Hey there! This blog post explores a range of fascinating topics, from leveraging local AI for podcast transcription and the history of a hard-working Manhattan font, to the future of content discovery and AI’s impact on technical debt. Dive into stories about the UX of AI assistants, strategies for newsletter monetization, and how multitasking may be affecting our brains. Plus, there are insights into DIY projects like building a wireless keyboard and the effects of caffeine on the brain. Enjoy!

• Blaizzy/mlx-audio: A text-to-speech (TTS) and Speech-to-Speech (STS) library built on Apple’s MLX framework, providing efficient speech synthesis on Apple Silicon.:
• How I Automated My Podcast Transcript Production With Local AI: I’ve been running a podcast for close to half a decade now, called The Work Item. Publishing new episodes generally takes a bit of time because of all the prep work that needs to happen beforehand. I now get to use AI to automate a pretty tedious part of the process.
• The hardest working font in Manhattan: A story of a 150-year-old font you have never heard of – and one you probably saw earlier today.
• The Demoralization is just Beginning: This is a map of primary trading partners, US vs China, and how it has evolved over the last 20 years. Think about it, and realize this probably reflects your experience.
• Bayleaf Wireless Keyboard Build: Read about my journey building a wireless ergonomic keyboard from scratch.
• The Deep Research problem — Benedict Evans: OpenAI’s Deep Research is built for me, and I can’t use it. It’s another amazing demo, until it breaks. But it breaks in really interesting ways.
• The all-new Alexa+ and more: All the news from Amazon’s 2025 devices event: Catch up on everything we announced at our Devices & Services event, including the new Alexa+.
• Hallucinations in code are the least dangerous form of LLM mistakes: A surprisingly common complaint I see from developers who have tried using LLMs for code is that they encountered a hallucination—usually the LLM inventing a method or even a full …
• Should managers still code?: Ah, the eternal question, straight from the mailbag.
• Deep: The UX of AI Assistants: Your ultimate guide to how top companies are designing the latest generation of AI Assistants
• How to Monetize Your Newsletter With Ads: Here are a few monetization suggestions for ad formats, rates, and tools that can help your newsletter ad strategy.
• How Katelyn Bourgoin Built an Audience of 220k Teaching People “Why We Buy” – Growth in Reverse: Katelyn Bourgoin writes the Why We Buy newsletter, where she shares fun and actionable tips on buyer psychology.
• How To Be Antifragile In The Age of AI – Leadership Minute: It’s important to understand AI’s likely disruptions. But certain mindsets, skills and professions will remain resilient during the changes of AI.
• Every Entrepreneur Needs to Unplug to Unlock Peak Productivity and Creativity — Here’s Why | Entrepreneur: A digital fast may sound impossible — especially for entrepreneurs. But as painful as it may seem to miss out on an important memo or viral meme, the benefits are even greater.
• How Multitasking Drains Your Brain: Renowned neurologist Richard Cytowic exposes the dangers of multitasking in the digital age.
• Your PR Process Is Killing Morale and Productivity: Code reviews are supposed to be about improving code quality and sharing knowledge, but when PR comments start hitting the hundreds, something is very wrong….
• 10-Step Anti-Procrastination Checklist: Despite recent strides in my productivity habits, I still catch myself procrastinating at work more often than I’d like. It’s not that I make a cons…
• ARC-AGI Without Pretraining:
• Apple’s Software Quality Crisis: When Premium Hardware Meets Subpar Software: Tim, I think we are not cooking
• Write to Escape Your Default Setting:
• Make Any File a Template Using This Hidden macOS Tool: Stationery Pad is a handy way to nix a step in your workflow if you regularly use document templates on your Mac. The long-standing Finder feature…
• How AI generated code compounds technical debt: GitClear’s latest report exposes rising code duplication and declining quality as AI coding tools gain in popularity.
• How your morning coffee is changing the structure of your brain: A novel study testing the effects of caffeine on the human brain found daily consumption can significantly reduce the volume of one’s gray matter. Whether this is a good or bad thing is unclear but that daily cup of coffee is certainly doing something.
• What’s the Goal of the Goal?: When we lock into a particular goal too quickly, we blind ourselves to alternate routes forward that might have been better and easier. We can avoid this trap by asking ourselves one simple question.
• Trump credit card freeze crippling cancer and Alzheimer’s research: On February 26, President Trump issued an executive order expanding the power of the United States DOGE Service (DOGE), the government initiative controlled by Elon Musk.
• 9 useful AI prompts ❤️: Discover 9 useful AI prompt templates for ChatGPT, Claude, Copilot and Gemini. Overcome writer’s block, edit well, and plan projects with these formulas.
