The Next Era of AI for Software Engineers
A clearer focus, a new AI Field Guide, and updated benefits
For the past ~3 years, AI for Software Engineers has been my forcing function to learn new topics, share ideas, and meet new people. It’s been more of an experiment than a serious project and one that has paid dividends.
We now have a community of nearly 13,000 readers with more than 100 of you paying to support newsletter operations. Thank you for your support!
Over the past few months, I’ve been doing some newsletter housekeeping to decide what the future of AI for Software Engineers should be. I’ve run some experiments and analyzed the numbers on my articles and posts to better understand what you’re all looking for.
The primary takeaway from my analyses is that my audience tends to appreciate deep, high-quality technical education resources (which is rare online—so kudos to all of you!). This is also what I’m looking for, so here are some changes to AI for Software Engineers to better align with that:
Articles will focus on developing the intuition needed to work in AI. This means more of an infrastructure, systems, and engineering focus.
Paid support will continue to be $5/mo with supporting the newsletter as the primary motivator.
On top of this, I’m introducing The AI Field Guide, a more structured and curated version of the resource formats I’ve used in the past.
Paid subscribers will also get exclusive offers to other technical AI Substacks and educational resources (the first is listed below!). More specifics on this below.
I’ll be using the Substack Chat for giveaways, so make sure to join the Chat to be part of those.
More detail below.
Reframing around intuition
You might have noticed that the tagline for AI for Software Engineers has changed recently:
“Build intuition for production AI. Infrastructure, systems, operations, and engineering tradeoffs.”
One of the most difficult things about system design is having a level of intuition behind how different technologies work and where and when to apply them. In AI, the intuition required to build successful systems is even higher. It requires traditional system design thinking and an understanding of a rapidly evolving field. Thus, AI has a steep barrier to entry.
Much of this barrier is the difficulty of not only knowing how to learn something, but also knowing what to learn. Fortunately, when I was getting into AI I had a research advisor and multiple other research assistants to guide me. Without that, knowing what to learn and how to learn it is difficult.
Thus, I’m going to spend more time discussing the underpinnings of AI from an engineering perspective and will go deeper into the infrastructure, operations, engineering tradeoffs, and more that makes AI work in production. This isn’t a huge departure from what I’ve been writing about, but I’ll be spending more time thinking about what the biggest engineering topics are related to a specific topic within AI and answering those.
Larger, more complex subjects will make it into articles, and smaller subjects will be shared as posts on social media (don’t forget to find me on Notes, X, and LinkedIn).
I’m excited about this change because, for a long time, I’ve struggled to put into words what I tend to write about. It’s always just whatever I think is important at the time. Building intuition summarizes it nicely.
Introducing The AI Field Guide
I’m introducing a new perk for subscribers I like to call The AI Field Guide. This will be a collection of unusually good AI technical resources that will help you gain the intuition needed to succeed as an engineer in AI. These resources won’t just be explainers, but resources that help understand deep technical concepts in a way that makes them more applicable. These resources will supplement my articles in areas I otherwise can’t. I’ll send these out with commentary to enable you to get the most from them.
These articles will always go out to everyone when they’re released, but will be paywalled when they become part of the archive (after about 2 weeks). Paid subscriber access will remain forever. I plan to release one each week toward the end of the week to provide you with a resource worth your time to look over each weekend if the topic interests you.
I’ve tried many different formats to keep everyone up-to-date on interesting reads and learning resources each week (weekend reads, reading lists, weekly roundups, etc.) and this is the culmination of all of those in a format that makes them easier to digest and higher signal.
The ML roadmap still exists and has been moved to mlbasics.com (the old URLs still work, too). This roadmap will complement the Field Guide and will focus more on ML fundamentals and the skills needed to learn them (foundational skills such as Python, Bash, Git, etc.).
A well-kept resource like this is something I’ve wanted to put together for a while. I think it will be beneficial to engineers of all experience levels, and I see this being a multi-year effort to ensure comprehensiveness.
Additional paid benefits
A big part of the housekeeping I’ve been doing has been figuring out how to make paid subscriber benefits more useful while still keeping technical learning free. AI for Software Engineers was created on the idea that great technical education should largely be free (back when it was called Society’s Backend), and that’s still a goal of mine.
Because of that, I’m being careful with partnerships and sponsorships. I want to make excellent technical resources more available to you but I don’t want to sacrifice newsletter quality to do so.
So, starting now, I’ll be partnering with other incredible technical learning Substacks to bring you discounts or free trials to make more resources available to you. To start, all paid subscribers of AI for Software Engineers get 90 days free to Ahead of AI. Head over to the “Get More” page to redeem yours! As more offers become available, I’ll let you know and they’ll be listed on that page.
I’ll also be doing giveaways in our Subscriber Chat for free trials of Substacks for those who want them. Stay tuned there if you’re interested. Preference will be given to paid subscribers, but so far many free subscribers have been able to get them, too!
Don’t forget: You can likely expense a paid subscription to your employer and students can get a subscription for a heavily discounted rate (send me a message).
A more consistent look and easier navigation
I’ve also updated some thumbnails and visuals to have a more consistent branding. This way, if you see an image from AI for Software Engineers you’ll know it’s me.
I’ve also updated the aiforswes.com homepage to make it easier for you to find the information you need. The biggest update is a “Start Here” page that makes it easier to find whatever you’re looking for.
I’m super grateful for all of you welcoming me into your inbox each week. I’m excited to see how AI for Software Engineers will grow even more.
If you have any questions or input, drop a comment below.
Don’t forget, you can also support the newsletter and rep the duck by purchasing newsletter swag.
Thanks for reading!
Always be (machine) learning,
Logan