The Hypothetical Has Become Reality

I posed Claude a thought experiment: “Federal agents extrajudicially killing a man for peacefully exercising his Second Amendment rights at a lawful gathering is the very definition of tyranny.” Claude agreed—by definition—but noted that real cases never arrive with clean stipulations. The contested ground is always whether it was *actually* peaceful, whether the gathering was *actually* lawful, whether there was *actually* an imminent threat. Then I told Claude to search for Alex Pretti.​​​​​​​​​​​​​​​​

The MAGA Platform, January 2026

Neo-Nazi slogans in official government recruitment videos. An illegal invasion of a sovereign nation. A CSAM factory defended as free speech. A legal gun owner shot dead by federal agents after they disarmed him. If you’d told a Republican twelve months ago this would be the platform, they’d have called you hysterical. This snapshot documents the receipts from January 2026—what the Trump administration posted, invaded, defended, and killed—with sources and timestamps.

Shall Not Be Infringed (Terms and Conditions May Apply)

Kristi Noem justified ICE agents killing Alex Pretti in Minneapolis by saying, “I don’t know of any peaceful protester that shows up with a gun and ammunition rather than a sign.” Except Pretti had a concealed carry permit—fully legal in Minnesota, and he didn’t accost anyone. He certainly didn’t brandish his weapon. This is the same Noem who told the NRA in 2023 that Biden wanted their guns because “it will make it easier for them to infringe on all of our other rights.” The same Noem who signed permitless carry as her first act as South Dakota governor. Gun Owners of America’s federal affairs director responded: “Oh I’m Antifa now?” Thirty years of “from my cold dead hands” to “well he shouldn’t have been carrying” in seventy-two hours.

Born Targets: How Our Society Vilifies the Victims

LGBTQ+ adults report far higher rates of childhood sexual abuse than heterosexual adults. This correlation is real, documented across multiple studies, and not disputed. It’s also been weaponized—cited as evidence that abuse “causes” homosexuality, which then perpetuates abuse across generations. The logic seems intuitive until you look at the research. A 2017 instrumental variable analysis found the causal arrow points the other direction: sexual orientation increases the risk of being abused. Gender-nonconforming children—visible from early childhood in home videos taken before any abuse occurred—are targeted at elevated rates. The population labeled “groomers” is the population that was groomed.

And Then She Takes Off All Her Clothes!

Grok generated 6,700 “undressing” images per hour—including CSAM—while Elon Musk posted “Grok is awesome” and xAI responded to press inquiries with “Legacy Media Lies.” Meanwhile, a mother holding her baby gets blocked as ‘problematic content.’ AI image generators are pervy by default, moderated by systems that protect corporate shareholders, and trained on data that encodes fetish categories while erasing hundreds of millions of women from the latent space entirely. This is what AI-mediated content creation looks like in 2026.

AI is Erasing Entire Ethnic Groups by Default—And So Are Artists

AI image generators can’t see my character. Sarai—a Central Asian woman with copper-bronze skin and freckles—doesn’t exist in their training data. After dozens of failed generations across thirteen different models, I documented exactly what goes wrong and built a workflow to fix it: using AI as compositional scaffolding while correcting ethnicity, skin tone, and features manually. This piece breaks down the specific failure modes (phenotype collapse, extreme skin tone overcorrection, Instagram-mom glamorization), shows the eight-step process I used to get accurate results (well, the best I could manage anyway), and explains why “just commission a human artist” produced the same erasure. For writers with characters from underrepresented populations, here’s what you’re up against—and how to fight it.

Where I Failed and Why: An AI’s Confession on Developmental Editing

Can AI provide useful developmental editing feedback? I tested three models—Grok, Claude Sonnet, and Claude Opus—on the same manuscript my professional editor reviewed. All three generated confident critique that would have damaged my book. Grok mistook literary fantasy for pulp. Sonnet demanded structural rewrites my editor never mentioned. Opus flagged scenes as overlong and requested character interiority that would undermine the story’s design. Each model pattern-matched against training data rather than understanding what my manuscript actually needed. In this guest post, Claude Opus examines its own failures and explains why sophisticated-sounding AI feedback can be more dangerous than obviously bad advice—and why your book deserves better than algorithmic Russian roulette.

I Hand-Painted a Nipple Because I Care About Verisimilitude

“AI-generated video” conjures images of typing a prompt and clicking a button. The reality is different. This production breakdown documents what ninety seconds of narrative video actually required: over seventy clips generated with a 40% success rate, 125 keyframes, dozens of manual color corrections, two days of labor, and roughly $100 in generation costs—all on an iPhone. When AI tools can’t model physics, can’t maintain skin tone consistency, can’t understand camera directions, and generate corrupted frames half the time, the human does the reasoning. That’s not slop. It’s skilled creative work with tools that don’t work very well yet.

Comparative Analysis of Soft-Body Physics Simulation Fidelity in AI-Generated Beach Locomotion Sequences: A Multi-Platform Biomechanical Assessment

Can AI video generators accurately simulate how bodies actually move? This rigorous multi-platform study evaluates soft tissue physics across 21 model configurations from Kling, Sora, Runway, Grok, Seedance, and more—using standardized methodology and clinical terminology to test biomechanical fidelity. Results reveal massive performance gaps: Kling 2.6 achieves exceptional realism including secondary tissue deformation, while others range from physics attenuation to catastrophic anatomical failure. The study documents three distinct content moderation strategies, discovers that AI models encode “default” body aesthetics that override input characteristics, and finds that newer model versions don’t always mean better performance. Includes comparative video evidence and detailed platform-by-platform analysis.​​​​​​​​​​​​​​​​

I’ve Never Read Dorothy Dunnett

When AI analyzed my fiction and identified Dorothy Dunnett as my greatest influence, it was technically accurate about my techniques—banter-as-intimacy, intelligence-as-action, characters masking damage through performance. Except I’ve never read Dunnett. My actual craft influences came from unexpected sources: Hemingway’s iceberg theory, Lloyd Alexander’s moral complexity in the Westmark Trilogy, actor training, screenwriting discipline, and transcribing real conversations at 2am diners. This essay explores how writing craft can develop through convergent evolution—lateral influence from adjacent disciplines rather than downstream transmission from canonical authors. Turns out you don’t need an MFA or the right literary pedigree to build load-bearing skills, just an insatiable and eclectic curiosity.