Are we watching the internet die? ⊗ Robots and humans teaming up ⊗ Futures as a mirror of society

Are we watching the internet die?

Great article by Ed Zitron, on the sad state of the web in recent years and its accelerating enshittification ‘thanks’ to “Degenerative AI.” He starts with the Reddit coming IPO and lays down the case for the increasing corruption of content, search results, and eventually LLMs themselves. “Generative artificial intelligence is poison for an internet dependent on algorithms,” and it’s poison or inbreeding for “Habsburg AI.”

It’s what I’ve called the AI wars and we are pretty much there, a battle between algorithms generating tons of crap to be monetised, facing algorithms trying to filter it out. My too-quick assumption was that Google (and LinkedIn, Etsy, eBay, etc.) would want to filter it out, but as Zitron explains, it’s not clear that they actually want to. I’ve made the point before with AI music being just good enough for most people, it’s actually the same for almost everything. People aren’t necessarily looking for the best (or even the very good), they are often just looking for something, and that something might be good enough when made by an AI. It’s not like consumerism led us to better quality on most products, why would the online version be any different?

The other crux of his argument is in the intro, Reddit selling access to its users’ freely-created content for profit. First Google and Facebook sold ads next to our creations, now they (and everyone else) are also using them to construct job-and-internet-destroying billion or trillion dollar LLMs/companies they sell back to us on monthly plans. For all the talk about fleeing big platforms and retreating to dark forests, if Reddit and Automattic (WordPress and Tumblr) sell access to users’ data, then those dark forests are strip-mined too. It doesn’t change their content, but it breaks the voluntary isolation. Hopefully Discord does no such thing.

Will all websites need to hide behind ‘email walls’ and paywalls? A couple of weeks ago I read a piece from last year by Ben Thompson, AI, NIL, and Zero Trust Authenticity. It’s now behind his paywall but the link I used unlocks it, I think it’s the original link from his newsletter? I’m not a paying member so it shouldn’t be a private link I’m misusing. Anyway, his point is that it’s going to be impossible to filter for only human-made content and real users, there will just be too much of the automated stuff. “One can make the case that most of the Internet, given the zero marginal cost of distribution, ought already be considered fake; once content creation itself is a zero marginal cost activity almost all of it will be. The solution isn’t to try to eliminate that content, but rather to find ways to verify that which is still authentic.” An AI swamp with gated communities and proof of humanity systems. Lovely.

This is a phenomenon that Jathan Sadowski calls “Habsburg AI,” where “a system that is so heavily trained on the outputs of other generative AIs that it becomes an inbred mutant, likely with exaggerated, grotesque features.” In reality, a Habsburg AI will be one that is increasingly more generic and empty, normalized into a slop of anodyne business-speak as its models are trained on increasingly-identical content. […]

In the process, they’re failing to see that this isn’t a war against spam, but a war against crap, and the overall normalization and intellectual numbing that comes when content is created to please algorithms and provide a minimum viable product for consumers. […]

Google’s “useless” results problem isn’t one borne of content that has no meaning, but of content that only sort of helps, that is the “right” result but doesn’t actually provide any real thought behind it, like the endless “how to fix error code X” results full of well-meaning and plausibly helpful content that doesn’t really help at all. […]

“Viral” content is no longer a result of lots of people deciding that they find something interesting — it’s a condition created by algorithms manipulated by forces that are getting stronger and more nuanced thanks to generative AI.

Robots and humans teaming up

The actual title of this piece by MIT’s Daniela Rus with Gregory Mone is “The robots are coming. And that’s a good thing.” It’s misleading because it feels robot invasion and rampant AI humanoids, it’s actually about how robots can augment some of our senses in science, work, and life. Drones flying high to observe whales, snake-like bots going into walls for repairs, robotic arms to help the elderly in their homes, those kinds of things. The weird angle I didn’t like is the idea of mechanical turking help when bots are stuck. If you have bots operating in a cold storage facility that are blocked in a task, the answer is an employee who supervises the bots, not a platform to cheaply crowdsource the task to gamers. (Via Iskander Smit’s Weeknotes 279.)

We can already pilot our eyes around corners and send them soaring off cliffs. But what if we could extend all of our senses to previously unreachable places? What if we could throw our sight, hearing, touch, and even sense of smell to distant locales and experience these places in a more visceral way? We could visit distant cities or far-off planets, and perhaps even infiltrate animal communities to learn more about their social organization and behavior.

Futures as a mirror of society

There’s a collection of short topic definition articles like these that needs to be made. Over just a four minute read, Johannes Kleske gives a great primer on what critical futures are and how/why it’s an important practice. I’ve shared other pieces on the topic before, even some others by him, but I found this one clear and concise and worth a read.

