written from scratch for free-threaded Python. Real threads, a handful of primitives instead of asyncio's pile of them, and it flat out refuses to start if the GIL is on. Episode sponsors Sentry Error Monitoring, Code talkpython26 Python in Production Talk Python Courses Links from the show Guest Giovanni Barillari: github.com Granian: github.com Hyper: github.com Free threaded Python: docs.python.org Sort of: labs.quansight.org did a whole course: training.talkpython.fm uvloop: github.com rloop: github.com TonIO: github.com your EuroPython 2026 talk: www.youtube.com Michael's Cutting Python Web App Memory Over 31% Article: mkennedy.codes Watch this episode on YouTube: youtube.com Episode #561 deep-dive: talkpython.fm/561 Episode transcripts: talkpython.fm Theme Song: Developer Rap

In this episode of Software People Stories, Gayathri Kalyanaraman speaks with Vasu Vadlamani, a technology leader, architect, entrepreneur and Professor of Practice, about a career that has evolved alongside the software industry itself. About his accidental start to an AI Era tech leader who's defining mantra has been Every role has an expiration date! Vasu takes us back to the early 1990s, when software meant client-server systems, handwritten notes and face-to-face communication. From his early work with Citibank to working with Sun Microsystems, building a professional services organization, experiencing the disruption that followed Oracle's acquisition of Sun, and later taking on technology and architecture roles at HCL and Cognizant, his journey reflects several waves of technology transformation.But perhaps the most interesting transition came later.While working at Cognizant, Vasu found himself asking a fundamental question: Why should the gap between college education and industry readiness be so large?This led him to work with educational institutions with a very practical goal — to make graduates industry-ready much sooner. Eventually, that became the driving force behind his move from corporate technology leadership to becoming a Professor of Practice.Today, Vasu sees himself less as a software professional and more as someone whose purpose is to help people do their best — in technology, in their careers and in their interactions with others.The conversation explores:The early days of software and how dramatically the industry has evolvedWhy business understanding helped Vasu become a better software professionalBuilding and scaling a technology services organization from zero to 150 peopleNavigating the difficult period when the Sun Microsystems ecosystem disappearedMoving from customer-facing roles to architecting internal enterprise systemsThe philosophy of making software development “zero maintenance”Why the gap between education and industry continues to matterMaking graduates “production ready” rather than simply degree readyWhy professors also need exposure to real-world engineeringCreating safe, sandboxed opportunities for faculty and students to work on real enterprise use casesWhat AI means for technology careersWhy roles will change, but technology itself isn't going awayThe importance of resilience, curiosity, teamwork and being “an engineer at heart”At the heart of the conversation is a simple message from Vasu:Don't worry about the future of technology. Keep learning, keep adapting, and keep building. There will always be a place for people who are willing to do the work and find their way.Quotable Quotes“Don't worry about AI. Roles will change. The industry will still be there.”“If you fail, use your knowledge to get up and rise again faster.”“IT is a team's play. One person alone cannot run the ship.”“The younger generation will always be smarter than me. If I can succeed, Gen Z will definitely succeed.”“If Vasu can find a way, you are much smarter than me. You will find a better way.”Vasu Vadlamani is an accomplished IT Professional offering 30 years of experience in integrating Technical and Domain expertise (Financial Services), focusing on Large Complex Digital Transformation Initiatives. An efficient Communicator possessing strong Coordination, Interpersonal, Mentoring and Team Management Skills. Cross Functional Services Leader managing multi geo, multi service line deliveries for over $440+M. Vasu was  CTO of ADM Insurance Domain, responsible to create and execute an Information Technology Strategy to modernize enterprise technological applications, moving away from their legacy apps to modern On Cloud tech stack. Co-authored the (C)ontain, (M)aintain (I)nvest philosophy so significantly enhance the business value of applications.His  research interests span Artificial Intelligence & Machine Learning, Cloud & Distributed Systems, Data Science, Software Engineering practices, and CS Education & Pedagogy. He is  particularly interested in how emerging technologies reshape both industry workflows and the way Computer Science itself should be taught at the graduate and postgraduate level.Vasu can be reached at https://www.linkedin.com/in/vasuvadlamani/

The debate about whether AI will replace human workers has already been settled - not by academics or futurists, but by the organisations that fired their humans, discovered AI couldn't do what they needed, and quietly hired them back. The future isn't AI replacing humans. It's humans and AI working together in ways that neither could manage alone.In this episode, Victor Coimbra joins Dr Genevieve Hayes to share what hybrid agentic organisations actually look like in practice, and what data scientists need to do to position themselves at the centre of them.You'll discover:Why thinking of AI as a tool rather than a coworker is the mindset holding most organisations back [03:58]The four archetypes that determine which tasks belong to humans and which to agents [08:17]How AI is turning data scientists back into scientists [18:24]The two skills that will define an indispensable data scientist in a hybrid organisation [27:01]Guest BioVictor Coimbra is a Partner and CTO at Artefact, the world's largest pure-play AI consulting firm and co-founded the firm's Latin American operations. In 2024, he was recognised in the Forbes 30 Under 30 Brazil list for his outstanding contributions to AI innovation.LinksConnect with Victor on LinkedInArtefact websiteConnect with Genevieve on LinkedInBe among the first to hear about the release of each new podcast episode by signing up HERE

