Martin Fowler

Can We Make AI Code Assistants Smarter by Asking Them to Write Their Own Rules? It seems logical, doesn't it?

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...

The Reformist CTO’s Guide to Impact Intelligence A guide for technology leaders and their teams to shift the organization

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...

How far can we push AI autonomy in code generation? An experiment to test the limits of autonomous code generation by LLMs

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...

There's definitely something to that. I like that I'm in a position to hear about non-SV tech orgs

My pulse will be quickenin'

Partner with the AI, throw away the code Notes from my Thoughtworks colleagues on AI-assisted software delivery

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...

smut
I love it when it's uncut
and unsubt
tle

and not to forget: "sliding down the razor blade of life"

Who is LLM? Is an LLM a stubborn donkey, a genie, a slot machine, or Uriah Heep?

NEW POST

Many people like to give their LLM assistant a persona, such as a stubborn donkey, a genie, a slot machine, and Uriah Heep

martinfowler.com/articles/who...

and a thoughtful general essay on why the hype and the immovable skeptics are both missing the train

blog.korny.info/2025/07/19/c...

Korny Sietsma has a great example of how using an LLM for coding is very helpful but with limitations.

blog.korny.info/2025/07/18/a...

Opinion | I Teach Creative Writing. This Is What A.I. Is Doing to Students.

Thoughtful essay on value and limits of AI.

"One of the real challenges here is the way that A.I. undermines the human value of attention, and the individuality that flows from that."

www.nytimes.com/2025/07/18/o... (gift link)

I've known Beth for ages: wicked smart, understands tech really well, and cares about its impact on ordinary people. We need people in Congress like that.

A moveable bollard to warn rapid walkers that the station floor is slippery, in the colour and shape of a (straight) banana.

Isn't this a perfect evolution of form and function? #Paddington #nonslip

Being afraid of failure doesn’t help people succeed. It helps them stay safe by succeeding less often.

When we make failure safer & cheaper, we can succeed more.

Doesn't ring any bells, I'm afraid

Thoughts on Motivation and My 40-Year Career I’ve never published an essay quite like this. I’ve written about my life before, reams of stuff actually, because that’s how I process what I think, but never for public consumption. I’ve been pus…

I really resonated with: "I think I’m less interested in my own happiness (whatever that means) than I am interested in doing work that feels worth doing" - @charity.wtf reflects on her career

charity.wtf/2025/07/09/t...

Leading your engineers towards an AI-assisted future Your CEO wants to know why AI isn’t writing half your code yet — are you ready for that conversation? I lay out a practical, proven strategy which will help your teams understand AI's capabilities and...

Pete Hodgson continues his excellent writing on Gen AI by proposing an approach to leading engineers towards an AI-assisted future, centered around a the concept of aligned autonomy.

blog.thepete.net/blog/2025/06...
@thepete.net

AI Tooling, Evolution and The Promiscuity of Modern Developers Zhixin Sun, Fangchen Zhao, Han Zeng, Cui Luo, Heyo Van Iten, Maoyan Zhu, CC BY 4.0, via Wikimedia Commons Historically, there have been two constants with developer tools. First, that their users were...

Gen AI breaks two common constants with developer tools: developers are willing to flit between Gen AI tools and they are willing to pay for them. Implies it's not too late for new tools , enterprise adoption will be slowed by a lack of consensus

redmonk.com/sogrady/2025... @sogrady.org

Makes me realize how hiring is not focused on this stuff: "how would this person make the team better."

I guess, it is hard to be focused on this.

But this was one of the *very* rare devs who made every team much better. Nicer. More motivated. More a "team."

Still think of it

I still care about the code Notes from my Thoughtworks colleagues on AI-assisted software delivery

NEW POST

Even with LLMs, @birgitta410.bsky.social still cares about the code: “LLMs are NOT compilers, interpreters, transpilers or assemblers of natural language, they are inferrers.

martinfowler.com/articles/exp...

Informative interview about a company that's leaned heavily into AI Augmented software development. Interesting to hear that this is leading them to increase the hiring of grads.

Expert Generalists Being an Expert Generalist should be treated as a first-class skill, one that can be assessed and taught.

FINAL §

In complex environments, the characteristics of Expert Generalists lead them to be particularly valuable in driving tasks to completion. Unmesh, Gitanjali, and I thus complete our article by summarizing the value of this skill.

martinfowler.com/articles/exp...

Expert Generalists Being an Expert Generalist should be treated as a first-class skill, one that can be assessed and taught.

NEW §

While we've spent this article praising the Expert Generalist, Unmesh, Gitanjali, and I simultaneously do not deny the value of specialist knowledge. We also note that EGs work well with LLMs.

martinfowler.com/articles/exp...

Expert Generalists Being an Expert Generalist should be treated as a first-class skill, one that can be assessed and taught.

NEW §

To grow Expert Generalists we need to focus attention on fundamentals rather tools. As an example, Unmesh, Gitanjali, and I describe a workshop we've used to break silos of application development, data engineering, and devops

martinfowler.com/articles/exp...

LLMs bring new nature of abstraction a short post

NEW POST

Gen AI for software development is the biggest jump up in abstraction since assembler to HLLs, but also a jump sideways.

martinfowler.com/articles/202...

Expert Generalists Being an Expert Generalist should be treated as a first-class skill, one that can be assessed and taught.

NEW §

We have two crucial checkpoints for spotting — and then nurturing — expert generalists. Unmesh, Gitanjali, and I look at hiring and career progression.

martinfowler.com/articles/exp...

Learnings from two years of using AI tools for software engineering How to think about today’s AI tools, approaches that work well, and concerns about using them for development. Guest post by Birgitta Böckeler, Distinguished Engineer at Thoughtworks

I wrote a guest post on The Pragmatic Engineer newsletter.pragmaticengineer.com/p/two-years-... rounding up 2 years of using AI coding assistants - how they evolved; ways of working; impact I see on speed, quality and team flow; and some thoughts on the future

I guess the question is what are the best set of characteristics to identify the most effective generalists. I think this 6 is a decent set, but I'm sure they aren't the only set, and probably not the best set. I do think we (meaning the industry) should spend more effort figuring that out.

That's a good point - but my sense is that these characteristics are being significantly undervalued at the moment. And our observation is that they are particularly important for effective generalists.

I don't have any problem with that, but you'll need to check with the authors of specific articles.