Bob Cornwall reviews Vulnerable God
bobcornwall.com/2026/10/vuln…
ICE supporter now being held by ICE
roysreport.com/christian-com…
Vatican criticizes bishops who pose for photo praying over Donald Trump
thelettersfromleo.com/p/past…
Bible Places Weekend Roundup, Part 1 bibleplaces.com/blog/2026/10…
Sibyls and Sibylline oracles. 2. The Sibylline books kiwihellenist.blogspot.com/2…
Trump Compares Himself To Alexander The Great joemygod.com/2026/10/trump-c…
In Reddit ModWorld 2026 there was a discussion of AI. Some commenters said "AI art is art." I replied, "Heating up a TV dinner is cooking."
James F. McGrath retweeted
By every measure, Opus 4.8 beats human experts as peer reviewers in a study of 124 Harvard social science papers covering a range from statistical analyses to formal models to qualitative work.
As academics, we need to figure out how to adapt the practice of science to AI, fast.
James F. McGrath retweeted
Replying to @rohanpaul_ai
Two links in that chain fail. An LLM predicts the next token from its context window, and eval-style prompts are a recognizable pattern from training. A spam filter detects spam too. Detecting a pattern isn't experiencing anything. #StopAnthropomorphizingAI #StopHumanizingAI
James F. McGrath retweeted
How bad is administrative bloat in higher ed?
At MIT, the “student body has grown by 8% and its faculty by 9% since the mid-1980s. By comparison, the tally of administrators jumped by 189%.”
AI predictions: seven ways you can tell useful thinking from sci-fi fantasies
naomialderman.substack.com/p…
Humans who think chatbots are trustworthy, and who have access to nuclear weapons, are an existential threat. LLMs themselves are not.
redletterchristians.org/2026…
Trump Calls CNN’s Kaitlan Collins “Low IQ Dummy” joemygod.com/2026/10/trump-c…
Three reasons why AI is not conscious and perhaps never will be
nature.com/articles/d41586-0…
To support a movement enriching billionaires and a trillionaires is bad enough. To project that on your opponents is silly and sad.