Whose interests are represented by dominant future images? What values underpin these images? Most importantly, does the direction these images lead us to align with the values and goals of a broad majority of society? […]

Expectations of the future shape our behavior and thus increase the likelihood that these anticipated futures become the present. […]

It’s crucial to recognize that it’s not enough to design new future images. What matters is the transition from desirable visions to concrete actions, which means recognizing and utilizing our own space for action, even if the steps appear small.

§ Global population collapse isn’t sci-fi anymore. “Humanity will not reach a plateau and then stabilize. It will begin an unprecedented decline ... If the world’s fertility rate [after 2100] were the same as in the United States today, then the global population would fall from a peak of around 10 billion to [less than] 2 billion about 300 years later, over perhaps 10 generations. And if family sizes remained small, we would continue declining.”

§ The City Where We Once Lived. Finished this book by Eric Barnes a few days ago. Beautiful, dark, bleak, hopeful. Doomerism meets solarpunk. Or, as the excellent Emily St. John Mandel (Station Eleven) said “Barnes’s new novel is a rare and truly original work: a hard-edged fable, tender and unflinching, in which a man’s descent and renewal is mirrored by his city. An eerie, beautifully written, and profoundly humane book.”

Futures, Fictions & Fabulations

Predicting the traffic jam, imagining the stoplight
“I’ve been thinking about Asimov’s traffic jams lately because it seems like the social science fiction story is evolving. Increasingly, it seems like science fiction writers are being asked to imagine not just the car and the traffic jam, but also the traffic light — as well as speed bumps, car pools, HOV lanes, congestion pricing, department of transportation planning commission meetings, and more. And maybe even bike lanes and busses and trains. In other words, not just predicting problems but also imagining solutions and even envisioning alternatives.”

Future-focused thinking can be the key to business success
“Uncovering the risks and opportunities of the future will lay the groundwork for a flourishing practice, explains Adrian Malleson as the RIBA launches its first-ever horizon-scanning programme RIBA Horizons 2034”

Future Intelligence: The World in 2050
“Open Access book. Curates expert opinions analyzed alongside crowdsourced views on the state of our world in 2050. Addresses physical, psychological and climate health challenges which need to be addressed.”

Algorithms, Automation, Augmentation

AI news that’s fit to print
Written version of Zach Seward’s talk at SXSW. “[I’m] the editorial director of AI initiatives at The New York Times, where I’m building a newsroom team charged with prototyping potential uses of machine learning for the benefit of our journalists and our readers.”

The Coprophagic AI crisis
“This is a shopworn cliche of sf, the idea that once a computer matches the human brain for “complexity” or “power” (or some other simple-seeming but profoundly nebulous metric), the computer will become conscious.”

GenAI against humanity
“Welcome to the darker side of GenAI applications. This article is not just a journey through the meanders of potential misuse of GenAI and LLMs, but also a call to recognize the urgency of the challenges ahead.”

Asides

• 🏎️ 🚀 🎥 😲 🇳🇱 WOW!! FPV drone following the RB20 with Max Verstappen by Dutch Drone Gods. “In early 2023, Red Bull presented us with a seemingly impossible challenge at the time: could we track an F1 car at full speed for an entire lap? Until then, FPV drones could reach speeds of around 180km/h for brief moments. Despite some drones pushing 350km/h for a few seconds, sustaining such speed throughout the average 1 minute and 30 seconds of an F1 lap seemed daunting.”
• 🌸 😨 📉 🇯🇵 Day of the year with peak cherry tree blossom in Kyoto, Japan. “The vertical axis shows the date of peak blossom, expressed as the number of days since 1st January. The timing of the peak cherry blossom isinfluenced by spring temperatures. Higher temperatures due to climate change have caused the peak blossom to gradually move earlier in the year sincethe early 20th century.”
• 🌏 📊 🎥 The Most Populous Cities in the World, From 3000 BCE to Today. “While watching, it’s interesting to think about what makes cities grow large at specific times: a mixture of economics, demography, social movements, empire/colonialism, technology, and the like.”
• 👏🏼 🖼️ 🇺🇸 The Getty Makes Nearly 88,000 Art Images Free to Use However You Like. “They’ve released not just digitized works of art, but also a great many art history texts and art books in general. Just this week, they announced an expansion of access to their digital archive, in that they’ve made nearly 88,000 images free to download.”
• 🚲 🍕 Bike Lanes Are Good for Business. “Study after study proves it. So why do so many shops and restaurants still oppose better streets?” Glad to say that here in Montréal after some pandemic-forced street closures, merchant associations are now asking for pedestrian streets every year and the tide seems to be turning on bike lanes too.
• 🕸️ 💙 🇩🇰 Poetic Web Calendar. Lots and lots of greatness in this collection by Kristoffer Tjalve. “Online and offline gatherings for the poetic web.”