Richard MacManus

@ricmac
3.8K Followers
901 Following
5.3K Posts
Technology Analyst, Web & AI · Internet Historian @ https://cybercultural.com · Founded ReadWriteWeb (2003–2012) · 🥝 in 🇬🇧
Homepagehttps://ricmac.org
Internet history bloghttps://cybercultural.com
Alt a/chttps://indieweb.social/@classicweb
Apparently it’s 20 years since Jack Dorsey sent the first tweet. On that same day, 21 March 2006, I was blathering on about “microcontent design” on ReadWriteWeb. This was the feeds world that web geeks like me wanted to become a reality: i.e. web standards-based feeds, using RSS/Atom, structured blogging, microformats, etc. Little did we know that Twitter would introduce the concept of proprietary feeds to the internet…and that would win out over web standards. https://web.archive.org/web/20060508071004/http://www.readwriteweb.com/archives/microcontent_de.php
Read/WriteWeb: Microcontent Design, Part 1

Richard MacManus on Next Generation Web and Media

New from my Web AI Lab: I’ve built an “article assistant” for my site that runs using local AI in the browser (via Chrome + Gemini Nano) when available — and falls back to a cloud model when it isn’t.

Before you dismiss this because it's AI, I think this has huge implications for the #OpenWeb. Instead of sending every interaction to BigTech clouds, users can increasingly run AI on their own device: better privacy, lower cost & more user control.

Check it out: https://ricmac.org/2026/03/19/article-assistant-local-ai-browser/ #WebAI

Building an article assistant: local AI in the browser with cloud fallback - Richard MacManus

How I built an article assistant for my website that uses local AI (via Chrome + Gemini Nano) when available, and falls back to a cloud model when it isn’t.

Richard MacManus

Paul Ford ( @ftrain) is blogging again and I’ve found his posts speaking to me, like they did 20+ years ago. Helps that I’m about the same age as him, so I can identify with the career and life angst he writes about. Also it’s something to do with me enjoying dipping into blogs again, since a lot of social media posts are really offputting nowadays (I hate this, I hate that, etc).

“…while I am many places right now, I am not in career heaven. So I took the train home.” https://ftrain.com/whoompf

Whoompf, by Paul Ford

is the sound my body made slamming into the car that jerked left in traffic in front of me. I was wearing a helmet. Maybe the Citibike scratched something. It w

Ftrain.com
“The future is much more likely to be AI embedded inside a million bespoke workflows, not a million bespoke workflows jammed into a single AI interface. For product leaders and designers, that’s a big opportunity.” https://bigmedium.com/ideas/links/saas-is-dead-benedict-evans.html
SaaS Is Dead? | Big Medium

Benedict Evans deflates the frothy talk that AI agents and assistants will eliminate vast swaths of software.

Big Medium
There was a time when Dril’s identity was leaked and we all just agreed to ignore that and mind our business as if it hadn’t happened and I think we owe Banksy at least that much.

I have no problem at all — in principle — with automating customer help inquiries. But what this *ought* to consist of, mostly, is exposing more information to the customer in a more accessible, searchable and understandable form.

Carefully written software could supercharge / personalise “help pages” and FAQs on a company’s web site, rendering a significant proportion of help desk phone calls unnecessary. Most of the time when I need to call a company, it’s because their web UI simply isn’t offering me some information that is in their system (and which the help desk staff can see), but that they’ve decided I don’t need to know.

Anyway ... that’s how I think these things *should* work. The link is to reporting on multiple cases where, instead, companies decided to pretend than their “AI” could just mimic human customer service staff and take their place, which of course it couldn’t, and it all went horribly wrong.

[H/T @lana ]

https://www.koreatimes.co.kr/southkorea/society/20260316/what-ai-overtaking-jobs-looks-like

What AI overtaking jobs looks like

“Are you actually a human?” The voice of a middle-aged man, thick with anger, roars through the phone. For Lee Eun-young — a pseudonym for a woman...

The Korea Times

Journal: That was Web Day Out

An excellent day of talks in Brighton exactly 37 years after the birth of the World Wide Web.

🔗https://adactio.com/journal/22465

That was Web Day Out

An excellent day of talks in Brighton exactly 37 after the birth of the World Wide Web.

The 49MB Web Page

A look at modern news websites. How programmatic ad-tech, huge payloads and hostile architecture destroyed the reading experience.

https://youtu.be/lA-wD89m6jM?si=6MvqDH_Y6SzdNGLg

I watched @zachleat's AMA during THE 11ty Meetup and I appreciate his accountability and honesty about the shift from #11ty to #BuildAwesome. Now that I've had a chance to simmer and think about the realities of an open source project funding specifically, I feel better about the shift for the longevity of 11ty.

In considering the scale of businesses and organizations that rely on 11ty, yet aren't adequately supporting the project (fuck them, honestly.), I feel grateful to Zach that he found a home for 11ty that will support continued development. I also feel for Zach navigating the balance between making a damn living and maintaining the homegrown feel of 11ty and it's community. It's an impossible paradox.

I still hate the new name, and that's okay. The shift is a response to problems bigger than me and I can accept that. I will forever be grateful to the 11ty community for helping me build my development skills and finding a web home that feels encouraging and kind. 💖

AMA about Build Awesome, an Open Town Hall with Zach Leatherman | 11ty Meetup

YouTube

Your 1990s files are slowly turning into magnetic dust. Leontien Talboom at Cambridge University Library is leading a rescue mission for our digital past. She uses hobbyist tools to pull data from moldy disks that standard drives can't touch. She even recovered lost lectures by Stephen Hawking. You can use her Copy That Floppy! guide to save your own archives. It requires specialized hardware like the Greaseweazle to capture raw magnetic pulses. Don't wait until the iron oxide flakes off for good. Check your storage boxes for those chunky rectangles before they become silent plastic.
🧠 Magnetic decay destroys data after 20 years.
⚡ The Greaseweazle captures raw flux signals.
🎓 Cambridge recovered Stephen Hawking's notes.
🔍 Follow the Copy That Floppy! guide now.

https://www.popsci.com/technology/floppy-disk-archivist-project/
#TechHistory #DataArchiving #VintageTech

The archivist preserving decaying floppy disks

It's a race against time (and magnetic decay) to preserve decades of cultural history stored on obsolete hardware.

Popular Science