Topics covered in this episode: OpenAI's Python SDK has migrated to HTTPX2 TMOG - Native Task Manager for macOS, Windows, and Linux wrapture - one wrapper for mocking, tracing, and observability linkedin2md: turn your LinkedIn export into 40+ Markdown files Extras Joke Watch on YouTube About the show Sponsored by us! Support our work through: Our courses at Talk Python Consulting from Six Feet Up Connect with the hosts Michael: Mastodon / BlueSky / X / LinkedIn Calvin: Mastodon / BlueSky / X / LinkedIn Show: Mastodon / BlueSky / X Join us on YouTube at pythonbytes.fm/live to be part of the audience. Usually Tuesday at 7am PT. Older video versions available there too. Finally, if you want an artisanal digest of every week of the show notes in email form? Add your name and email to our friends of the show list, we'll never share it. Calvin #1: OpenAI's Python SDK has migrated to HTTPX2 The OpenAI Python SDK has migrated to HTTPX2, the Pydantic-stewarded fork of httpx. Pydantic picked it up citing "limited activity recently" in the original project, promising "a reliably maintained path forward." If you just use the default client, nothing to do. No code changes. The catch is TLS. Quoting the guide: HTTPX "previously verified certificates against the CA bundle provided by certifi. HTTPX2 instead uses the operating-system trust store, and the SDK no longer installs certifi." That "can break certificate verification in minimal container images without system CA certificates, environments using corporate TLS-inspecting proxies, and deployments that relied on a custom or modified certifi bundle." The fix is SSL_CERT_FILE or SSL_CERT_DIR, or pass your own ssl.SSLContext via verify. Deeper integrations need real edits: custom clients, auth handlers, hooks, and request mocking all take HTTPX2 objects now, and plain httpx is no longer pulled in transitively. So import httpx in your own code means declaring it yourself or moving over. Temporary escape hatch: a legacy HTTPX client Michael #2: TMOG - Native Task Manager for macOS, Windows, and Linux A native, deeply instrumented system monitor for macOS, Windows, and Linux, now in public beta - from Plummers' Software, i.e. Dave Plummer, who wrote the original Windows Task Manager and donated it to Microsoft in 1995. Wikipedia Three real native apps: Swift/AppKit on macOS, Win32 on Windows, C++/Qt 6 on Linux, with a shared C++ core keeping metric semantics aligned - no browser shell anywhere. One dense summary: CPU, clocks, thermals, GPU, memory, storage, network, energy, and the processes responsible for the load, all click-through. Per-core honesty: logical processor and NUMA views, P and E cores color-coded, optional kernel time, 60 FPS live meters. Memory with context: pressure, wired, compressed, cached, committed, available, and swap, plus configurable scrolling history. Processes that act like processes: tree view, filtering, sorting, follow mode, and native verbs including service and launchd control. Phosphor themes: light, dark, green, amber, blue, or mono, with color and saturation you tune yourself. Calvin #3: wrapture - one wrapper for mocking, tracing, and observability Graham Dumpleton, author of wrapt and the original New Relic Python agent, has released wrapture. The name is wrapt plus capture. The core idea: wrap real code instead of replacing it, so the real code still runs while you watch every call. Name a method with wrapture.binding(Class, "method"), open a timeline(), and you get a tape of what actually happened. Real return values, real nesting, arguments normalised against real signatures. tape.tree() prints the call graph as it ran. One mechanism, three jobs: monkey patching with a real lifecycle (apply, remove, suspend, plus returns, raises, transforms_args), unit testing that asserts on real call flow instead of a flat MagicMock call list, and ad-hoc tracing of a running app. The testing pitch is error paths. Inject TimeoutError at the payment gateway, then assert the ledger was never written. Stubs and mocks are strict and spec-required, and there is deliberately no bare Mock(). Tracing needs no code at all. A wrapture.toml naming targets and a sink, run with python -m wrapture main.py, and you get a live call tree with timings. It captures ordinary logging calls as nested events, and with the otel extra it exports spans, metrics and correlated logs with W3C trace ids that join across services. Every line of code and docs was AI-written under their direction, and they say so up front. Two weeks from first commit, eleventh alpha, over 1000 tests, 150+ pages of docs. Alpha on PyPI, needs Python 3.12+ and wrapt 2.4.0+. Michael #4: linkedin2md: turn your LinkedIn export into 40+ Markdown files Via Juan Manuel Daza - a Python CLI that unpacks LinkedIn's data-export ZIP into clean, per-category Markdown you can drop straight into an LLM. One command: linkedin2md Complete_LinkedInDataExport.zip, plus o for output dir, -lang en|es, and -pdf. 40+ output files: profile, experience, education, skills, connections, posts, comments, reactions, recommendations, endorsements, job applications, even ad targeting and LinkedIn's inferences about you. Built for LLM analysis: the README pitches NotebookLM, Claude Projects, Obsidian, and Ollama, with example prompts like "what patterns do you see in my career transitions?" PDF resume mode: -pdf renders an A4 CV via weasyprint, and degrades gracefully to Markdown-only if it isn't installed. Dependency note: "pure Python / zero-dep" holds for the Markdown path only - the PDF path needs weasyprint and markdown installed. Install: pipx install linkedin2md recommended, pip in a venv otherwise - 86% Python, 10 releases, v0.3.1 in May. Agentic dev angle: repo ships opencode config and an N3RV subagent pipeline, including a "judgment day" dual-model adversarial PR review. Extras Calvin: EVE Online Migrates to Python 3 Michael: Dinkus by Will McGugan Joke: Tao of Programming: Book 5 Maintenance

Dr. Lauren Ponisio is an Assistant Professor in the Department of Biology at the University of Oregon. Lauren's research revolves around preserving and restoring bee populations in agricultural areas and other natural habitats. She is interested in understanding the distribution and health of different populations of native bees. When she's not working, you can often find Lauren in her garden. She has been an avid gardener since childhood, and she currently has a thriving garden with lots of vegetables and plants to attract bees and other pollinators. She received her B.S. degree in biology with honors in ecology and evolution, as well as her M.S. degree in biology, from Stanford University. Lauren was awarded her Ph.D. from the Department of Environmental Science Policy and Management at the University of California, Berkeley. She conducted postdoctoral research at UC, Berkeley afterwards, and she served on the faculty University of California, Riverside before recently accepting her current position at the University of Oregon. Lauren received graduate fellowships from the National Science Foundation and the National Institute for Food and Agriculture, as well as a Postdoctoral Fellowship from the Berkeley Institute for Data Science. She was also named among the Global Food Initiative's "30 Under 30" in Food Systems in 2016. In our interview, Lauren shares more about her life and science.

