🌍 Dev.to

Cloudflare Worker redirect never runs, because static assets are served first

I wrote a www → apex redirect in my Worker's fetch handler. It was dead code from the moment I deployed it, and I didn't notice for weeks. Matches if: your Cloudflare Worker isn't running; the fetch handler never fires; a redirect in Worker code doesn't work; console.log in a Worker produces nothing on real pages. The symptom The redirect logic is right there in the Worker. It's deployed. You can read it in the dashboard. And www.example.com/about serves the page instead of redirectin...
🌍 Dev.to

AI Agents vs Chatbots: They're Not the Same Thing

A few months ago, if someone mentioned AI, most people immediately thought of ChatGPT. Today, the conversation has changed. Everywhere you look, people are talking about AI Agents . Tech companies are building them. Startups are raising millions around them. Developers are integrating them into products. And suddenly, everyone claims that "AI Agents are the future." But here's the interesting part. Many people still use the terms chatbot and AI agent as if they mean ...
🌍 Dev.to

My Work Week Starts Before I Sit Down: Running Kiro Crew as a Technical Program Manager

The tool I have quietly run my work week on is now open source. Kiro Crew is a persistent agent workspace built on Kiro. I have been on the internal version since April, and it changed how I think about what AI at work actually is. A while back I wrote about terminal productivity as a TPM and how far the command line can take a non-engineer. This is the next chapter: the terminal now runs itself. I am a technical program manager, not a software engineer. My setup today: around a dozen...
🌍 Dev.to

S3-Compatible Storage Explained: The De Facto Standard for Object Storage in 2026

"S3-compatible" is one of those phrases that gets thrown around so much that people nod along without checking whether everyone means the same thing. Here's what it actually means, why it became the de facto standard, and what changes when you pick one implementation over another. What "S3-Compatible" Actually Means S3 compatibility means implementing Amazon's S3 REST API as a protocol — not as an AWS service you pay for. Specifically, it means supporting: The shared data m...
🌍 Dev.to

Why AI Agents Are Replacing Traditional Software

A few years ago, building software meant writing a series of rules. If a user clicked a button, the application executed a function. If a condition was true, another function ran. Everything was predictable because every possible outcome had already been programmed. That approach has worked for decades. But AI is changing the way software behaves. Instead of building applications that simply follow instructions, we’re beginning to build systems that can understand goals, make decisio...
🌍 Dev.to

Encrypting personal data at rest in Symfony: a 60-line Doctrine type, and the four things it breaks

A review form asked me a yes/no question I could not answer with yes: do you encrypt personal data at rest and in transit? In transit, sure, TLS everywhere, that part is free in 2026. At rest, no. The app runs on a plain VPS with no full-disk encryption, and the database held customer names, addresses and email addresses in a json column, in clear, exactly as PostgreSQL received them. There are two ways out of that question. One is to click yes and move on, because everybody knows the ...
🌍 Dev.to

Your Cloud Has Attack Paths No Scanner Can Find. Because They Don't Exist in Any Single Resource

✓ Human-authored analysis; AI used for formatting and proofreading. Most cloud security tools work by checking individual resources against a library of known-bad patterns. Is this bucket public? Is this port open? Is this key unrotated? These checks are necessary. They're also insufficient. Because the most dangerous violations don't exist in any single resource. They exist in the RELATIONSHIPS between resources. They're detectable from the configuration structure alone, without any use...
🌍 Dev.to

5 Things Developers Get Wrong About Securing AI Agents

If you've shipped an AI agent to production in the last year, you've probably already felt the gap between "this works in the demo" and "this is safe to run unsupervised." Agentic AI systems that plan, call tools, and take actions with minimal human input breaks a lot of assumptions that traditional app security is built on. Here are five mistakes developers keep making, and what to do instead. 1. Treating the agent's credentials like a service account It's tempting to give you...
🌍 Dev.to

OpenAI Just Solved a Problem Open Since 1999. It Still Can't Ask Its Own Question.

Four days after I published a piece arguing LLMs can't make the jump, OpenAI announced that an internal model called Astra had solved ten open problems in mathematics and theoretical computer science. One of them had been open since 1999. I'm not going to pretend that's a comfortable coincidence to sit with. So let's sit with it properly instead of pretending it didn't happen. The headline result is a non-sofic group. Mikhail Gromov introduced the concept of soficity in 1999 and asked...
🌍 Dev.to

The LLM is an ALU - Lessons from a ZX Spectrum

Every time your LLM agent calls a tool and reads the result, you're paying a tax. Not in dollars — in tokens, latency, and reliability (ok, dollars then). The tool's output gets serialized into text, tokenized into the context window, attended to by the model, and then re-serialized when the model decides what to do next. If the next step is another tool call, the cycle repeats. Data that never needed the model's judgement flows through the most expensive component in your system — twice. Thi...
🌍 Dev.to

Why Publishers Are Blocking AI Crawlers and Reshaping Access to Web Content

Major publishers have increasingly restricted AI crawler access to their websites, turning a once largely technical robots.txt decision into a strategic question about training data, licensing, and control of digital content . The shift does not mean publishers have adopted one common position on artificial intelligence. It does mean that AI companies can no longer assume that publicly reachable web content is available for automated collection. One early, high-profile example came in Sept...
HackerNews

My thoughts on the foreign policy AI evaluation gap

Article URL: https://www.karlsnotes.com/my-thoughts-on-the-foreign-policy-ai-evaluation-gap/ Comments URL: https://news.ycombinator.com/item?id=49180466 Points: 1 # Comments: 0
HackerNews

AI's Arctic Land Grab

Article URL: https://www.arctictoday.com/ai-arctic-land-grab/ Comments URL: https://news.ycombinator.com/item?id=49180508 Points: 2 # Comments: 1
HackerNews

The Economist Problem and AI

Article URL: https://pithandpip.com/blog/the-economist-problem-and-ai Comments URL: https://news.ycombinator.com/item?id=49180573 Points: 1 # Comments: 0
HackerNews

AI Agents for Logistics, Pitfall?

Recently we've all seen the chart showing different industries and the amount of AI solutions being created for it in percentages. I remember that something that stood out was that the supply chain and transportation industry had less than 2%. Now recently Happy Robot reached unicorn status and they started as AI voice agents but for almost a year now they've rebranded completely to "superintelligence" (among other similar iterations before they landed on that word). But something here doesn't ...
HackerNews

What AI doesn't say about AI

Article URL: https://claudiocaletti646669.substack.com/p/what-ai-doesnt-say-about-ai Comments URL: https://news.ycombinator.com/item?id=49180923 Points: 2 # Comments: 1
HackerNews

Trump's AI protectionism has come for robotics

Article URL: https://www.technologyreview.com/2026/08/03/1141056/trumps-ai-protectionism-has-come-for-robotics/ Comments URL: https://news.ycombinator.com/item?id=49181297 Points: 1 # Comments: 0