• Creating a Sense of Stability: How to help your team stay calm when things feel rocky
• 20 years working on the same software product: I released version 1 of my table seating planning software, PerfectTablePlan, in February 2005. 20 years ago this month. It was a different world. A world of Windows, shareware and CDs. A lot has changed since then, but PerfectTablePlan is now at version 7 and still going strong. PerfectTablePlan v1 PerfectTablePlan v7 I have released…
• An Expected Goals on Target (xGOT) Metric as a New Metric for Analyzing Elite Soccer Player Performance: Introduction: Football analysis is an applied research area that has seen a huge upsurge in recent years. More complex analysis to understand the soccer players’ or teams’ performances during matches is required. The objective of this study was to prove the usefulness of the expected goals on target (xGOT) metric, as a good indicator of a soccer team’s performance in professional Spanish football leagues, both in the women’s and men’s categories. Method: The data for the Spanish teams were collected from the statistical website Football Reference. The 2023/24 season was analyzed for Spanish leagues, both in the women’s and men’s categories (LigaF and LaLiga, respectively). For all teams, the following variables were calculated: goals, possession value (PV), expected goals (xG) and xGOT. All data obtained for each variable were normalized by match (90 min). A descriptive and correlational statistical analysis was carried out. Results: In the men’s league, this study found a high correlation between goals per match and xGOT (R2 = 0.9248) while in the women’s league, there was a high correlation between goals per match (R2 = 0.9820) and xG and between goals per match and xGOT (R2 = 0.9574). Conclusions: In the LaLiga, the xGOT was the best metric that represented the match result while in the LigaF, the xG and the xGOT were the best metrics that represented the match score.
• AI coding tools are quietly reshaping software development: They’re making an economic impact, despite not being very good yet
• 10 Things I Learned from Sending an Email Every Day for a Year: It takes work to send an email every day — but it also opened new doors for my business. Here’s what I learned from the process.
• Robotics for software engineers: What does it take to build and program robots? A look into the exciting, increasingly popular field of robotics. Guest post by humanoid robot expert, Sandor Felber
• Wow, have you seen the stock market lately?:
• How Do You Plan for a Future That Might Not Exist?: What collective grief looks like.
• How dismantling silos gave Metro’s newsletter team room to experiment: Metro’s Newsletter Editor Sophie Laughton explains how a behind-the-scenes reshuffle enabled its newsletter team to find new opportunities.
• Should we stop AGI before it destroys us, or just naively hope for the best?: AGI has lots of promise, but there are also a lot of threats. Are we blithely walking into a future we won’t be able to control?
• How digital natives changed the game on legacy media: TikTok, YouTube and Substack have rewritten the rules of media, leaving legacy outlets scrambling to survive.
• New Neuroscience Reveals 3 Secrets That Will Make You Happier: Your brain doesn’t want you to be satisfied. Seriously. But we can trick it into making you happier. Here’s how…
• The Pentium contains a complicated circuit to multiply by three: In 1993, Intel released the high-performance Pentium processor, the start of the long-running Pentium line. I’ve been examining the Pentium’…
• macOS Tips & Tricks: A collection of 150+ nifty tips and tricks for macOS, for both system features like the Dock, Spotlight, and Mission Control and built-in apps like Safari, Mail, and Photos.
• nRootTag – Tracking You from a Thousand Miles Away!: Tracking You from a Thousand Miles Away! Turning a Bluetooth Device into an Apple AirTag Without Root Privileges
• Fine tuning in Apple MLX, GGUF conversion and inference in Ollama?:
• Opportunities and perils in the coming content earthquake: the future of content discovery: Content discovery is about to go through a remarkable transformation as AI agents take over the work. Goodbye search. Goodbye websites.
• New Jersey Newsrooms Could Be Hit By Walkout: Publishers In Action: They’re Exploring AI And Ways To Grow Revenue – 02/25/2025
• Hard problems that reduce to document ranking: There are two claims I’d like to make: LLMs can be used effectively1 for listwise document ranking. Some complex problems can (surprisingly) be solved by transforming them into document ranking problems. I’ve primarily explored both of these claims in the context of using patch diffing to locate N-day vulnerabilities—a sufficiently domain-specific problem that can be solved using general purpose language models as comparators in document ranking algorithms. I demonstrated at RVAsec ‘24 that listwise document ranking can be used to locate the specific function in a patch diff that actually fixes a vulnerability described by a security advisory, and later wrote on the Bishop Fox blog in greater defense of listwise ranking by publishing a command-line tool implementation (raink) to prove the idea.
• A leap forward for AI in journalism from an unexpected place — small-town Missouri:
• A system to organise your life: Johnny.Decimal is a system to organise your life. Find things, quickly, with more confidence, and less stress. It’s free to use and the concepts are the same at home or work.