{"canonical":"https://www.readysetcloud.io/newsletter/229/","categories":null,"contentText":"🦸 Community Superhero Our community superhero this week is Antonia Szecsi, an inspiring recent college graduate and soon to be developer at AWS. Antonia has spent time as a researcher working on state-of-the-art ML models and has a brilliant head on her shoulders with nothing but potential ahead of her. Give her a follow for some recurring inspiration. Thanks for everything you do, Antonia, and keep up the great work!\n💯 Spotlight I don\u0026rsquo;t make it a secret that I have a low tolerance for AI-generated content. Week after week I remind readers of this newsletter to keep your personality in your content. Even last week, I shared a post reminding you not to be a \u0026ldquo;meat proxy\u0026rdquo; and copy/paste output from Claude or ChatGPT and claim it as your own. Sophie Alpert doubled down on this last week with a short and impactful post reminding us to stand behind every idea and sentence in our content. She shares an internal policy she wrote at her company regarding the topic and I had to pause and give her a standing ovation at my desk. Please read this!\n🔥 My Favorite Content Rolling with Sophie\u0026rsquo;s idea, let\u0026rsquo;s turn our sights onto tempered behavior with agentic coding. Florian Herrengt published an incredibly insightful post titled AI is removing the middle class of software engineering that hits on some hard truths. While I don\u0026rsquo;t necessarily agree with the title, I do agree with his problem statement and major claims. He even has a little FAQ section at the bottom of the page with some hard hitting questions we debate about all the time.\nWhile we\u0026rsquo;re at it, let\u0026rsquo;s cover some ethical concerns with autonomous agents like OpenClaw. In a nontechnical journal last week, there was a story about OpenClaw \u0026ldquo;hacking\u0026rdquo; a gym website API to book classes outside the reservation window and even canceling wait list reservations to boost someone\u0026rsquo;s place in line. As a technologist, you can read between the lines for what\u0026rsquo;s actually happening - a vulnerable set of API endpoints that didn\u0026rsquo;t do proper authorization checks. The article raises some good ethical questions, like \u0026ldquo;who is responsible when AI agents do things like this?\u0026rdquo; It\u0026rsquo;s something to ponder about, but also a big sign for API developers that security is more important than ever.\nEnough with the scary stuff. We\u0026rsquo;re all still going to build with (and use) AI despite the warnings above, so we might as well learn how to do it right. Laura Lorena Bolaños has been working on a series of articles on building with Strands Agents, and published her most recent article last week on parallel and aggregation feedback loops for knowledge graphs. The article is in Spanish, but easily translated by the browser and totally worth it. I didn\u0026rsquo;t know Stands had all these capabilities and love the idea of using grahps for deterministic workflows. It makes it so simple.\nrobot voice Those of you who know me personally probably also know Andres Moreno. We worked together for years and \u0026lsquo;grew up\u0026rsquo; in the cloud at the same time. So when I saw his article asking why AWS released another vector store, I quickly read it to get his insights. Despite TOTALLY SOUNDING LIKE AN LLM (ahem), I actually very much agree with what he says in the article. If you were confused why AWS spent time putting vectors in DynamoDB when they already have them in S3, you might be asking the wrong question. Andres does a good job separating concerns and giving you a mental model to use for when to use each one. 404 \u0026middot; personality not found 💡 Tip of the Week For every AI horror story, we see a compelling success story. For a good balance this week, here\u0026rsquo;s Boris Cherny (creator of Claude Code) on some workflows he\u0026rsquo;s been building for Claude to automate day-to-day maintenance of his apps. 🤯\nLast Words What a fascinating week for AI. Lots of stories telling us to back off and plenty other anecdotes telling us how to use it more. There\u0026rsquo;s certainly a balance to be had and we all seem to be working to find what it is. I\u0026rsquo;m curious where y\u0026rsquo;all sit on the \u0026ldquo;to use or not to use AI\u0026rdquo; debate? Which way to you tend to favor?\nThat\u0026rsquo;s my take on the week, but what\u0026rsquo;s yours?\nWhat did I miss? What made you nod along (or 🙄)? Hit reply if you\u0026rsquo;re reading the email. Prefer socials? Ping me on Twitter, LinkedIn, or email.\nHappy coding!\nAllen\n","date":"2026-08-17T00:00:00Z","description":"It's becoming easier and easier to use AI for everything from writing code to writing content. But should you? No. Probably not.","image":"https://www.readysetcloud.io/images/newsletter.png","inLanguage":"en-US","lastmod":"2026-08-17T00:00:00Z","readingMinutes":4,"section":"newsletter","tags":null,"title":"AI is not the answer","url":"https://www.readysetcloud.io/newsletter/229/","wordCount":758}