Again with the AI? Your weekend reading from Web Directions
This week marked the 2 year anniversary of the launch ChatGPT. It feels like much longer ago, and so recent. So much hype in so little time. And yet also so much promise. The challenge is separating those out.
So perhaps it’s no surprise that this week or two my reading has included numerous articles on the topic.
And it seems that Bluesky is having a moment–you can find me there. I’ve been primarily on Mastodon for the last year and change, so if you’re on Mastodon, where I post relatively frequently. Come say hi!
Against the dark forest
The complex of ideas I’m going to call the Dark Internet Forest emerges from mostly insidery tech thinking, but from multiple directions—initially in Kickstarter co-founder Yancey Strickler’s freeform noticings that apply science fiction writer Liu Cixin’s dark forest theory of the universe to social media, then in humanist all-arounder Maggie Appleton’s illustrated tech notes. It names an experience of paranoia and anxiety that by the end of the 2010s was widespread among people with meaningful connections between their online personas and their ability to maintain their standard of living. It hit a nerve, especially within some corners of tech-and-society thinking that influence internet makers. It even shows up in a New York Review of Books piece: a coup for something so initially modest.
Source: against the dark forest
Erin Kissane interrogates the idea of the dark forest theory of the web. It’s one of the most thoughtful pieces on this challenging and open ended topic, and she leaves us with a quest–instead of withdrawing, we should strive to reform the structural issues of the internet to create safer, more inclusive spaces.
Bonus Video: Watch Maggie Appleton’s talk from last year’s Web Directions Summit (no need to sign in or signup!).
The Coming Bifurcation of the Web: AgentWeb vs ArtisanWeb
The web is splitting into two distinct futures: one driven by relentless AI agents, the other shaped by human hands. It fills me with both excitement and sadness – excited by the possibilities of AI-driven efficiency, yet nostalgic for the craftsmanship of a human touch. Let’s dive into what this means for the future of web development and ownership.
Source: The Coming Bifurcation of the Web: AgentWeb vs ArtisanWeb | Michael Antczak
The aligns with Maggie Appleton’s talk just a year ago (I originally wrote ‘a couple of years back’) “The Expanding Dark Forest and Generative AI”, though perhaps not as sombre as that.
Why are we using LLMs as calculators?
So why are we trying to get these enormous models, good for natural text completion tasks like summarization, translation, and writing poems, to multiply three-digit numbers and, what’s more, attempt to return the results as a number?
Two reasons:
- Humans always try to use any new software/hardware we invent to do calculation
- We don’t actually want them to do math for the sake of replacing calculators, we want to understand if they can reason their way to AGI.
Source: Why are we using LLMs as calculators?
Vicki Boykis explores the history of using computers for mathematics. A great read.
Building LLMs is probably not going be a brilliant business
In the 1960s, airlines were The Future. That is why old films have so many swish shots of airports in them. Airlines though, turned out to be an unavoidably rubbish business. I’ve flown on loads of airlines that have gone bust: Monarch, WOW Air, Thomas Cook, Flybmi, Zoom. And those are all busts from before coronavirus – times change but being an airline is always a bad idea.
Source: Building LLMs is probably not going be a brilliant business
I once read that in the entire history of commercial aviation, collectively airlines have not made a profit. Sure, there are lots of profitable aspects of aviation–running privatised airports is one (essentially government granted rights to tax people are pretty good ‘businesses’) and aircraft manufacturers have done well (more supplying governments with military aircraft than commercial aviation, but that was for a while a good business too). Cal Paterson thinks the same may be true of the business of Large Language Models–maybe they just aren’t great businesses? Bit early to tell, but a well reasoned argument.
Principles of Developer AI Product Development
How do you build products and platforms for developers in a world that contains probabilistic black boxes that surprise you with what they can and can’t do, and when they decide to show you.
From my own trial and error, I have found that most of my mistakes are in not understanding the two pieces: developers and AI systems.
Source: Principles of Developer AI Product Development
Dion Almaer reflects on the nature of product engineering in the era of large language models.
AI for Web Work
Kevin Powell pitted several AI models against each other in a video focused on CSS questions and found Claude to be the best. I think a ton of coding usage of AI is in VS Code which leans Microsoft Copilot which uses GPT-4o by default although you can change it in the UI, so it might be worth giving others a shot. I like how using different models is pretty developer friendly these days.
Source: AI for Web Work – Frontend Masters Boost
Excellent evidence based investigation into the use of LLMs for web development work.
For The Love of God, Make Your Own Website
Browsing the internet used to be a hobby of mine. Ever since my dad got us a modem when I was around ten, I spent hours at a time just looking at different websites. The internet felt like a limitless expanse of free expression. Now, despite how many more people use the internet, I usually end up at the same three or four websites, and I end up a lot more bored.
Source: For The Love of God, Make Your Own Website – Aftermath
Related in a subtle way to all of the above, a heartfelt plea I entirely concur with. Right now so much our online life is mediated by algorithms and the whims of those who get to decide them–yes, we still deeply rely on Google’s search algorithms, but regain some control over your web, our web, the web by POSSE (Publishing (on your) Own Site, Syndicate Elsewhere).
Great reading, every weekend.
We round up the best writing about the web and send it your way each Friday.
