Emily Bache

Earlier today, I weighed in on the Great Instructional Debates of 2026 by sharing my mental model of effective instruction. You can make of it what you will, but make sure to check the reflective and insightful comment left by @biblioracle.bsky.social about his own teaching.

A big part of the PHP fix was just one guy in Czechia, Ondřej Surý. What he had to say about it is worth a read www.theatlantic.com/technology/2...

"TDD is actually easier than writing the tests afterwards, so why is it so hard to learn?" - @emilybache.com What are your thoughts? 👇

How does learning, technical coaching and training change with AI? In today's blog post, I'm reflecting on @emilybache.com 's "Samman Society Statement on Agentic AI" and adding my own thoughts. Read it here: www.devteams.at/gen_ai/2026/... German version and newsletter link below 👇

This month it's "software factory". Four months ago it was "agent harnesses". Four months before that it was "loops" and "gastowns". Four months before that it was "agent skills". Four months before that it was "MCP". Four months before that... Ughck. Feeling nostalgic for 2010s JavaScript fatigue.

The Industry Is Catching Up: The Only Way to Generate AI Code That Works | @davefarley77.bsky.social 📅 TONIGHT AT 7PM (UK) Subsribe & turn on notifications 🔔 so you NEVER MISS AN UPLOAD ➡️ youtube.com/@ModernSoftw...

As I'm filling my queue for blog posts again, I'm slowly getting back into the virtuous cycle that I loved about blogging, back then, when I did it more regularly: As soon as I create a draft for a future blog post, I get more ideas for more drafts. Highly recommended. Start blogging.

IMHO Postman is the worst thing to happen to API testing. In the old days devs used to write & share tools (ie scripts) to test APIs which were easily extensible and, more importantly, automatable. Now every dev has their own customised Postman collection and testing is manual 😔

A little piece about mining my own (Claude/Codex/VIBE) sessions for mistakes my agent made: blog.code-cop.org/2026/10/impr... #AI #ProcessImprovements

My thoughts on the current state of AI driven software development, why I'm no longer concerned about a runaway intelligence explosion, and what I think is a far worse outcome we could be walking into maevyn.codes/decoherence

Kudos to Netflix! They did a great job experimenting and measuring the effect of the code intelligence tools on the agents. I wish we could see more studies like this from other big companies, too: "20–30% more efficient (fewer tokens, less time)" blog.jetbrains.com/ai/2026/10/h...

Agentic AI coding BREAKS Mob Programming (Here’s How To Fix It) | @emilybache.com 📽️ AVAILABLE NOW Watch HERE ➡️ youtu.be/ydAin2IWaQo

Collaborative coding in an ensemble or mob has been a cornerstone of my work, and when you add agentic AI it needs different interaction patterns. My latest video and blog explain some of the things I've found to work. youtu.be/ydAin2IWaQo & coding-is-like-cooking.info/2026/10/agen...

Agentic AI coding BREAKS Mob Programming (Here’s How To Fix It) | @emilybache.com 📅 TONIGHT @ 7PM (UK) Subscribe & turn on notifications 🔔 so you never miss an upload! ➡️ youtube.com/@ModernSoftw...

The Boots Theory of Software Architecture "Organisations that seem not to have the time or the money to address technical concerns, repay their technical debt or improve their architecture are typically caught in a trap." kevlinhenney.medium.com/the-boots-th...

The Samman Technical Coaching Society has published a statement about Agentic AI: sammancoaching.org/society/agen...

New blogpost: Another experiment. Since April I've been handing more and more work to AI agents, and turning up the practices I've trusted for years so it stays safe. It feels like it's working really well. Wrote it up, failures included: www.eferro.net/2026/10/turn...

Mark your calender: Saturday, 14th Nov 2026 Global Day of Coderetreat in Vienna -> insights.squer.io/gdcr-2026

This is wild - somebody did a deep dive into the real stories behind this famous XKCD: sheets.works/data-viz/hol...

You’re Writing Legacy Code Right Now (And Don’t Even Know It) | @davefarley77.bsky.social 📽️ AVAILABLE NOW Watch HERE ➡️ youtu.be/jScItKhruN0

My thoughts on the recent TDD with LLM discussion. Theatre? gregorriegler.com/2026/10/03/t...

This is outstanding both as a piece of parody and as a more digestible explanation of what happened with the Hugging Face hack. https://www.frogandtoad.ai/

Blown away by the interest and positive feedback! The Cleaning Up episode on the Primary Energy Fallacy with me and Michael Liebreich is now in the top 10 of all 276 episodes of the show available on YouTube. www.youtube.com/watch?v=7xDH...

I've built and rebuilt this "thing" over the past 2.5 years ~4 times. It's finally taking real shape as a thing that other people will be able to use. A code / entire platform inspection tool - that builds code graphs and supports agents with context via MCP. "What if SonarQube was useful?"

