NEW POST Powerful context engineering is becoming a huge part of the developer experience of modern LLM tools. @birgitta410.bsky.social explains the current state of context configuration features, using Claude Code as an example. martinfowler.com/articles/exp...

Fragments: teaching your pair is learning, use LLMs as a GPS rather than a destination, future libraries may be specs, and LLMs of the future funded by advertising martinfowler.com/fragments/20...

NEW POST These days I'm seeing a lot of bold fonts, especially in LLM-written prose. But when bold is used frequently, it quickly becomes ineffective. martinfowler.com/bliki/Excess...

NEW POST Erik Doernenburg maintains CCMenu: a Mac application that shows the status of CI/CD builds in the Mac menu bar. He assesses how using a coding agent affects internal code quality by adding a feature using the agent, and seeing what happens to the code. martinfowler.com/articles/exp...

Fragments: Thoughtworks launches AI/works™, electricity consumption of LLMs, how LLMs shift where we put discipline, and thoughts on The Situation in Minnesota martinfowler.com/fragments/20...

NEW POST I had a conversation with Unmesh Joshi, and Rebecca Parsons where we talked about the feedback loop between the "what" and "how" in software development, and the role LLMs play in it. martinfowler.com/articles/con...

NEW POST Jim Highsmith sees teams turning into tribes wedded to adaptation or optimization - but we should manage the tension between them by blending modes of explore and exploit, based on uncertainty, risk, cost of change, & evidence threshold. martinfowler.com/articles/sto...

My favorite albums from last year. Balkan brass, an acoustic favorite of 80s returns, Ethio-jazz, Guatemalan singer-guitarist, jazz-rock/Indian classical fusion, and a unique male vocalist. martinfowler.com/articles/202...

Fragments: How AI is changing Anthropic's internal development, a detailed account of using LLM to program a knowledge management tool, obvious-easy-possible buckets for interface design, specifications cannot be complete, and lightweight tools to work with LLMs martinfowler.com/fragments/20...

I continue to enjoy playing board games: here's my summary of 2025 gaming on my BGG blog. boardgamegeek.com/blog/13064/b...

Two legends confirmed for The Pragmatic Summit: Martin Fowler (@martinfowler.com) and Kent Beck (@kentbeck.com) We'll talk about what past booms+busts taught them, and their take on AI+software engineering: what they see working, and what not so much 11 Feb, SF: www.pragmaticsummit.com

Tonight at around 1815 EDT, we are featuring #Indonesia again! With the 3rd Edition delivering, now seemed a great time to revisit the all-timer, but now, in green & 4k! www.youtube.com/watch?v=JkI5... See y’all tonight! @jeroendoumen.bsky.social @martinfowler.com

Fragments: Illustrated guide to mainframe modernization, what the second value of code reviews means for LLMs, LLMs write and then port an HTML parser martinfowler.com/fragments/20...

Fragments: The nature of AI's prose, what makes good tests, EUs fines on X aren't about free speech, and the perils of being a reverse-centaur martinfowler.com/articles/202...

Some fragments on AI, its impact on maintainability, the dangers from its jailbreaking, developer workflows - and how it can be an ally against Parkinson's disease martinfowler.com/articles/202...

Thoughtworks Radar, on a podcast with @gergely.pragmaticengineer.com , another way at looking at feature/time curves, and the return of @heavycardboard.bsky.social martinfowler.com/articles/202...

Tomorrow - pod episode with @martinfowler.com coming!

Today, @nicolefv.bsky.social and Abi Noda's new book "Frictionless" is officially published. I liked it enough to write a foreword for it: martinfowler.com/articles/fri...

Unmesh Joshi finds LLMs to be a useful tool, but explains why their help becomes illusory if they shortcut the essential learning loop of our professional practice. martinfowler.com/articles/llm...

Concerns about LLM-enabled browsers, the different perspectives on AI-Coding, and looking back at the loss of an iconic helmet. martinfowler.com/articles/202...

There's got to be a massive chance of public backlash against all this AI assistant stuff It's totally being forced down people's throats and it barely works yet Those pushing it have a massive megaphone But one killerapp of capitalism is people vote with their wallets

NEW POST Agentic AI systems are amazing, but so are the security issues they raise. Korny has written a great primer on the problem - and what we can do to mitigate it.

