In my latest blog post roundup, I dive into a treasure trove of insights and breakthroughs across various fields. From future-proofing your business with Jessica Norell Neeson’s strategies, decoding the art of perfect prompts for large language models, to the intriguing world of monorepos and their use in tech giants like Figma and Spotify. I also explore the cutting-edge application of AI in data journalism, share the journey of building Bluesky—a novel distributed social network, and offer personal revelations on overcoming the dreaded idea drought. Additionally, I discuss the hard-learned software design principles, demystify the complex realm of recommendation systems, reveal the underbelly of AI’s impact on social media, chuckle at the whimsical units of measurement, and ponder the privacy-centric future of home assistants. Join me as I sift through these fascinating topics, shedding light on innovation, challenges, and a bit of humor in the tech world.

• Future-proof your business: No matter what you do or which sector you’re in, change is constant and will only continue. The winners will be the businesses with the mindset to embrace transformation, says Jessica Norell Neeson.
• The Perfect Prompt: A Prompt Engineering Cheat Sheet: Large language models can produce any sequence of characters. Literally any. In any idiom, data format or programming language.
• 🧠Knowledge Series #32: What’s a monorepo?: Everything you need to know with insights from Figma, Spotify, Airtable and more
• AI for Data Journalism: demonstrating what we can do with this stuff right now: I gave a talk last month at the Story Discovery at Scale data journalism conference hosted at Stanford by Big Local News. My brief was to go deep into the …
• Building Bluesky: a Distributed Social Network (Real-World Engineering Challenges): Bluesky is built by around 10 engineers, and has amassed 5 million users since publicly launching in February this year. A deep dive into novel design decisions, moving off AWS, and more.
• What To Do When You’re Out of Ideas: I’ve only been writing online as a marketer for about four years. And already, I feel like I’m out of ideas. In this photo I am holding a light bulb but
• 4 Software Design Principles I Learned the Hard Way: If there’s two sources of truth, one is probably wrong. And yes, please repeat yourself.
• Learning RecSys through Papers- Implementing a Candidate Generation Model: Sometimes I’m asked how to start learning about recommender systems, and there’s a few go-to papers I always mention; however, without a proper map, it could be alittle difficult for the uninitiated. So, to try to make a gentle introduction, I will walk through an implementation of the candidate generation model based on Deep Neural Networks for YouTube, which I will sometimes refer to as the “Deep Nets” paper. This paper (and its talk) is jam-packed with all sorts of interesting practical recommendation system knowledge and is the perfect starting place for people hoping to understand large-scale recommendation systems. I will implement the key components of the candidate generation model and train the model on the MovieLens dataset in PyTorch, a typical benchmark dataset in the RecSys space. There will also be a few modern flourishes and comments on new developments as applicable. I’ll conclude with outlining a natural extension to this approach to predict multiple outcomes.
• AI Is Poisoning Reddit: A market for manipulating Reddit using AI have emerged.
• ‘Butt’ is an actual unit of measurement: The imperial system is a funny thing. Like, really funny. So funny that there is actually a unit of measurement for wine (or whiskey) casks called a “butt.” That means if you fill the barrel up, you technically have a buttload of wine—though you’d probably just call it a full butt. Are you laughing yet?
• Learning RecSys through Papers- Implementing a Candidate Generation Model: This paper (and its talk) is jam-packed with all sorts of interesting practical recommendation system knowledge and is the perfect starting place for people hoping to understand large-scale recommendation systems.
• Home Assistant has a new foundation and a goal to become a consumer brand: Can a non-profit foundation get Home Assistant to the point of Home Depot boxes?
• Tales from the .NET Migration Trenches – Authentication: Of all the topics in .NET migration, authentication, like always, is the one that is most characterized by “It Depends”. The solution for addressing authentication is wholly dependent on what the current authentication solution is in the current .NET 4.8 application. If you’re doing external SSO, then it’s likely quite simple – the new solution is simply a new client for your external SSO.