Judging beautiful docs, AI fatigue, and tool slop
In this podcast, I chat with Fabrizio Ferri-Benedetti about a variety of topics related to AI and docs, such as applying Italo Calvino's literary principles of lightness and quickness to evaluate docs, the reality of AI review fatigue versus creator fatigue, whether vibe-coded tools are tools slop, developing internal skills for repeatable doc processes, and the utility of running local AI models.
Review of Max Tegmark's 'Life 3.0: Being Human in the Age of Artificial Intelligence'
This post is my review of Max Tegmark's Life 3.0, a book we read in the AI Book Club. Tegmark, an MIT physicist and founder of the Future of Life Institute, says Life 1.0 relies entirely on biological evolution to change its hardware and software, Life 2.0 (humanity) can design its own software (through learning and culture) but is stuck with its evolved hardware, and Life 3.0 is a technological stage where life can design both its software and its physical hardware. Although Tegmark's book was published in 2017 (before ChatGPT), the book's core questions about recursive self-improvement, AI alignment, steering beneficial AI, etc., are still highly relevant today.
This is a recording of the AI Book Club discussion of Max Tegmark's Life 3.0: Being Human in the Age of Artificial Intelligence. Our discussion explores Tegmark's visions of superintelligence, the unpredictability of AI goals and subgoals, and how the Asilomar principles fare against the relentless market pressures of today's AI arms race. We also contrast the book's theoretical focus on cosmic energy (and how superintelligence might more efficiently extract energy from matter) with the immediate, tangible threats AI poses to cybersecurity and global financial systems.
Developing internal skills for recurring documentation processes like release notes
My hypothesis this year around AI was that if I develop some agent skills to speed up repeatable processes, it might clear up my bandwidth and free up time for me to work on non-repeatable doc tasks. It appears to be working.
Looking back at the AI Book Club one year in
This month marks the one-year anniversary of the AI Book Club. We’ve now read 12 books in the club — see the list of books. For each session you can view recordings, notes, transcripts, discussion questions, etc. By some measures, the book club might not be considered much of a success. For each session, I record the discussion using Google Meet. Then I publish the recording on YouTube, along with a transcript and other notes. For the 12 vide...
On pace and value -- why is moving slow boring?
After my post about how work expands to fill the space allotted, for a few days afterwards, I kind of slipped into the mindset that it didn’t matter how hard I worked, since my work and the work of those around me would all be viewed relatively the same, and if it comes time for someone to do massive layoffs (like we see in the news with Meta, Microsoft, Nike, etc.), it’ll be someone who doesn’t know me, who likely doesn’t even consult my mana...
One of the things I dislike about AI is how it takes over everything — every conversation, every process, every tool. But this week I noticed it started taking over my thinking as well. Here’s what happened. While building reference docs for an SDK I support, one of the APIs failed due to some conflicting visibility conditions. I was jumping the gun on writing the release notes before I had all the launch entry info and Gerrit previews, so I ...
Work expands to fill the space allotted
Last year I had a goal of getting down to bug zero and started a whole series on this quest. If this is the first you’re hearing of this goal, bug zero really means clearing out all doc requests (~JIRA tickets), finishing them. It’s similar to the idea of inbox zero, but with work. I actually hit a brief moment where I hit bug zero, and I felt pretty accomplished. But within 1 or 2 days of clearing out that bug queue, it filled back up to 10 ...
I don’t entirely understand my relationship with coffee, but I think I have some character flaws that could be advantageous in some situations and disadvantageous in others. In some areas of my life, I’m an all-or-nothing person. This seems to be true with coffee. When I’m drinking coffee, I can’t seem to stop myself from drinking 3-4 cups a day. Before long, I can sense my body basically becoming desiccated (from the way coffee acts as a diur...
AI Book Club discussion recording of 'Breakneck: China's Quest to Engineer the Future', by Dan Wang
This is a recording of the AI Book Club discussion of Dan Wang's Breakneck: China's Quest to Engineer the Future. Our discussion tries to tie some of the book's themes to AI (to align with the book club's theme), but only in places where it makes sense. Some topics we discuss include social engineering, electric vehicles, contrasting energy infrastructures and the ability to power AI compute, mass surveillance networks, the differences between America's lawyerly society and China's engineering state. Fortunately, one of the book club members had recently been to China and could share firsthand experiences and impressions of cities like Shenzhen and Beijing. Among many observations, she describes the tech fatigue of everyday life, where everything (even prices at stores) requires you to use an app.
Some thoughts after using AI to help with taxes
Apr 20, 2026 • ai
After receiving a letter from the IRS stating that I owed more taxes on a previous year, I used AI to identify the issue -- a missing cost basis on stock sales. I then successfully disputed the notice, and ultimately claimed a sizable refund by carrying over capital losses. AI tools like Claude and Gemini have become surprisingly capable at analyzing uploaded tax returns and guiding novices like me through complex scenarios.
Podcast: How valuable are agent skills? Conversation with Larah Vasquez and Fabrizio Ferri-Benedetti
In this podcast, I chat with Larah Vasquez and Fabrizio Ferri-Benedetti about using skills to extend AI capabilities, the future of agentic engineering, local models like Qwen and Gemma, and whether the tech writer role is shifting into automation architecture. We get into the memory problem in LLMs (and why some of us actually prefer the no memory to extended memory), the progression from prompt engineering to context engineering to compound engineering to orchestrating whole agent systems, and how skills are quietly forcing engineers to write down knowledge they'd never documented before.
I recently gave a presentation to students and faculty in person at Louisiana Tech University on March 30, 2026, focusing on what I call the cyborg model of technical writing. The idea is that the emerging model for tech writing isn't one in which AI replaces tech writers but rather one in which AI augments tech writers. Tech writers interact with AI in a continuous back-and-forth, conversational, iterative manner. This post contains the recording, slides, transcript, summary, notes, and more from my presentation.
Will tech writers survive AI? Perspectives from two professors, Nupoor Ranade and Jeremy Merritt
In this podcast, I chat with two professors — Nupoor Ranade (Carnegie Mellon) and Jeremy Merritt (James Madison University) — about how AI is reshaping the technical writing profession from the academic side. We discuss dropping enrollments, misconceptions about what tech writers do, historical parallels to past disruptions, agentic AI and organizational restructuring, the cyborg model of human-machine collaboration, and how academics and practitioners can bridge the divide to solve real problems together.
AI Book Club recording of 'If Anyone Builds It, Everyone Dies'
This is a recording of our AI Book Club discussion of If Anyone Builds It, Everyone Dies: Why Superhuman AI Will Kill Us All by Nate Soares and Eliezer Yudkowsky, held March 15, 2026. Our discussion touches on a variety of topics, including whether the book's use of parables strengthens or weakens its argument, the question of whether AI can develop genuine intentions, the competitive dynamics that prevent any single company from pumping the brakes, the limits of recursive self-improvement, and what ordinary people should make of wildly conflicting predictions from leading AI thinkers. This post also includes discussion questions, key themes, and a full transcript.
