Artificial Ignorance

Content highlights

• Weekly AI Roundup: a numbered weekly (issue 150 and counting) that uses a few key developments to explain "what happened and why it matters", refusing news-link bombardment
• Deep essays and analysis: pieces like "LLMs are getting dumber" and "The Field Guide to AI Slop" that other media frequently cite and discuss
• Engineering practice tutorials (paid): project walkthroughs with code examples and real-world architecture cases — a continuously updated "AI engineering field manual"
• De-hyped perspective: the author explicitly promises no dry ML theory, no junk links, and no anxiety-selling content — only actionable AI practice

Pricing

• Free subscription: $0, includes the weekly Roundup and most deep essays
• Paid subscription: $10/month or $100/year, unlocking AI engineering tutorials, code examples, funding data, paid-subscriber-only workshops, and direct messages to the author
• Note: prices are not shown on the homepage — visible only after entering the Substack subscription flow

Who it's for Software engineers, product managers, founders, and indie developers; people who want to follow AI efficiently without being whipsawed by hype; content consumers who prefer "in-depth long-form from a real author + email reading" over short-video news feeds.

Pros

• The author is a first-hand practitioner (Stanford CS, repeat YC founder, OpenAI DevEx engineer), making his viewpoint genuinely scarce
• 26,000+ subscribers and growing; featured in "best AI newsletter" lists from Georgia Straight, Unite.AI, and others, with public endorsements from fellow AI newsletter writers

Cons / caveats

• Text-only format, no video or interactive sandboxes — not for visual learners
• Deep tutorials, code examples, and funding data all sit behind the paywall, capping free readers' value
• The author works at OpenAI and coverage skews toward OpenAI products — a potential bias; cross-check sources like The Batch and Import AI
• The name "Artificial Ignorance" also belongs to an academic book and a critical term — disambiguate when searching