Welcome to episode #1051 of Thinking With Mitch Joel (formerly Six Pixels of Separation). Dr. Sebastian Wernicke has built his career at the intersection of data science, artificial intelligence, economics, and decision-making. He leads the Data Science and AI team at Oxera, one of Europe's leading economics consultancies, and has become widely recognized for his ability to make complex ideas about data both accessible and deeply human. Many people first encountered Sebastian through his memorable TED Talk, which cleverly used data to satirize TED Talks themselves, but his work has always been driven by a much bigger question: how can organizations use data to challenge assumptions instead of reinforcing them? His new book, Data Inspired - Building An Organizational Culture Of Inquiry For Lasting Transformation, argues that businesses have become too focused on being "data-driven" and not focused enough on becoming "data-inspired." In this episode, Sebastian explains why more data doesn't automatically lead to better decisions, why confirmation bias often causes us to use data to strengthen existing beliefs, and why the greatest value of data lies not in optimization but in transformation. We also discuss the promises and disappointments of big data, the rise of generative AI, synthetic data, AI agents, and why the hardest challenge facing organizations isn't technological... it's cultural. Along the way, Sebastian makes a compelling case that leadership is ultimately about designing environments where people are encouraged to ask better questions, embrace uncertainty, and use evidence to change their minds rather than defend them. It is a thoughtful conversation about data, intelligence, curiosity, and why the organizations that thrive in the AI era will be the ones that build cultures of inquiry instead of cultures of certainty. Enjoy the conversation... Running time: 57:16. Hello from beautiful Montreal. Listen and subscribe over at Apple Podcasts. Listen and subscribe over at Spotify. Please visit and leave comments on the blog - Thinking With Mitch Joel. Feel free to connect to me directly on LinkedIn. Check out ThinkersOne. Here is my conversation with Dr. Sebastian Wernicke. Data Inspired - Building An Organizational Culture Of Inquiry For Lasting Transformation. Sebastian's viral TED Talk. Oxera. Follow Sebastian on LinkedIn. Chapters: (00:00) - Introduction to Sebastian Wernicke and Data Inspired. (02:52) - The Meta Nature of TED Talks and Storytelling. (06:12) - Confirmation Bias and Data Interpretation. (09:03) - The Limits of Data-Driven Decision Making. (12:10) - The Concept of Data Inspired. (14:58) - The Role of AI in Data Utilization. (18:07) - The Promise and Pitfalls of Big Data. (21:12) - Cultural Elements in Data Transformation. (24:02) - Trust in AI and Machine Learning. (26:52) - Human Nature and the Quest for Perfection. (29:30) - The Balance of Autonomy and Control in AI. (30:30) - The Rise of Synthetic Data. (32:00) - Skepticism Towards Synthetic Data. (34:54) - Exploring Personas and Assumptions. (37:40) - The Role of Data in Decision Making. (39:27) - The Complexity of Uncertainty in Business. (41:52) - Rethinking Intelligence and Creativity. (46:54) - Building Beautiful Questions with Data. (51:30) - Navigating the Data Dilemma. (54:46) - Cultural Attitudes Towards Privacy and Surveillance.

What are the key characteristics of complex systems, and what are practical patterns for tackling complex coding problems? Christopher Trudeau is back on the show this week with another batch of PyCoder's Weekly articles and projects.

In 2020, a gastroenterologist in Glasgow did the math on his new research study and came up with 30,000 samples, arriving over two years from three cities and a dozen hospitals. He asked around about how researchers keep track of that. The answer was Microsoft Excel. Shaun Chuah had written some HTML by hand in Notepad back in high school and that was about the whole of his programming experience, so he opened the Django tutorial and started reading. Six years later that app is Foundry120, holding 10 terabytes of clinical and genomics data with an agentic AI running on top of it. Episode sponsors Sentry Error Monitoring, Code talkpython26 Talk Python Courses Talk Python Courses Links from the show Guest Shaun Chuah: github.com Up and Running with Rust Course: talkpython.fm Foundry120: www.foundry120.com Designing Data Intensive Applications: www.oreilly.com Microsoft Foundry: ai.azure.com ChatIBD: www.chatibd.com Blog: shaunchuah.github.io @drshaunchuah: x.com github.com/shaunchuah: github.com Watch this episode on YouTube: youtube.com Episode #560 deep-dive: talkpython.fm/560 Episode transcripts: talkpython.fm Theme Song: Developer Rap

