Zed Adams on AI Copyright Limitations

Since F(r)ictions, I have been thinking about how new technology changes the way we see ourselves. Sometimes it gives us a new identity to grow into, the way writing gave us the writer. Other times it gives us something to define ourselves against, the way being called a computer once meant you were skilled and now sounds like an insult. Zed Adams discusses that idea in this video from the symposium.

Ever been fascinated by the name “Claude”? I used to wonder how Anthropic landed on a name that fits their product so well. Turns out, it’s not random. It’s named after Claude Shannon, widely regarded as the father of information theory. What’s even more interesting is that his Master’s thesis at Massachusetts Institute of Technology showed how electrical circuits could represent logical operations, essentially laying the foundation for modern digital computing. Decades later, he went on to formalize how we measure and transmit information, introducing ideas like “bits” and showing that communication is really about managing uncertainty. So in a way, the systems powering today’s AI trace back to ideas he formalized long before the internet even existed. Crazy how something as simple as a name can carry that much history. If you’re curious to read more about his impact, this is a great deep dive: https://lnkd.in/eFfVsfZU

Why resist GenAI? come to listen to a fantastic conversation that delves into that very topic and see how others are finding their way through the quagmire of social, environmental, and economic justice issues.

Today’s technocrats would have us believe that AI represents an unprecedented solution to society’s governance woes, but trust a historian to set the record straight. In her new book, The Rise and Fall of the Artificial State, Jill Lepore argues that the “artificial state”—government by machines—is far from a new idea. It’s actually something we’ve considered and rejected for centuries. “The book,” writes philosopher Carissa Véliz in this week’s issue of Science Magazine, “is a journey through the narratives that built this framework—and a reminder that just as it was assembled, it can be dismantled.” https://lnkd.in/efj_-b65

Lot of talk about engineer productivity with AI tools. Wonder what that really looks like for a large, complex code-base? Here is an in-depth review of what building a practical, complex AI-native codebase can look like. https://lnkd.in/gRuy_9PD

“How the Internet Disrupted Science” is essential reading for anyone seeking a clearer understanding of academic journal publishing. Most discussions of the topic begin with a perfunctory bow to the 17th century royal societies, and then leap to Robert Maxwell, the Serpent that supposedly put an end to 300 years of static publishing Eden. Not only is that picture absurd on its face, but it treats publishing as an autonomous enterprise separate from the influence of three key drivers: the institution of the university, the role of government, and the influence of technology. Kent and Joy’s book uniquely places journal publishing in the context of Silicon Valley’s development and metastasis. Any aspirations towards autonomy were subordinated many years ago to the desire to become mini versions of Google. An appropriate subtitle would be: the decline of “content” and the rise of “data.” https://lnkd.in/gv4MvMYW

I just finished Cory Doctorow’s The Reverse Centaur’s Guide to Life After AI, and it gave me a distinction worth sitting with: a centaur is someone technology makes better and keeps in charge. A reverse centaur is someone technology uses up — paced, timed, buried reviewing output faster than any human can verify it. Same tools. Opposite relationship. The real question isn’t what a technology does — it’s whom it’s built to serve. That’s the same question I’ve been asking about social media for over a year, most recently in my “Why Slowfit?” series. Doctorow just gave it a sharper name. I also folded in a real-time example from Cal Newport: a senior engineer who went all-in on AI coding, crashed his product twice, and went back to writing most of it by hand. His point — software is the best-case scenario for AI. If it can’t safely be handed the keys there, that’s a signal about everywhere else. Which led me to an honest audit of my own human performance app, specifically the Coach Keir AI feature— not the marketing version. Full post is linked below. #AI #ReverseCentaur #HumanPerformance #SlowfitMethod

The book sits at the intersection of computer science, AI engineering, and philosophy, specifically exploring. Rather than standard feedback, I am looking for rigorous critiques regarding the technical assumptions, philosophical frameworks, and engineering realities presented in the text. You can access and download the full PDF draft directly via the Google Drive link in the first comment below. I look forward to your thoughts, pushback, and discussions in the comments or via DM. #ArtificialIntelligence #ComputerScience #Philosophy #AIEngineering #MachineLearning #TechEthics

Claude's naive research is not very reliable. It never challenges its own results. That's why I think everybody should have a proper research skill. Right now I'm running the one Nate builds in this video. First as it is. Then a variant targeted at technical problems. 3 takeaways from "Stanford's Method Turns Claude Into a PHD Level Research Team" (Nate Herk | AI Automation): ▸ A research method is only trustworthy if something challenges it. One prompt means one angle and a fixed set of blind spots. STORM runs five expert lenses to create disagreement on purpose, then a contradiction pass to surface it. ▸ Trust the V2, never the V1. Nate is blunt that his first pass contained claims that were simply wrong. The verification stage is what makes the output usable. ▸ Findings should carry provenance and dissent. Each one is ranked for reliability and annotated with which lenses supported it and which challenged it, which beats a bare confidence score. I share this to invite you on my knowledge journey. 👇 Full wiki summary in the first comment. Source: https://lnkd.in/dT6y6pbT Written with Claude Code. Content carefully curated by me. #ClaudeCode #AIAgents #MultiAgent #Research

A dancing bear, a colleague, or a sharpened toolbox? The cautious adoption of generative artificial intelligence technologies in digital humanities research Rongqian Ma, Meredith Dedema, Andrew Cox https://buff.ly/EsCo6vq