Hacker News: james_ross

New comment by james_ross in "Understanding Is the New Bottleneck"

james_ross — Sun, 05 Jul 2026 06:12:11 +0000

I'm a huge fan of looking back to sources outside our own field for ideas on how to cope with the changes within it. An influential one on me comes from education, too, "Learning How to Learn" by Joseph D. Novak. The main take away is that our brains work associatively, and you can both extract and share mental models into a visual form called a concept map. I've been experimenting with capturing them as plain text files that both humans and agents can read as a very dense form of shared memory. When trying to build understanding, the idea is you start by extracting the associations people already hold, and then build bridges one proposition at a time to the new concepts.

New comment by james_ross in "Understanding the Dynamics of the AI Ecosystem with Pace Layers"

james_ross — Sun, 05 Jul 2026 05:52:18 +0000

This makes sense to me, and reminds me of the more familiar variation in rates of chage in building business systems. I think I first came across this in the writing of Ronald G. Ross in relation to his business rules approach. At the bottom layer, changing at the slowest pace, are the core business terms/ concepts, above which sit business rules, and above those business processes, and above those user interactions, which change the fastest of all. The current trend we see to provide agentic end user access to business processes formerly intermeidated by web pages is an example of the friction from above placing demands on the levels below to adapt. If the business process layer isn't set up well to serve completely new kinds of user interatction, then it will come under pressure to be refactored in order to do so from the faster pace of change at the higher layer.

New comment by james_ross in "Why do teams keep losing context, and why hasn't any tool fixed it?"

james_ross — Sat, 04 Jul 2026 04:59:35 +0000

Yes it's a product of mine, but it's actually a recent rebuild in desktop app form of an app I built as a web app well over a decade ago. And yes, I used it to get hmans on the same page at some scale, which you can hear me describe in talk I did at the time like this one: https://www.infoq.com/presentations/concept-map/

I haven't done formal evals in this AI era about how having a concept map alongside the codebase changes what the agents do, but informally I know mine read them because I tell them to and they seem to take it into account, and I seem to have less trouble than the average geek I talk to keeping my code base adequately coherent. It's just plain text context, after all so there's no magic, just very concise and accurate context.

New comment by james_ross in "Best Simple System for Now (2025)"

james_ross — Sat, 04 Jul 2026 04:55:11 +0000

This article describes one of the classic polarities in our industry, the tension between the short term and long term. There are many others, such as the tension between autonomy and alignment in organisations with lots of teams. The article did not mention polarity management, but it would help it if it did IMHO. The essence of polarity management is that the problems described can never be solved once and for all, because it's not a problem that you can fix, like you can fix, say, the damage to your car after an accident. The short term and long term tension will be ever present, and the key to managing them is to acknowledge that both have positive attributes, but both, when taken too far, have negative ones. The best advice in these kinds of situations is to work out in which direction you may have leaned too far, and lean back the other way. The key is not to dogmatically lean back too far the other way because you've treated it as a problem that can be "fixed". It can only ever be managed, and each fix creates the next imbalance.

New comment by james_ross in "Show HN: PMB – local memory for coding agents that shows if it is used"

james_ross — Thu, 02 Jul 2026 04:58:12 +0000

this looks very interesting to me and I plan to try it. I was looking at the web site (pmbai.dev) earlier today and it looked very thorough but when I went to show someone this afternoon it no longer loads in my browser, saying it can't establish a secure connection. Verified by clicking the website link on the repo page on github so I am pretty sure I have the url right...

New comment by james_ross in "Why do teams keep losing context, and why hasn't any tool fixed it?"

james_ross — Wed, 01 Jul 2026 21:31:57 +0000

Depending on your definition of context, my attempt to solve just the foundational context of what words mean in a given domain, and sharing it cleanly, quickly and reliably with both new people in the team and AI agents, is the humble concept map stored in plain text. Talk here: https://www.infoq.com/presentations/concept-map/ Product here: https://thinkingtools.software/concepticon

New comment by james_ross in "Engineering for Bounded Cognition"

james_ross — Mon, 29 Jun 2026 10:12:43 +0000

Ha — you actually did it. Genuinely made my day; you've now read it more recently than I have. And I think you've landed where I did: the concept-mapping part is the keeper and the rest I have not put to practical use.

New comment by james_ross in "From Prompts to Loops: Building Autonomous Coding Agents"

james_ross — Mon, 29 Jun 2026 09:55:26 +0000

I'm intrigued by this idea, and plan to test it to build a new product that is a sibling of an existing one, but with a different targeted purpose. I believe it's verifiable enough to try a goal oriented approach but I'm slightly nervous about this all just being a way to get us to burn the next order of magnitude in tokens!