Topics covered in this episode: Web UIs for your reverse proxy Wagtail 8.0 is hot off the presses RISC-V is now officially supported by CPython Django's annual releases make every version an LTS Extras Joke Watch on YouTube About the show Sponsored by Logfire from Pydantic: pythonbytes.fm/logfire Connect with the hosts Michael: Mastodon / BlueSky / X / LinkedIn Calvin: Mastodon / BlueSky / X / LinkedIn Show: Mastodon / BlueSky / X Join us on YouTube at pythonbytes.fm/live to be part of the audience. Usually Tuesday at 7am PT. Older video versions available there too. Finally, if you want an artisanal, hand-crafted digest of every week of the show notes in email form? Add your name and email to our friends of the show list, we'll never share it. Michael #1: Web UIs for your reverse proxy Traefik, nginx, and Caddy all sit in front of a lot of self-hosted infrastructure, and all three are configured by hand-editing files. Three active projects put a control plane on top: Traefik Manager (Python + Flask), Nginx UI (Go + Vue), and caddy/ui (React + Node). All three are additive rather than replacements - none of them take ownership of your config away from you - which is the part that matters when the thing has write access to production routing. Traefik Manager is the Python one: Flask 3.1 and Gunicorn for the control plane, a lightweight Go agent for remote instances, currently v1.10.0 with an Android companion app. Nginx UI is a single Go binary at 11.3k stars, with a block-style config editor, an Ace editor doing LLM completion on nginx syntax, and an MCP server so agents can drive it. caddy/ui runs as two containers next to your existing Caddy, reads and writes your Caddyfile directly, and uses Caddy's /adapt API to validate before reload - no Docker socket required. Each one edits the config the underlying server already reads, so your files stay the source of truth and you can drop the UI without unwinding anything. Undo is a first-class feature across all three - timestamped backups with optional Git history, config version compare and restore, Caddyfile snapshots with one-click rollback. Observability is where they diverge: Traefik Manager does CrowdSec and a visual route map, Nginx UI does server metrics, caddy/ui streams access logs over SSE and pulls p50/p95/p99 off Caddy's Prometheus endpoint. Maturity spread is wide - Nginx UI has 11.3k stars, caddy/ui has 4 and was built in a single Claude session - and caddy/ui ships with auth off by default, so set CADDY_UI_USER and JWT_SECRET before it goes anywhere near a public interface. Calvin #2: Wagtail 8.0 is hot off the presses Link: https://github.com/wagtail/wagtail/releases/tag/v8.0 Custom base page models are now supported, so projects aren't locked into subclassing Wagtail's Page as shipped (Matt Westcott). New v3 REST API handles both read and write CMS operations, a first for Wagtail's API. A global registry for permission policies, plus full customizability for the remaining page views via PageViewSet. AVIF and WebP images are no longer auto-converted to PNG by default, a real behavior change to watch on upgrade. Five security fixes: page admin API restrictions, document identification by SHA1 hash, descendant collections in the Documents/Images API, snippet copy permissions, and the page translation endpoint. Formalized Django 6.1 support, and CI now runs on uv with a lockfile. Sponsor: Logfire from Pydantic Your AI agent failed at 2am. Was it the model? A tool call? The database? Most observability tools can't tell you, because they only see part of your stack. Pydantic Logfire sees all of it. One trace across your agents, LLMs, APIs, and database. Down to the infrastructure: services, Kubernetes, and hosts. It's built on OpenTelemetry, with SDKs for Python, TypeScript, and Rust, and it works with any OTel-compatible language. Every prompt, token count, and cost, right next to your vector searches and API calls. You query everything with Postgres-compatible SQL. And so can your coding agent, through the Logfire MCP server. Stop guessing. Read the trace. Pydantic Logfire. AI, it's still just engineering. Visit pythonbytes.fm/logfire today and sign up today. Get 10M records free every month, no card required. You can even click “Onboard with your coding agent” to copy a prompt to have claude or codex integrate Logfire into your app. Thanks to Pydantic for supporting the show. Calvin #3: RISC-V is now officially supported by CPython Link: https://blog.python.org/2026/08/riscv-now-officially-supported/ CPython added RISC-V as a tier 3 platform under PEP 11, specifically the 64-bit Linux target riscv64-unknown-linux-gnu. RISC-V is an open ISA anyone can implement, unlike x86 and ARM, and its market is projected to quadruple by 2032. The RISE Project donated real RISC-V machines for buildbots; the author's work was funded by a Sovereign Tech Agency fellowship. What changes: the port is now a maintained compatibility target, so CPython changes are less likely to quietly break it. What doesn't: no python.org installers, no binary wheel parity for native extensions. Next up: RISC-V runners in CPython CI for pre-merge feedback, then a push toward tier 2, plus architecture-specific optimizations. The ask is testing. If you have RISC-V hardware, build CPython, run your test suite, file what breaks. Tier 3 is the weakest support tier. PEP 11 tier 3 requires a core developer contact and a buildbot, but failures on tier 3 platforms explicitly do not block a release. Saying "ongoing CI/testing expectations" oversells it. The honest bit is "someone is now on the hook for it, and breakage gets noticed," not "it's guaranteed working." Worth the caveat that this is Linux SBCs, not microcontrollers. A VisionFive 2 counts, an ESP32-C6 or Pico 2 does not. Those are 32-bit non-Linux parts where MicroPython is still the answer. Michael #4: Django's annual releases make every version an LTS Starting with Django 2028, Django will move to one January feature release per year, adopt calendar-based version numbers, and support every release for three years. The old distinction between standard and LTS releases disappears, giving teams a predictable annual upgrade path that aligns more closely with Python's own release and support cadence. Every Django release becomes the safe, long-supported choice, so teams no longer need to wait for a specially designated LTS version or absorb two years of changes at once. Each release gets one year of mainstream bug fixes followed by two years of security and data-loss fixes. New releases support the three latest Python versions and add the next Python release during their first year. Calendar versioning begins with Django 2028, followed by Django 2029 and so on. Three Django versions will be supported at any time, giving third-party packages a clearer rolling target. Nothing changes before 2028, and existing commitments for Django 5.2 LTS and 6.2 LTS remain in place. Extras Calvin: The Python docs now document the time complexity of built-in types https://docs.python.org/3.16/library/time-complexity.html Thinking in Python - Bruce Eckel's free book https://thinkinginpython.com/ Michael: prune_uv_pythons.py - Prune uv-managed Python installs, keeping only the newest patch per minor version Runs automatically in my system “upgrade” script: upgrade-output-2026.png Started using Ollama cloud models for my Hermes assistant. Thanks to Jeff Triplett I learned they are not just local models. Joke: The Tao of Programming - Book Seven: Corporate Wisdom

