I've dabbled a few times in writing bitmap font parsers for both technically constrained and artistic projects. There is a reason that design has resolved to the same few cliches, because expectability, latent understanding, and 'obviousness' reduces onboarding curve and fatigue. It's a cognitive accessibility issue before you even get to legitimate accessibility concerns. Render a .F16 at anything larger than 16px in a modern application and you're introducing issues which are solved, quantitatively and qualitatively, by vector graphics and antialiasing. There's an optimistic naivety which is nice to have, but misunderstands design as a conduit for informed action vs design as an aesthetic function independent of intent is legitimately dangerous if you're doing anything other than building narrative products emulating older tech.
Something I've been tinkering on for the last few weeks is absent.dev[1], was working towards getting the guts to share this on HN but there's a few things I want to polish first. This is the MVP, but it will essentially be moving towards what you're suggesting, much more customisable news aggregation and more importantly, where you consume it. Feed agnostic, channel agnostic, in the form(s) you want (longform, shortform, summary, tl;dr etc). You can check the clustering logic on the digest[2].
I think that the inverse thesis is true, you make websites more accessible (a11y, wgac, aria labels etc) for humans, then the interaction heuristics are clearer for agents functioning off browser-use or similar. If a screen reader can understand your site, then an agent can. Reinventing the wheel to facilitate the current state of agents makes the web worse for everyone, it's not a preemptive move, it's actually a decline in almost objective and measurable quality, and potentially one which removes access to the internet by people who just want to.. use the internet.
This. The accessibility tree is a superpower for agents when it's good. Screenshots are "robust" but low performance. Like so many other things, making stuff better for humans indirectly makes it better for agents.
reply