- The OKR TRAP Most Companies Fall Into 🔗 talk notes (Dan North) [Company Culture, Management, Strategy, leadership] [Duration: 00:08] (⭐⭐⭐⭐⭐) Dan North traces OKRs back to Drucker's management by objectives and to Andy Grove's rework of them at Intel, then shows the three places organizations break them: defining objectives as activities, tracking work done instead of distance to the goal, and scoring key results at the end of the quarter as if they had been targets.
- 34. Pensar en sistemas - Donella Meadows - Libros en 20 minutos 🔗 talk notes (Andrés Aguilar, Donella Meadows) [Mental models, Strategy, Systems Thinking] [Duration: 00:21] (⭐⭐⭐⭐⭐) Recorrido por Pensar en sistemas, el libro póstumo de Donella Meadows: qué es un sistema, cómo los bucles de refuerzo y de equilibrio con sus demoras producen comportamientos que nadie quiso, y por qué los puntos de apalancamiento que de verdad cambian algo son los menos obvios.
- The CTO's Guide to Fixing Slow Delivery 🔗 talk notes (Dave Farley) [Continuous Delivery, Feedback cycles, Management, Nature of Software Development, Small Safe Steps (3s)] [Duration: 00:13] (⭐⭐⭐⭐⭐) Every obvious lever a technology leader reaches for when delivery feels slow — more people, more pressure, more checkpoints — tends to make it slower, because software development is discovery rather than production. Dave Farley makes the case for optimising the speed of learning instead, and letting delivery speed fall out of it.
- Residues: Time, Change & Uncertainty in Software Architecture 🔗 talk notes (Barry M. O'Reilly) [Architecture, Evolutionary Architecture, Mental models, Resilience, Systems Thinking] [Duration: 00:46] Randomly stressing an architecture produces designs that are more likely to survive events nobody predicted than requirements engineering or risk management manage. Barry M. O'Reilly explains why through complexity science — Kauffman networks, attractors and criticality — and walks through the stressor analysis, the incidence matrix and the PhD experiment built to falsify the whole thing.
- Building Software That Survives 🔗 talk notes (Michael T. Nygard, Charles Humble) [Architecture, Engineering Culture, Legacy code, leadership, team topologies] [Duration: 00:38] From the 3 a.m. outages that produced Release It to running architecture enablement at Nubank, Michael T. Nygard talks with Charles Humble about autonomy as something you grant per activity rather than wholesale, about Conway's law being a law about communication and not org charts, and about how much a leader changes by what they celebrate rather than by reorganising.
- Product Thinking on Data and Platform Teams: Is It Worth It? 🔗 talk notes (Jacqueline Lee) [Data Engineering, Legacy code, Observability, Platform engineering, Product] [Duration: 00:41] Product thinking was built for external customers, and porting it to internal data and platform work breaks in specific ways. Jacqueline Lee contrasts a greenfield bank for underbanked Canadians with getting 37 teams off a legacy table at Yelp, and lands on a developer-friendly definition: understanding the user, the business and the tech, and writing code that lets you test, learn and adapt fast.
- Thinking in Product Outcomes | The Shift Every Engineer Needs to Make 🔗 talk notes (Peppe Silletti) [Feedback cycles, Lean Product Management, Mental models, Product, Product Engineer] [Duration: 00:25] Business outcomes are lagging indicators an engineering team cannot move directly; product outcomes are leading indicators tied to user behaviour that it can. Peppe Silletti works through the difference, the four tests of a good product outcome, and what an engineer can actually do to stop taking orders — with the key concepts mirrored in engineering analogies.
- The creator of OpenClaw: "I ship code I don't read" 🔗 talk notes (Peter Steinberger, Gergely Orosz) [AI Assisted Engineering, Developer Productivity, Engineering Career, Generative AI, Technical Practices] [Duration: 01:54] Peter Steinberger spent thirteen years building PSPDFKit, burned out, sold his shares and disappeared for three years — then came back straight into Claude Code having missed the years everyone spent deciding AI was useless. He explains to Gergely Orosz why he no longer reads most of the code he ships, why closing the loop is the whole secret, and why he now calls pull requests prompt requests.
- Beyond the Warehouse: Why BigQuery Alone Won't Solve Your Data Problems 🔗 talk notes (Sarah Usher) [Architecture, Big Data, Cloud, Data Engineering, Software Design] [Duration: 00:44] Scale-ups buy the idea that one tool can do everything and make the warehouse that tool, which works beautifully until latency forces the first team to bypass it — and then everyone copies that bypass. Sarah Usher redraws the architecture around a raw / curated / use case lifecycle, argues that a single source of truth is a place you design rather than a system you own, and makes one request she repeats twice: store your raw data.
- B2B Startup Metrics | Startup School 🔗 talk notes (Tom Blomfield) [Business models, Lean Startup, Management, Product, startup] [Duration: 00:23] Launching without metrics is flying an aircraft with no instruments, but a dashboard of 500 metrics at a few thousand users is just as useless. Tom Blomfield picks the handful that matter for a B2B startup — revenue, burn, runway, then retention, net dollar retention and gross margin — and explains what each one hides when you get it wrong.
- How To Start A Dev Tools Company | Startup School 🔗 talk notes (Nicolas Dessaigne) [Business models, Devex, Management, Product Strategy, startup] [Duration: 00:32] Algolia's co-founder, now a YC group partner, walks through the whole arc of a developer tools company: why runtime ideas beat build-time ones, why you almost certainly do not need a business co-founder, when open source is mandatory rather than optional, and why at a dev tools company the documentation, the support queue and the marketing should all be done by engineers.
- Starting A Company? The Key Terms You Should Know | Startup School 🔗 talk notes (Dalton Caldwell) [Business models, General, Management, startup] [Duration: 00:17] A plain-language glossary of the vocabulary founders are expected to already know — MVP, venture capital, angel, burn rate, seed and series rounds, product market fit, bootstrapping, SAFEs and convertible notes, equity, TAM, valuation, IPO and ARR — with the caveats that matter more than the definitions.
You can explore all my recommended talks and podcasts on the interactive picks site, where you can filter by topic, speaker, and rating:
- 🌐 Interactive Site: https://eferro.github.io/eferro-picks-site/