AI hallucinations get all the attention. But hallucinations are relatively easy to catch because the output is obviously wrong. The failure mode that should worry data scientists more is when the agent uses facts that are true to draw conclusions that are false, producing outputs that look perfectly fine. This is known as silent correctness.In this Value Boost episode, Jia Huang joins Dr Genevieve Hayes to explore why silent correctness is the most dangerous failure mode in agentic AI systems and what data scientists can do to catch it before it causes serious harm.You'll discover:Why silent correctness is harder to catch than a hallucination [04:17]Why sampling and auditing are non-negotiable in agentic systems [07:09]Four techniques data scientists can use to catch silent failures [09:28]The one safeguard every agentic AI system should have [11:10]Guest BioJia Huang is a lead research engineer at A*STAR, Singapore's Agency for Science, Technology and Research, and is the author of multiple books on AI engineering and agent design, including Designing AI Agents and RAG from First Principles. His work focuses on turning agentic AI from impressive demos into reliable, auditable, and value-producing engineering systems.LinksConnect with Jia on LinkedInFollow Jia on SubstackAgent Design Pattern Society (ADPS) websiteJia's AI agent design position paperConnect with Genevieve on LinkedInBe among the first to hear about the release of each new podcast episode by signing up HERE

Dr. Hooman Rashidi is Associate Dean of AI in Medicine at the University of Pittsburgh and Executive Director of CPACE. Before Pitt, he founded Cleveland Clinic's Center for AI and Data Science and directed AI at UC Davis Medical Center. Ed asks him whether medical staff are prepared for AI. His answer is that even the ones who feel ready are not, because people bundle every kind of AI together and the newer generative tools don't behave like the predictive models the field has been using for decades. In this episode of DGTL Voices, he explains what's missing from tumor boards, why he leads with education before deployment, how Pitt runs a hybrid strategy rather than committing to vendors alone, and why his center puts junior contributors on patent filings when most institutions don't. https://bio.marxadvisory.com/

Your site is down. It's 3am. Is it a bug, a bill, or a breach? You can't tell yet, and everyone is watching you find out. Matt Lea has spent fifteen years being the person companies call when an outage is costing them real money per hour, and his whole argument is that everything you'd want in that moment gets decided months earlier, on ordinary afternoons, when someone chose the convenient thing. We walk his top twelve dos and don'ts in AWS - infrastructure as code, IAM roles instead of access keys, private subnets, no wildcards, no public buckets - and I push on which of them actually matter if you're one person on a small VPS. Then we get to Cloud War Games, where Matt breaks things on purpose so your team's first real incident isn't their first incident. Let's get into it. Episode sponsors Sentry Error Monitoring, Code talkpython26 Talk Python Courses Talk Python Courses Links from the show Guest Matt Lea: linkedin.com Talk Python Certificates: training.talkpython.fm/certificates Schematical: schematical.com CloudWarGames.com: cloudwargames.com Zero to Hero on AWS Security: www.oreilly.com Repo: github.com Custom Wheel Offset: customwheeloffset.com 2012 TechCrunch Disrupt Hackathon: techcrunch.com tech comics: schematical.com shhgit: github.com Zero Trust in 200ms: Implementing Identity-Per-Transaction: us.pycon.org Coolify: coolify.io returned to full GA Nov 2025: aws.amazon.com Signed URLs/cookies: docs.aws.amazon.com Cloudflare: www.cloudflare.com Bunny Shield: bunny.net Cloud War Games One: www.youtube.com Cloud War Games Two: www.youtube.com LinkedIn: linkedin.com YouTube: youtube.com KnocKnoc: knocknoc.io Watch this episode on YouTube: youtube.com Episode #559 deep-dive: talkpython.fm/559 Episode transcripts: talkpython.fm Theme Song: Developer Rap

