Weekly Bookmarks

These are some things I’ve wandered across on the web this week.

🔖

AI Whistleblower: We Are Being Gaslit By AI Companies, They’re Hiding The Truth! - Karen Hao

The truth about Sam Altman. AI Critic Karen Hao reveals what 90 OpenAI employees told her.

Karen Hao is an AI expert, award-winning investigative journalist, and former reporter for The Wall Street Journal covering American and Chinese tech companies. She is also co-host of the podcast The Interface and freelances for publications like More Perfect Union and The Atlantic. Her latest book is the bestselling ‘EMPIRE OF AI: Inside The Reckless Race For Total Domination.’

🔖

Introduction to Compilers and Language Design

This is a free online textbook: you are welcome to access the chapter PDFs directly below. If you prefer to hold a real book, you can also purchase a hardcover or paperback below. The textbook and materials have been developed by Prof. Douglas Thain as part of the CSE 40243 compilers class at the University of Notre Dame. Join our mailing list to receive occasional announcements of new editions and other updates.

🔖

Inside Claude Code With Its Creator Boris Cherny

A somewhat bizarre interview with the creator of Claude Code, where he talks about the origins of the tool, and how its current development fits in with Anthropic’s business plans – which seem pretty vague other than taking over the world.

🔖

Mitchell Hashimoto’s new way of writing code

Mitchell Hashimoto on building HashiCorp, navigating the cloud giants, and how AI agents have transformed his day-to-day engineering workflow.

🔖

pi-mono contributing guide

Some open source projects that accept AI contributions are moving to a model where PRs need to reference an issue that has been marked approved by an existing maintainer with a lgtm comment. This then triggers a Github Action that adds the user to the .github/APPROVED_CONTRIBUTORS file. Then when a PR comes in, it isn’t immediately closed.

https://newsletter.pragmaticengineer.com/p/mitchell-hashimoto

🔖

badlogic / pi-mono

AI agent toolkit: coding agent CLI, unified LLM API, TUI & web UI libraries, Slack bot, vLLM pods.

(apparently Claude Code was built with this?)

🔖

The Coral Bones

The Coral Bones is a tale of three women from different times in Earth’s history, each of whom has a special relationship with the Great Barrier Reef. Judith is the daughter of a 19th Century English sea captain and is desperate to study the natural world, just like the famous Mr. Darwin. Hana is a Japanese-Australia scientist from the present day, studying the dying reef. And Telma is a descendant of refugees in a near future Australia where most forms of animal life except humans are functionally extinct.

🔖

In Ascension

In Ascension is a 2023 novel by Martin MacInnes, published in the UK by Atlantic Books and in the US by Grove Atlantic.[1] It is published or forthcoming in ten languages. The novel tells the story of Leigh, a young girl who grows up in the Netherlands amid the specter of climate change and eventually becomes a marine scientist exploring ocean trenches and investigating an anomaly at the edge of the Solar System.

🔖 Artemis II

Mission Dashboard

A dashboard for what is going on right now for the mission and space weather, and hand livestream.

🔖

Papers, Please: The toll of age verification laws on digital sex work

“The only point [of these laws] is to restrict access to content,” Riana Pfefferkorn, an attorney and policy fellow at Stanford University’s Institute for Human-Centered Artificial Intelligence, told me. “I think that the ubiquity of [age verification] lately has warped people’s views of what online safety means, so that now everything is just like, ‘Why don’t we just do [age verification]? Won’t that fix it?’” she continued. She was alluding to the use of AI to generate adult content, as well as the trend of users on X requesting that the platform’s AI assistant, Grok, non-consensually undress photos of potentially underage girls. In response to outcry, X chose to paywall access to its AI tools

🔖

Content Neutrality for Kids: Intermediate Scrutiny for Social Media Age-Verification Laws

The First Amendment imposes a high, but not insurmountable, hurdle for states to overcome in regulating minors’ social media use. By focusing on specific features that lead to harmful effects on minors, states can craft content-neutral laws that will merit only intermediate scrutiny. The solutions to the LinkedIn Problem proposed above — naming platforms directly under TikTok’s revival of the “special characteristics” standard or regulating specific harmful features without reference to content — are the two likeliest ways for states to have their laws upheld in court. Like California, states must be creative and flexible as they respond to a rapidly developing legal doctrine. If “[s]ocial media is a cancer on our society,”213 then seeking a constitutional cure is crucial even if current efforts “dwell only on the suffering of children.”214

🔖

The machines are fine. I’m worried about us.

for someone who doesn’t yet have that intuition, the grunt work is the work. The boring parts and the important parts are tangled together in a way that you can’t separate in advance

🔖

On The Enshittification of Audre Lorde: “The Master’s Tools” in Tech

This is not an argument to reject the enshittification analysis. It is an argument to extend it. A decolonial critique of technology is not simply “the internet was always bad.” It is rather: the conditions that made the internet harmful to specific communities were never peripheral to its design; they’re an integral part of it. And any politics that aims to restore something like the pre-enshittification internet without reckoning with those conditions is doomed to reproduce them.