Half of Europe's population was exposed to poor or worse ground-level ozone levels so far in 2026, as successive summer heatwaves drove up concentrations, according to new @CopernicusECMWF analysis. 🔗 atmosphere.copernicus.eu/air-quality/... #AirQuality #Ozone

Is asking AI agents to do TDD "theater" or actual value? Join @emilybache.com for a free live workshop: "From Traditional TDD to AI-Assisted, Without Losing the Plot" 🤖 Learn how to drive AI with TDD discipline. Book your spot ➡️ www.subscribepage.com/eb-webinar

I've published detailed notes and an annotated transcript to accompany the video of the keynote I gave at @wearedevelopers.bsky.social World Congress North America in San Jose on Friday - here's my rundown of everything that's happened with LLMs in 2026 so far simonwillison.net/2026/Sep/27/...

New post/newsletter: Harms of the AI ​​ecosystem – infographic A visual aid for discussions and learnings about the impact of AI development on everything from individual to global to planetary wellbeing. #DigitalEthics #AIEthics axbom.com/aieco/

Yes, you read that right. NINE times cheaper.

The Last Year Has Changed Everything I Knew About TDD After 20 Years | @emilybache.com 📽️ AVAILABLE NOW Watch HERE ➡️ youtu.be/wK5WgbqtI50

About TDD inside the agent loop - my response to @birgitta410.bsky.social 's recent article. video: youtu.be/wK5WgbqtI50, article: coding-is-like-cooking.info/2026/09/the-...

The Last Year Has Changed Everything I Knew About TDD After 20 Years | @emilybache.com 📅 TONIGHT @ 7PM (UK) Subscribe & turn on notifications 🔔 so you NEVER miss an upload! ➡️ youtube.com/@ModernSoftw...

We are not screwed. We can still stop the worst impacts of climate change and build a better future – if we act soon and act at scale. drawdown.org/insights/wer...

I structured my thoughts for months about My Writing and GenAI - read here blog.code-cop.org/2026/09/my-w...

At Lunar, we've been appreciating a flexible and broad skill set over a deep one. For years, it felt like an uphill battle. Enter AI. Suddenly, the job market is all like, "You were right all that time." pawelbrodzinski.substack.com/p/product-de...

We all have to stop treating technical practices as a side quest when they are, in fact, the mission. "If the only problem were translating the cost of AI tools into end-to-end value, it would be serious enough. But something far worse is happening..." thenewstack.io/ai-co...

Was the Hexagonal Architecture invented by Alan Key?

When AI generates the job description and AI rewrites the CV, the system is broken.

What is the first trait that comes to mind when you think of the BEST TEAM you ever worked in? 👇

And, in 2025, Europe produced more electricity from wind and solar combined than from fossil fuels for the first time.

🚨 Announcing Pushin 🚨 Now officially: I'm announcing Pushin.eu, a git hosting platform built and run entirely in Europe. Think GitHub/GitLab but European, better, and faster. Built with #ElixirLang and #RustLang. peterullrich.com/announcing-p...

2026 really was a fundamentally different summer to 1976 in terms of temperature across the UK and Europe Data: ERA5

What is the BEST case scenario for AI adoption in the tech industry? Say, in 10 years, if you were the ultimate optimist, what would you want it to look like?

10 tips if you're moderating a panel. Trust me I have learned these the hard way by watching 1000 bad ones (including with me in the chair), and a few great ones. Panels need to be lively, engaging and enlightening. Here's a few pointers on how to make that happen (if you're lucky).

I'm speaking at Code.talks in Hamburg in November. They currently have a special "back to the office" discount on tickets - take a look! codetalks.com

Did you miss training your juniors during your adoption of GenAI? Regular "learning hour" sessions with whole teams are a great way to teach developers to take (back?) control of the software quality if their work. See sammancoaching.org/training/del... (15 September 2026, 14:00 CEST)

The car industry didn’t see this coming. In 2015, 330k fully electric cars were sold globally, about 1% of the market. In 2025 it was 14m, close to 1/5 of all new cars. The fleet on the road tells you about the past. Sales figures tell you about the future. janrosenow.substack.com/p/the-car-in...

Verify now supports inline snapshots github.com/VerifyTests/...

Seeing a lot of "mutation testing is too slow" posts from people who are trying to do it with an LLM. People, the existing mutation analysis tools are literally orders of magnitude faster. And that's true even if you exclude all the time used by the LLM.

Both the Economist and others suggest this study demonstrates that AI can help students. So I contacted the lead researcher to figure out how many. Answer: Out of ~26,000 kids, a total of 20. And five months after starting to use AI, only four continued to spend 65 minutes on homework. Four.