The #1 criteria for my browser is that I trust it, and it’s secure. It’s logged into my email, Stripe account, stores my passwords and credit cards. A browser is useful if it does all these. These AI browsers are interesting but… with prompt injections, no way I trust them…

The primary and most pervasive form of speech suppression that this Administration employs is chill; it's pervasive but much harder to quantify than outright censorship. But stories like this must be happening everywhere. www.politico.com/news/magazin...

👀Domains and Bounded Contexts in DDD, tech industry focusing on the wrong kind of AI, story of Direct File, and U.S. government crippling its cybersecurity martinfowler.com/articles/202...

The #NoKings protest in Wakefield MA (a suburb of Boston). There was another crowd up by the lake as well.

„…the tech industry should stop focusing so heavily on these one-size-fits-all tools, and instead concentrate on narrow, specialized A.I. tools engineered for particular problems. Because, frankly, they’re often more effective.“ www.nytimes.com/2025/10/16/o...

This is so good

I tried to make sense of "spec-driven development" by looking at 3 tools: Amazon's Kiro, GitHub's spec-kit, and the Tessl Framework martinfowler.com/articles/exp...

Can we use AI to help services update with their templates?

Don't want to say I called it, but... you know... sometimes I know what I'm talking about. www.techdirt.com/2024/11/27/b...

Beth certainly will!

NEW POST I'm about to head off on travels for the next few weeks. Before I go, here's a bunch of scattered thoughts about Gen-AI and software development martinfowler.com/articles/202...

NEW POST Legacy Modernization is hard, especially when we've lost the source code. Thiyagu Palanisamy and Chandirasekar Thiagarajan describe how AI helped them reverse engineer. martinfowler.com/articles/bla...

NEW POST The Bahmni open-source hospital management system was built with an AngularJS front-end over OpenMRS REST API. Rahul Ramesh is using AI tools to convert it to React over FHIR. martinfowler.com/articles/res...

NEW POST CLI coding agents are powerful tools but don't understand your particular environment. Ben O'Mahony shows how to assemble your own. martinfowler.com/articles/bui...

NEW POST Unmesh Joshi and I had an interesting email conversation about how when programming with an LLM he likes to grow a language of abstractions. martinfowler.com/articles/con...

NEW POST When I used to give talks, I was very good at keeping them to the allotted time, even with a last-minute change. Key tool: Expansion Joints. martinfowler.com/bliki/Expans...

NEW POST OKRs are popular but often set in a top-down cascade. Paulo Caroli describes a path for teams to create their own, in a collaborative process to align with broader strategy. martinfowler.com/articles/tea...

FINAL § @srny.bsky.social concludes his article in impact intelligence by addressing five common objections to this activity, including slowing down, lack of agility and collaboration, and the unpredictability of innovation. martinfowler.com/articles/imp...

This is why the whole thing is so hard to trust and if you have any critical thinking, you stay away from such rankings. The same is true for Gartner, where many vendors also pay very heavy sums to Gartner to have Gartner evaulate them

So to be impressed by a machine going way beyond the question asked feels like the technical implementation of this exact anti-pattern. Worse, all that generated output creates significant gravity/path dependence, your brain turns off all that much faster. Why do we assume this is good?

Five Quick Reads: Durable v Disposable code, Domain Models as memory, getting AI to generate its own coding rules, what AI should stand for, and what loses more time than you save with AI martinfowler.com/articles/202...

NEW § @srny.bsky.social continues his article on impact intelligence by outlining five actions that can be done to improve impact intelligence including: demand management, impact validation, and an alternative to ROI. martinfowler.com/articles/imp...

Nik Malykhin found it useful to get an AI assistant to write its own coding rules by analyzing his code, and then asking it to refine them as worked with it. nik1379616.substack.com/p/can-we-mak...

NEW POST Sriram Narayan looks at how to properly assess the effectiveness of software development - by developing an intelligence about its business impact martinfowler.com/articles/imp...

We recently ran an experiment to explore how far GenAI can currently be pushed toward autonomously developing high-quality, up-to-date software without human intervention, and gather observations about where it breaks down. martinfowler.com/articles/pus...

My pulse will be quickenin'

NEW POST @xpmatteo.bsky.social shows why the common metric of AI code acceptance has big hole. LLM can be helpful even if you throw away its code. martinfowler.com/articles/exp...