About ABV — AI · Books · Validation
ABV is a publication about applied AI, machine intelligence, and decision-making under uncertainty.
This is a newsletter about applied AI: what ships, what breaks, and what compounds.
I cover AI tools, LLM-first visibility (LLMO), agent workflows, and the boring-but-important parts: docs, automation, distribution, measurement.
Most posts are either: (1) reviews/teardowns, or (2) build logs from my own OSS and products.
No hype. No “future of everything”. Just systems, shipping, and evidence.
I focus on validation — not demos, not hype, not architectural diagrams for their own sake, but the moment where ideas, tools, and systems meet real constraints: incentives, human behavior, incomplete data, and time pressure.
Here you’ll find:
critical analysis of AI concepts, architectures, and research
practical breakdowns of tools, workflows, and decision frameworks
case-based validation: why some ideas survive contact with reality — and others don’t
occasional writing about books, systems thinking, and long-form reasoning where AI meets practice
This is not a newsletter about “the future of AI”.
It’s about how AI actually behaves now, in real products, organizations, and messy human environments.
Why subscribe
Subscribe if you care about:
understanding AI beyond surface-level explanations
separating useful ideas from impressive-looking ones
seeing how systems fail, adapt, and evolve in practice
long-form thinking instead of hot takes
No spam. No growth hacks.
Just structured thinking, validation, and honest analysis.
About the author
I’m Anton Biletskiy-Volokh — working at the intersection of applied AI, systems design, and strategic decision-making.
More links and projects: abvx.xyz