New comment by james_ross in "Finding It Challenging to Maintain Software Created with Coding Agents?"

james_ross — Mon, 29 Jun 2026 08:17:37 +0000

I'm about to try Matt Pocock's skills for more rigorous agentic engineering now that my code base is climbing towards 100kloc, but can't yet testify to their effectiveness. People I trust swear by them: https://github.com/mattpocock/skills

New comment by james_ross in "Reality has a surprising amount of detail (2017)"

james_ross — Sun, 28 Jun 2026 07:39:50 +0000

a good read indeed! Makes me think about my use of coding agents differently, as the main thing they do is deal with a lot of details that matter to the execution but don't matter to me personally enough to figure them out. Would love to see this author's more recent take as this was written pre-LLMs taking over the world.....

New comment by james_ross in "It took two weeks to make Claude's "overnight solution" for flaky tests useful"

james_ross — Sun, 28 Jun 2026 04:27:49 +0000

This matches my experience as well, as the one person team on a desktop app with thousands of unit tests and hundreds of playwright e2e tests. I had a number of flaky tests that Claude was self selecting to isolate when running the tests and this was concerning. The breakthrough for me was using the superpowers debugging skill and setting a focused goal to fix one particular test that was failing most often. It ended up being a race condition that I'd never have found on my own, and it then went and found the dozen or so other similar issues in the code base. No e2e failures now. This is a very satisfying use of an AI agent for me.

New comment by james_ross in "The Exhaustion of Talking to a Tool"

james_ross — Sun, 28 Jun 2026 04:14:51 +0000

I agree, but I get the impression talking to others and from articles like this one that everyone's experience is real, it's just that the diversity of experiences is so vast that there are some really good ones and some really bad ones and the best thing we can do is share more on how to get the better ones more often. I'm in a weird situation where my day job involves running a 400 person dev team and my evenings and weekends involves playing with my one person startup and my virtual team of AI agents, so I get a lot of heavy in person stuff and plenty of AI stuff too. The key for me in this last few months on the virtual side, which has been anything but fatiguing, is to use a framework a friend of mine made that treats agents not as order takers but senior collaborators who push back, have explicit autonomy levels for different tasks, and are by design proactive and helpful. It's made me able to confidently do things like product management and marketing that I lacked confidence in while leaving me to be as pedantic and particular as I like as the engineer on the team. This has been energising, not fatiguing. The framework is here: https://github.com/normannoble/agent-framework

New comment by james_ross in "How to Write an Effective Software Design Document"

james_ross — Fri, 26 Jun 2026 11:20:50 +0000

Thanks for sharing; that's a really thorough approach. These days I wonder if you have thought of creating an AI skill to guide people through the creation of such a document using an interview style interaction, and perhaps a separate one to assess the quality of a proposed design document. At my current workplace, we're trying to dial up the "definition of ready" along these kinds of lines. My other favourite thing to add is to use a concept map rather than just a glossary

New comment by james_ross in "Mixing Visual and Textual Code"

james_ross — Thu, 25 Jun 2026 09:42:07 +0000

for a long time I've believed that the way to bridge the visual and the textual worlds is to generate the visuals from a plain text source of truth. I've never gone so far as to make a visually editable programming language, but I have done multiple projects where, for example, process diagrams in an IDEF0 style were generated from human readable text propositions of the form:

Prepare Meal is a Process Oven enables Prepare Meal Utensils enable Prepare Meal Prepare Meal transforms Raw Ingredients Prepare Meal produces Finished Meal Menu governs Prepare Meal Customer Order governs Prepare Meal

You can map out a business process using a very simple (and XML-free, and diffable) plain text DSL and generate interactive diagrams from it. My most recent example is for concept maps along similar lines.

New comment by james_ross in "Agentic coding and persistent returns to expertise"

james_ross — Wed, 24 Jun 2026 08:30:34 +0000

this seems to be a load bearing claim in the article: "success is determined by how well a person understands the problem they are trying to solve, not whether they’re trained in coding."

I would normally assume that my understanding of the problem I'm trying to solve is a blend of domain knowledge and system knowledge; that is, some of the what and some of the how. And just as important as bringing domain knowledge to the table is bringing the ability to articulate it with conceptual clarity in a way that bridges to the software. I have met many very deep domain experts who don't have that skill.

My personal interest of late is in capturing domain knowledge and managing it like source code, and using it to share with both the humans (in visual form) and the agents (in plain text form) some of that foundational understanding to make them both better at both the what and the how. This seems aligned with the article when it says "the more understanding a worker brings to an agent, the more quality work the agent is able to do" but I would also say that the more understanding we can bring to future humans new to the domain has pretty good ROI as well.

New comment by james_ross in "Google proposes Open Knowledge Format based on Markdown"

james_ross — Tue, 23 Jun 2026 21:24:22 +0000

I'm a massive fan of capturing domain knowledge in a plain text format that both humans and AI can use, one of my favourites being concept maps stored in human readable form. I did a talk on this exact topic long before we all got AI pilled, this version from 2015 for example: https://www.infoq.com/presentations/concept-map/

The web app mentioned in that talk that I built to help with that no longer exists, but I recently built a new desktop app (Apple Silicon Mac only so far, sorry) for this exact purpose: https://thinkingtools.software/concepticon