Topics covered in this episode: Python 3.12.14, 3.11.16, 3.10.21 - security releases Codeberg's AI-code ban tests its role as a GitHub alternative Brett Cannon: what's missing for reproducible builds on PyPI nothing records the source code a distribution came from. direct_url.json captures it when you install from a repo or archive, so the fix is putting the same info in sdist/wheel metadata. recording the build tools. Wheels can already do this via PEP 770 SBOMs in .dist-info/sboms/ - sdists can't, since they're a tarball plus a precalculated PKG-INFO with nowhere to hang extra metadata. Either "don't use sdists" or an sdist v2. Extra extra extra, hear all about it Extras Joke Watch on YouTube Sponsored by Logfire from Pydantic pythonbytes.fm/logfire This episode is brought to you by Pydantic Logfire. It's observability for AI apps from the team behind Pydantic - agents, LLMs, APIs, database, and infrastructure in a single trace, queried with Postgres-compatible SQL. Your coding agent can query it too, through their MCP server. I'll tell you more later. Connect with the hosts Michael: Mastodon / BlueSky / X / LinkedIn Calvin: Mastodon / BlueSky / X / LinkedIn Show: Mastodon / BlueSky / X Join us on YouTube at pythonbytes.fm/live to be part of the audience. Usually Tuesday at 7am PT. Older video versions available there too. Finally, if you want an artisanal, hand-crafted digest of every week of the show notes in email form? Add your name and email to our friends of the show list, we'll never share it. Calvin #1: Python 3.12.14, 3.11.16, 3.10.21 - security releases https://blog.python.org/2026/08/python-31214-31116-31021/ Source-only security releases for the three branches now in security-fix-only mode; release team blamed the European solar eclipse for the timing. tarfile hardening. Multiple path-traversal bypasses of the data filter closed, including a symlink escape that bypassed the CVE-2025-4330 fix; extract() now applies the filter to link targets too. Four fresh CVEs: CVE-2026-2297 (SourcelessFileLoader not using io.open_code() for .pyc), CVE-2026-4224 (expat crash on deeply nested content models), CVE-2026-3644 (control chars in http.cookies.Morsel), plus the completed CVE-2021-4189 fix in ftplib.ftpcp. Quadratic-complexity DoS cleanup across the stdlib: HTMLParser, configparser regexes, unicodedata.normalize(), csv.Sniffer.sniff(), and ElementTree XPath index predicates. Header/injection fixes: CR/LF rejected in HTTPConnection.set_tunnel(), control chars blocked in wsgiref.handlers status, and webbrowser now rejects leading dashes (plus a %action prefix bypass). http.client now caps chunked trailer lines and 1xx interim responses at 100 each - a hostile server could previously hang the client forever despite a socket timeout. Memory-safety odds and ends: stale pointers in lzma/bz2/zlib decompressors after MemoryError, a bz2 stack overflow on reuse-after-error, and bundled libexpat bumped to 2.8.3. If you're still on 3.10, 3.11, or 3.12 - and you extract tarballs from anywhere you don't fully control - this one's not optional. Michael #2: Codeberg's AI-code ban tests its role as a GitHub alternative Armin's article “Codeberg Divides” Armin Ronacher argues that Codeberg's new terms, which prohibit projects mostly written with generative AI, create a vague and difficult-to-enforce boundary. His larger concern is that a democratically governed host can still be unpredictable or ideologically narrow, weakening Codeberg's potential as a broad European alternative to GitHub. The strongest question for Python developers is whether repository hosting should judge legal open source by how code was produced, or focus on behavior and resource abuse. “Mostly generated” is hard to measure in modern codebases where developers mix handwritten code, completions, agents, and generated refactors. Ronacher suggests clearer alternatives: ban all LLM involvement, or target autonomous repository spam, abusive resource use, and low-quality generated contributions directly. Codeberg is free to choose a values-driven community, but that may conflict with being predictable, neutral infrastructure and a serious GitHub competitor. Worth discussing: can open-source communities set meaningful AI boundaries without driving maintainers and projects into opposing camps? Very first search for these terms lands on this page. Codeberg looked like a viable alternative. … Unfortunately, the latest update to its terms of service seems to mark a first step in changing one part I moved there for, namely the “freedom” part. Sponsor: Logfire from Pydantic Your AI agent failed at 2am. Was it the model? A tool call? The database? Most observability tools can't tell you, because they only see part of your stack. Pydantic Logfire sees all of it. One trace across your agents, LLMs, APIs, and database. Down to the infrastructure: services, Kubernetes, and hosts. It's built on OpenTelemetry, with SDKs for Python, TypeScript, and Rust, and it works with any OTel-compatible language. Every prompt, token count, and cost, right next to your vector searches and API calls. You query everything with Postgres-compatible SQL. And so can your coding agent, through the Logfire MCP server. Stop guessing. Read the trace. Pydantic Logfire. AI, it's still just engineering. Visit pythonbytes.fm/logfire today and sign up today. Get 10M records free every month, no card required. You can even click “Onboard with your coding agent” to copy a prompt to have claude or codex integrate Logfire into your app. Thanks to Pydantic for supporting the show. Calvin #3: Brett Cannon: what's missing for reproducible builds on PyPI Framing came out of his 2026 Python Packaging Council nomination - the secure-supply-chain gap he found is that Python has no defined way to do reproducible builds at all. Design goal is zero friction: producers uploading to PyPI shouldn't have to do anything. The work lands on build backends and installers. Gap #1: nothing records the source code a distribution came from. direct_url.json captures it when you install from a repo or archive, so the fix is putting the same info in sdist/wheel metadata. Gap #2: recording the build tools. Wheels can already do this via PEP 770 SBOMs in .dist-info/sboms/ - sdists can't, since they're a tarball plus a precalculated PKG-INFO with nowhere to hang extra metadata. Either "don't use sdists" or an sdist v2. The replay mechanism already exists: [build-system] in pyproject.toml is a defined entry point, so if backends recorded their own environment, you could reinstall and re-run the build. Payoff idea: trusted third parties report successful reproductions back to PyPI, which displays "independently reproduced by X" - surfaced in the index API so installers could prefer reproduced files. Explicitly framed as a perk, not a requirement - roughly SLSA build level 1, no shaming projects that don't opt in. Verbal kicker option: "And don't think pure-Python wheels are off the hook. Something built that wheel, and if that something was compromised, so is your wheel. SolarWinds was a build-process attack." Michael #4: Extra extra extra, hear all about it Python 3.14.7 Upgraded the MCP servers to 2026-07-28 v2 protocols (talk python, python bytes) Got agentsview running synced via postgres Talk Python courses, teams trial offering Talk Python courses, government procurement offering Lean TDD audio book is out Extras Calvin: uv now prefers post-quantum key exchange - https://github.com/astral-sh/uv/releases/tag/0.12.4 Joke: Beware of dog

