“I believe life is a continuum, and that no one really dies, they just drop their physical body and we'll all meet again, like the song says. It's sad but it's not devastating if you think like that... It's a continuum, and we're all going to be fine at the end of the story.”
— David Lynch
Simon Willison shares his approach to running a link blog. “I don’t like to recommend something if I’ve not read that thing myself, and sticking in a detail that shows I read past the first paragraph helps keep me honest about that.”
Power of Community Posts: Conversation is the New Influencer
Research by Reddit, about Reddit – At CES 2025, Reddit released research outlining how conversation is becoming an influencer throughout the purchase journey. “In the last year, our research has shown that 47% of social media users say “irrelevant search terms” are the most frustrating aspects of their product research). That is in part why people are coming to communities like Reddit from search looking for advice, answers, and recommendations. In fact, Reddit is the number one social media platform for finding “immediate answers to specific questions” (71% agree or strongly agree). Overwhelmed by online choice and AI-generated content, the 2025 research outlines why back-and-forth, bidirectional conversation is shaping a new landscape of decision making to bring back confidence to buying:
- Discovery: Reddit conversations are the new landing page for business discovery. According to new Reddit research, Reddit ad investments lead to more organic posts, creating a multiplier effect of discoverability.
- Consideration: Nearly 1-in-4 recommendation posts elicit a redirection, where the original poster considers a brand they had not considered previously (23% of all recommendation posts and up +30% YoY).
- Decision: 2 out of 5 social media users (42%) find Reddit recommendations more valuable in their decision to purchase a product vs. all other forms of recommendations.
42% of Reddit users say that the ability to ask questions within communities they trust is in the top 3 most important aspects of their product research…”
Meta Secretly Trained Its AI on a Notorious Piracy Database
Wired – [unpaywalled] Newly Unredacted Court Docs Reveal – One of the most important AI copyright legal battles just took a major turn : “Meta just lost a major fight in its ongoing legal battle with a group of authors suing the company for copyright infringement over how it trained its artificial intelligence models. Against the company’s wishes, a court unredacted information alleging that Meta used Library Genesis (LibGen), a notorious so-called shadow library of pirated books that originated in Russia, to help train its generative AI language models. The case, Kadrey et al. v. Meta Platforms, was one of the earliest copyright lawsuits filed against a tech company over its AI training practices. Its outcome, along with those of dozens of similar casesworking their way through courts in the United States, will determine whether technology companies can legally use creative works to train AI moving forward and could either entrench AI’s most powerful players or derail them. Vince Chhabria, a judge for the United States District Court for the Northern District of California, ordered both Meta and the plaintiffs on Wednesday to file full versions of a batch of documents after calling Meta’s approach to redacting them “preposterous,” adding that, for the most part, “there is not a single thing in those briefs that should be sealed.” Chhabria ruled that Meta was not pushing to redact the materials in order to protect its business interests but instead to “avoid negative publicity.” The documents were originally filed late last year remained publicly unavailable in unredacted form until now…”
Before these documents were made public, Meta previously disclosed in a research paper that it had trained its Llama large language model on portions of Books3, a dataset of around 196,000 books scraped from the internet. It had not previously publicly indicated, however, that it had torrented data directly from
- Mary the color scientist and her friends — The Philosophical Quarterly puts Jackson’s original article and 8 others it has published about it over the years in an open-access collection
- The obstacle course of knowledge — a list of things about us and the world that help explain why “knowing things is hard” (via The Browser)
- “We argue that there are several pragmatic reasons—based on a combination of biological, social and normative considerations—to classify pregnancy as a disease” — so argue Anna Smajdor and Joona Räsänen
- “Your relation to yourself when you peruse forgotten pages of a journal kept by a forgotten you, uncanny but not unpleasant, addressed by the inverse of a ghost…” — Kieran Setiya (MIT) on writing for oneself
- An example of how well AI can summarize a philosophy paper, devise objections to it, reply to those objections, etc. — ChatGPT o1 Pro on C. Thi Nguyen, with an ironic twist, from Kelly Truelove
- “Can one portray a nervous man, not by showing him manifesting nervousness in his appearance in any way, but only by inducing, through ‘formal’ means, nervousness in the viewer?” — Brad Skow on a type of artistic “magic act”
- “While public intellectuals generally rise to prominence by being an average person’s idea of a smart person, this usually is because they are, in fact, a smart person” — in defense of public intellectuals
