Martin Alderson
Web development, AI tooling, and building better software
martinalderson.comMartin Alderson is an independent blog covering programming. It publishes on a weekly or bi-weekly basis, with 37 posts in its archive.
Regular
Publishes weekly or bi-weekly
0
Independent Blog
English
How this blog's content is accessed through Blogs Are Back.
Excerpts Only
RSS feed provides excerpts — visit the blog for full posts
Direct Access
Feed can be fetched directly from your browser
Direct Post Links
Post pages can be loaded directly in the reader
Embeddable
Posts can be displayed inline in the reader view
Recent posts from Martin Alderson's RSS feed.
Why on-device agentic AI can't keep up
There's a growing narrative that on-device AI is about to free us from the cloud - the pitch is compelling. Local inference means privacy, zero latency, no API costs. Run your own agents on your computer or phone, no cloud required. Indeed, the pace of improvements in open weights models has been spectacular - if you've got (tens of) thousands to drop on a Mac Studio cluster or a high end GPU setup, local models are genuinely useful. But for the other 99% of devices people actually carry around,...
Using OpenCode in CI/CD for AI pull request reviews
Most existing AI code review tools require you to grant them access to your GitHub or GitLab repositories. While some of these tools are interesting, the security implications of handing over repo access to a third party are significant - and they're typically GitHub or GitLab-first. If you're working on projects that don't use either of those platforms, you're out of luck. I run a few projects where we don't use GitHub or GitLab, so these tools simply aren't an option. That led me to explore an...
Which web frameworks are most token-efficient for AI agents?
I wrote an article a couple of months ago about which languages are the most token efficient. I've been thinking about this for quite a while - and many others have too, thinking through what happens to programming languages now increasingly agents are writing code, not humans. However, it did occur to me that maybe this is the wrong angle to look at the question. These days, frameworks tend to matter far more than the language itself, so I thought I'd see if I could repeat the previous research...
Who fixes the zero-days AI finds in abandoned software?
Anthropic's red team released research showing that Claude Opus 4.6 can find critical vulnerabilities in established open source projects. They found over 500 high-severity bugs across projects like GhostScript and OpenSC - some of which had gone undetected for decades. This is impressive, and genuinely useful work. But their research focused on maintained software - projects where patches can actually be shipped. The scarier problem is the enormous long tail of abandoned software that nobody wi...
Attack of the SaaS clones
I cloned most of Linear's core functionality in 20 prompts using Claude Code. It took a couple of evenings and a few million tokens. Here's what I think that means for the future of SaaS economics. A quick recap In my previous two posts, AI agents are starting to eat SaaS and a more recent one about the sharp decline in software company valuations, I've covered some of the risks to existing SaaS businesses now agentic coding capabilities have increased so much (and perhaps more importantly, cont...
Follow Martin Alderson
Add this blog to your reading list on Blogs Are Back, or visit the blog directly.