Steven Song is the founder and CEO of Diald, an AI-powered decision intelligence platform for commercial real estate. Before founding Diald, Steven worked on both the investing and development sides of real estate, and built his career across architecture and urban planning, training at Carnegie Mellon and the University of Pennsylvania. He was a founding principal at SCAAA, a global strategy, planning, and design firm, and is a partner at Axle Companies, a family office focused on real estate investment and social impact ventures. Steven is based in Los Angeles.(02:26) Why CRE Decisions Are Still Judgment-Driven (04:41) The Signal That Killed an Atlantic City Deal (07:44) Contextual Drift: The Risk Nobody Models(10:20) Diald's approach (11:59) AI Token Costs and Asking Better Questions (14:27) Tools vs. Workflows (15:52) Diald's Underwriting (17:58) Killing Bad Deals Earlier (19:18) How AI Upgrades the Analyst Role (20:44) Where General Purpose AI Fails at Underwriting (24:23) Does AI Make CRE More Efficient or More Competitive (26:02) What Underwriting Looks Like in 5 Years (27:14) Where Human Judgment Still Matters (29:06) The Local Signals Investors Miss (31:15) Collaboration Superpower: Denise Scott Brown and Reyner Banham

0:30 - Mike Koolidge filling in for Dan 9:11 - DSA 27:59 - Can the Republicans Beat Socialism in the Midterms? 46:10 - WRFH 101.7 FM General Manager and Hillsdale College journalism lecturer Scot Bertram gives his on-the-ground view of the Michigan Senate race. Follow Scot on X @ScotBertram 01:00:19 - The Heritage Foundation's Steven Bucci dismisses the alleged complaints from the USS Lincoln: "It's a shame because they are, frankly, whining." 01:15:50 - More reaction to the USS Lincoln 01:31:34 - Judge Glock, director of research at the Manhattan Institute, makes the case for data centers. Follow Judge on X @judgeglock 01:49:50 - Professor of the Practice of Data Science at Washington University, Liberty Vittert Capito, analyzes Why nutty candidates are underperforming their polling so badly. Follow Liberty on X @libertycapitoSee omnystudio.com/listener for privacy information.

Topics covered in this episode: Claude Code /insights Post-quantum crypto lands in Python MCP goes stateless — and FastMCP gets renamed inshellisense - IDE style command line auto complete Extras Joke Watch on YouTube About the show Sponsored by Xweather Xweather combines enterprise-grade weather intelligence with agent-ready APIs, natural language capabilities, and an MCP server so your agents can adapt workflows, automate responses, and make better decisions based on real-world conditions. Michael will tell you more about them later in the show. Get started for free at pythonbytes.fm/xweather Connect with the hosts Michael: Mastodon / BlueSky / X / LinkedIn Calvin: Mastodon / BlueSky / X / LinkedIn Show: Mastodon / BlueSky / X Join us on YouTube at pythonbytes.fm/live to be part of the audience. Usually Tuesday at 7am PT. Older video versions available there too. Finally, if you want an artisanal digest of every week of the show notes in email form? Add your name and email to our friends of the show list, we'll never share it. Michael #1: Claude Code /insights Michael's Insights: michael-kennedy-claude-code-insights-2026-08-09.html Be careful sharing these outputs, they include details references to your projects, errors, security findings, etc. ;) /insights reads your last 30 days of local session transcripts and hands back an interactive HTML report on how you actually work. One command, zero setup: type /insights in a session, or run claude -p "/insights" from the shell for a non-interactive version that just prints the path Reads what's already on disk: pulls session logs from ~/.claude/projects/, skipping agent sub-sessions and anything under 2 messages or 1 minute Project areas: clusters your sessions into themes like "CLI Tooling" or "Documentation" with session counts Friction analysis: categorizes where things went wrong by root cause - and quotes your own prompts back at you Interaction style: tells you whether you're a delegator or a micromanager, plus which workflows are worth doubling down on Actually actionable: suggests concrete CLAUDE.md additions and Claude Code features you're not using The catch: Haiku does the per-session classification, so the first run takes several minutes; results cache to ~/.claude/usage-data/facets/ and the report lands at ~/.claude/usage-data/report.html Calvin #2: Post-quantum crypto lands in Python pyca/cryptography 48 ships ML-KEM (key establishment) and ML-DSA (signatures) — NIST's post-quantum standards, now one pip install away. Big deal because it's the 11th most-downloaded package on PyPI (~1.2B downloads/month) and sits under Ansible, Certbot, Airflow, and paramiko. No PQ there, no PQ anywhere in Python. Trail of Bits did the work (Rust bindings, cross-backend API, tests, AWS-LC backend support), funded by the Sovereign Tech Agency. Timing tracks a June 22 White House order setting federal deadlines: PQ key establishment by end of 2030, PQ signatures by end of 2031. Not a drop-in swap — the wire sizes explode. ML-DSA-65 signatures are 3,309 bytes vs Ed25519's 64; ML-KEM-768 public keys are 1,184 bytes vs X25519's 32. Hardcoded field sizes and length prefixes will bite. API looks like the existing asymmetric primitives, except ML-KEM is encapsulate/decapsulate rather than a Diffie-Hellman exchange. SLH-DSA (the hash-based conservative backstop) is still in progress. The primitives are here, but protocols haven't caught up — so you won't be running post-quantum Certbot this week. Sponsor: Xweather You're using agents that can write code, summarize documents, and automate workflows. But they're missing one thing: awareness of the world around them. This is where today's sponsor, Xweather comes in. Xweather combines enterprise-grade weather intelligence with agent-ready APIs, natural language capabilities, and an MCP server built for tools like Claude, Codex, Copilot, and modern IDEs – so your agents can adapt workflows, automate responses, and make better decisions based on real-world conditions. Backed by Vaisala, whose instruments fly on NASA missions to Mars, Xweather delivers trusted data and unique insights that go beyond conditions to actual impact – from real-time lightning strikes to road surface forecasts. Start with 15,000 free API calls each month and pay only for what you use as you grow. Xweather is your full weather stack, for developers by developers. Start building for free today at pythonbytes.fm/xweather. The link is in your podcast player's show notes and on the episode page. Thanks so much to Xweather for supporting Python Bytes. Calvin #3: MCP goes stateless — and FastMCP gets renamed From Philipp Acsany over at Real Python The 2026-07-28 spec landed July 28 and the Python SDK shipped 2.0.0 the same day. Biggest rewrite since MCP launched, and it's breaking on purpose. Context for scale: the Tier 1 SDKs are pulling close to half a billion downloads a month, with TypeScript and Python each past a billion total. The headline is the stateless core. The initialize/initialized handshake and the Mcp-Session-Id header are both retired — protocol version, client identity, and capabilities now ride in _meta on every request, with an optional server/discover RPC if a client wants capabilities up front. Any request can land on any instance behind plain round-robin, no shared storage. Server-initiated calls are the hard part of the migration. Sampling, elicitation, and roots/list no longer call back to the client; instead the server returns resultType: "input_required" and the client retries with inputResponses attached. Multi Round-Trip Requests, MRTR. Also: Mcp-Method and Mcp-Name are now required headers so gateways route on headers instead of cracking JSON bodies, and missing-resource errors move to standard 32602. Deprecation sweep with an actual policy behind it — Roots, Sampling, Logging, and the legacy HTTP+SSE transport all deprecated with a twelve-month minimum offramp. Tasks graduated out of the experimental core into a real extension, which is what the formalized extensions framework was for. MCP Apps is now an official extension too, so a tool call can return sandboxed interactive HTML. Auth picked up RFC 9207 issuer validation, issuer-bound credentials, and a shift from DCR toward CIMD. Python SDK 2.0 is where it gets personal: FastMCP is now MCPServer, no alias, no shim. McpError → MCPError. Wire types went snake_case (is_error, input_schema) and moved to a standalone mcp_types package, with mcp.types kept as a permanent alias. One Client object replaces the old transport + ClientSession + initialize() stack. httpx became httpx2. Sync handlers run on worker threads now, so asyncio.get_running_loop() raises inside them. The good news: one MCPServer serves both protocol eras, so 2025-era clients keep working with nothing to configure, and a Resolve(fn) parameter lets one tool body cover MRTR and the old path. 1.x is maintenance-and-security-fixes only — pin mcp>=1.28,

