Why long-form writing and blogs will continue to thrive in spite of AI (greyenlightenment.com)

Despite predictions that large language models would kill writing, long-form blogs and explainer posts are thriving. The author—who runs an RSS aggregator (pubstack.site) built from 40 blogs—reports a marked surge in both the quantity and length of human-produced posts since the AI debate intensified in 2024–25. Market signals back this up: in October 2025 Paramount paid $150 million for theftp.com (Bari Weiss), underlining that unique voices, brands and reader loyalty still command huge value that firms haven't been able to replace with AI or cheaper freelancers. The piece also notes an overall improvement in online long-form quality since the early 2010s and the rise of platforms like Substack that have normalized paid subscriptions. For the AI/ML community this matters: LLMs appear to amplify demand for original analysis rather than eliminate it. A large fraction of the surge—perhaps 50–70%—is content about AI itself (safety, education, policy, law), creating ongoing publishing opportunities for researchers and informed commentators. Economic effects of automation (more leisure, disposable income, subscription patronage) and the creator economy further sustain readership and monetization. Practically, that means AI should be seen as augmentative: it raises noise and production volume but increases the premium on trustworthy expertise, narrative voice, curation, and verifiable research—areas where human authors still outperform generic LLM outputs.