How I Use AI to Write
Joe's Nerdy Rants #68 - Weekend reads, podcasts, and other stuff
No doubt AI is changing how people write. Some people use AI to write nearly everything. Others use it as a writing assistant. Some even avoid AI. I get questions about how I use AI for my writing these days. It’s an interesting question, because for me, writing isn’t just the physical act of writing. Writing is also about researching and thinking. Even if I’m not clanking on a keyboard, I’m “writing.” Let me explain.
Research is a big chunk of writing a book, or even a complicated article. You might think you understand a topic. Heavy research into a topic (reading, interviewing, and doing) is how you realize how much you need to learn. The big challenge is that research often has many false paths and dead ends. Some resources are valuable, and most are a waste of time. The latest generation of AI research assistants is almost ready for prime time. I’m starting to use AI research as a starting point, which helps curate the materials I can further read. Is it perfect? No, but I’m guessing it saves days of work.
Another part of research is talking with people. I’m fortunate to have a network where I can often chat with top experts in data and AI, which provides a lot of education that’s hard to find in books, articles, or videos. These discussions happen either during podcasts or in private conversations. There are backstories that people don’t write about or share publicly. While I’m sensitive to sharing these backstories, it helps fill the blanks where one might have questions about how things came about. While talking with peers is great, we often discuss things from the perspective of our bubble. Increasingly, in addition to chatting with my friends and colleagues, I also use AI for conversations. Chatting with AI is excellent for getting out of the bubble.
I spend a lot of time getting hands on with a topic. For data modeling, this means not just theorizing about it, but doing it. While I have a lot of hands-on work experience, there’s still always something new to learn in our field. I find the AI code-generation tools are very handy for scaffolding the databases I need for research. They’re also getting good at schema design and troubleshooting. While not perfect, AI saves a ton of time and false starts. At the same time, I feel like some of the best ways to learn are smashing your head against the wall with debugging stack trace errors.
Thinking through a topic is time consuming, and I probably spend more time on this than anything else. This might involve going for a long walk, hiking, or scribbling on a legal pad. My home is stocked with notepads and pens (usually Parker or Kaweco brands). I’ll jot down an idea and discuss my thoughts with an AI. This helps shed light on some blind spots in my thinking.
Finally, there’s physically writing. That hasn’t changed. I shoot for a specific word count daily, usually 500 to 1000. That’s around 1-2 pages, which should keep you on pace for an article or a book. Keep in mind this is often just “zero draft” material, which might not go anywhere. I’ve written thousands of pages that never see the light of day. But I do this because it keeps my mind sharp and promotes continuous thinking on a topic. I use Grammarly for final grammar edits but usually keep 50% of its suggestions. If I accept all of its suggestions, grammarly forces my writing into a homogenous and bland style. Quirks and mistakes are a feature, not a bug.
I don’t let AI write for me. I think that cheats me and the reader. It cheats me because having an AI write for me defeats the purpose of writing. Writing is thinking. Writing with one’s voice is the essence of writing. And reading the author’s thoughts and voice is the essence of reading.
Which brings me to my final point in this rant. Writing is changing. Books will change. For example, I don’t believe code-heavy technical books are relevant anymore. Things change too fast and AI does a decent job creating technical examples. Those books should instead be video courses. In some ways, I think we’ll see a new golden age of writing precisely because people will get tired of reading AI slop. More than ever, people want to read stuff with personality, by humans and for humans.
Have a wonderful weekend,
Joe
Cool Weekend Reads & Listens
Gen Z and the End of Predictable Progress - by kyla scanlon
Skype is dead. What happened? – On my Om
Hallucinations in code are the least dangerous form of LLM mistakes
The Secret Society Raising Your Electricity Bills - The American Prospect
In defense of simple architectures
The Differences between Deep Research, Deep Research, and Deep Research
Some thoughts on autoregressive models
Podcast
Freestyle Friday - How I Use AI for Writing (Spotify)
Matthew Kelliher-Gibson - The Data Cynic (Spotify)
Carly Taylor - The True Cost of Replacing Engineers with AI (Spotify)
Freestyle Friday - The Cult of Scrum (Spotify)
John Thompson - The Path to AGI, Writing Books, and More (Spotify, YouTube)
Freestyle Friday - The Great Pacific Garbage Patch of AI Data Slop (Spotify)
Eric Broda - AI Agent Ecoysystems, the Death of Consulting, and More (Spotify, YouTube)
Hugo Bowne-Anderson - Exploring the Future of AI and Automation (Spotify, YouTube)
Ghalib Suleiman & Kevin Connolly - Data Team "What Ifs" (Spotify, YouTube)
Anne-Claire Baschet & Yoann Benoit - Unifying AI and Product Development - (Spotify, YouTube)
David Jayatillake - Semantic Layers, Proving Value in Data Work, and More - (Spotify, YouTube)
Freestyle Fridays - Old School vs New School Data Modeling (Spotify)
Evan Wimpey - On Being a Data Comedian, a Bayesian, and Other Priors (Spotify)
Data Day Texas Recap w/ Tony Baer, Matt Housley, and Juan Sequeda (Spotify)
Freestyle Fridays - Unhinged Rants w/ Carly Taylor (Spotify)
Way more episodes over at the Joe Reis Show, available on Spotify, Apple Podcasts, or wherever you get your podcasts. It will soon be available on YouTube.
Upcoming Events
Atlanta, HEDW Conference. Data Engineering Workshop - April. Register here.
Las Vegas. Google Next - April TBA
San Francisco - April TBA
Deep Dish Data w/ Matillion - April TBA
London - May TBA
Snowflake and/or Databricks - June TBA
Iceland - June TBA
Australia, Data Eng Bytes - July TBA
Big Data London - September TBA
Helsinki Data Week - October TBA
More to be announced soon…
Thanks! If you want to support this newsletter
The Data Engineering Professional Certificate is one of the most popular courses on Coursera! Learn practical data engineering with lots of challenging hands-on examples. Shoutout to the fantastic people at Deeplearning.ai and AWS, who helped make this a reality over the last year. Enroll here.
Practical Data Modeling. Great discussions about data modeling with data practitioners. This is also where early drafts of my new data modeling book will be published.
Fundamentals of Data Engineering by Matt Housley and I, available at Amazon, O’Reilly, and wherever you get your books.
The Data Therapy Session calendar is posted here. It’s an incredible group where you can share your experiences with data - good and bad - in a judgment-free place with other data professionals. If you’re interested in regularly attending, add it to your calendar.
My other show is The Joe Reis Show (Spotify and wherever you get your podcasts). I interview guests on it, and it’s unscripted, always fun, and free of shilling.
Want me to speak at your event? Please submit a speaking request if you want me to speak or give a workshop at your event.
If you’d like to sponsor my newsletter or podcast, please reach out to me.
Be sure to leave a lovely review if you like the content.
Thanks!
Joe Reis
"Writing is thinking" is exactly why I feel that using AI generated content in school should be considered cheating. Especially in school, the final output is merely an artifact that represents the real goal; the acquisition of new knowledge or new skills.
“Writing is thinking”. This is so true. It’s when I start writing my articles where I start scratching my head saying “wait a second…this hardly makes any sense to me. How can I frame this to make sense to others”.
True growth on a subject comes from taking the time to explain it to others as if they have never heard of it before.