Every company has one. The little internal tool that Jane built back in 2021, and then Jane left. Nobody understands it, nobody will touch it. There are two unwritten rules around it: don't change it, it's working. And if you break it, you bought it. That's dark-matter enterprise software. For every app you can actually see, there are ten of these sitting in the shadows, frozen. Michael Booth thinks that just changed. He read my article on hyper-personal software and ran with it, writing about hyper-team software: small teams inside big companies finally building the tools that were never going to get built. We cover where this works, where it quietly goes wrong, and the guardrails that keep it from turning into a mess. Let's get into it. Episode sponsors Sentry Error Monitoring, Code talkpython26 Python in Production Talk Python Courses Links from the show Guest Michael Booth: github.com Talk Python AI Integrations: talkpython.fm/blog From Hyper-Personal to Hyper-Team Software: Small Team-Built, AI-Assisted Tools Inside the Enterprise: www.databooth.com.au What hyper-personal software looks like (MK's article): mkennedy.codes Databooth Site: www.databooth.com.au Wall Street just lost $285 billion because of 13 markdown files: martinalderson.com SaaSpocalypse is real but everyone is panicking about the wrong thing: www.reddit.com Warp Terminal: www.warp.dev Watch this episode on YouTube: youtube.com Episode #558 deep-dive: talkpython.fm/558 Episode transcripts: talkpython.fm Theme Song: Developer Rap

On this Summer Friday, we've put together some of our favorite recent interviews, including: Each year WNYC hosts a "health convening," with support from the Alfred P. Sloan Foundation, as an opportunity for healthcare experts and practitioners to inform WNYC's health reporting. This year, the topic is ultra-processed foods and how they affect our health.  First, Kevin Hall, Ph.D., former senior investigator at the National Institutes of Health (NIH) , discusses his groundbreaking, tightly controlled metabolic ward trials and mathematical models tracking how human bodies respond to ultra-processed foods. Then, Fang Fang Zhang, M.D., Ph.D., cancer epidemiologist and chair of the Division of Nutrition Epidemiology and Data Science at the Friedman School of Nutrition Science and Policy at Tufts University, discusses her population‑based research on how ultra-processed foods influence cancer prevention, cancer survivorship and long‑term health outcomes. Then, David Kessler, MD, former FDA Commissioner, breaks down a citizen petition he filed with the Food and Drug Administration last summer which would, among other things, declare that some of the core ingredients in ultra-processed foods were no longer “generally recognized as safe (GRAS)” -- a classification  that means ingredients are exempt from strict premarket approval process. Plus, Ilan Stavans, publisher of Restless Books, professor of humanities and Latin American and Latino culture at Amherst College and the author of A Nation Wrestles with God: American Prophets, Philosophers, and Firebrands (Restless Books, 2026), talks about the broad spectrum of American religious thought, from Cotton Mather to Lana Del Rey.   These interviews were lightly polished up and edited for time, the original versions are available here: What Are Ultra-Processed Foods and What Are They Doing to Us? (June 16, 2026) Ultra-Processed Foods and Colon Cancer (June 16, 2026) Ultra-Processed Foods Policy Proposals (June 16, 2026) American Believers (July 9, 2026) photo: Potato chips and other junk food in Walmart, Wenatchee Washington (Thayne Tuason, CC BY-SA 4.0, via Wikimedia Commons) Hosted by Simplecast, an AdsWizz company. See pcm.adswizz.com for information about our collection and use of personal data for advertising.