🌍 Dev.to

Building a Schema With an AI Agent Without Naming a Single Column

Ask an AI agent for a discussion-board schema and you'll get one in about four seconds. Here is the part nobody mentions: you'll also get is_deleted in this table and del_yn in the next one, because the model is pattern-matching against every schema on GitHub, and those schemas disagree with each other and with you. The fix is not a better prompt. It is giving the agent the same thing you'd give a new hire — the list of words your team actually uses — and making it impossible to ignore. ...
🌍 Dev.to

A Rule Can Be Specific and Still Be Too Broad

Update — v0.2.0 released. CauterRule is now live on GitHub and PyPI . It turns repeated agent failures into permanent standing rules — extract, replay-test, promote. pip install cauterule gives you the full CLI, TUI review, observability, 7 export formats, adversarial corpora, and a bundled git rule pack. The field test report evaluated 4 models across 745 trajectories and is the source for every number in this article. Release notes · Changelog CauterRule is an open-sour...
🌍 Lobsters

A decade of rustls

Context: rustls is the TLS implementation in Rust. Comments
🌍 TheVerge

Apple’s foldable ‘iPhone Duo’ will reportedly start at $2,000

With only a few hours left until the first iPhone launch event since John Ternus took over as the CEO of Apple, Bloomberg reporter Mark Gurman says the name of the company's long-rumored folding phone is the iPhone Duo. The midnight report suggests it will have a price starting at $2,000 for a model with […]
🌍 Dev.to

What AI Engineer Job Ads Actually Screen For

Decide to become an AI engineer and the first thing you meet is a framework list. Learn LangChain? People say LangGraph is rising. What about CrewAI, AutoGen? Picking wrong feels like it will cost you a year. The posting data settles that anxiety first. An agent-specialist job board tallied 1,135 of its own listings from March–May 2026 (the site doesn't name its operator, so read it as a snapshot). More than half of those ads name no framework at all. The ads that do name them average 2.31 ...
🌍 Dev.to

Be Confident in How You Work, Not What You Know

Ground your confidence in how you work, not in what you currently know. Knowledge based confidence is a trap, and it gets more dangerous the better you get. You learn a stack deeply, you become the person everyone asks, and your sense of being competent quietly fuses with that specific body of facts. Then the company changes direction, or the tool is replaced, or you move teams, and your confidence walks out of the door with the thing you knew. You have seen this happen to people with fift...
🌍 Dev.to

Writing Maintainable Code in Large Teams: Principles, Patterns, and Pipeline Architecture

Writing code that works is relatively straightforward. Writing code that can be safely modified, extended, and maintained by a team of 50+ engineers across multiple time zones without breaking production is an entirely different discipline. As engineering organizations grow, the primary bottleneck stops being individual developer speed and shifts to cognitive load, coordination overhead, and architectural drift . When every team member implements features using their own preferred style ...
🌍 Dev.to

We Renewed That Contract Because Nobody Had Time to Read It

The renewal notice arrived sixty days before expiry, as the contract required. It went to a shared mailbox, was read by somebody who assumed the service was still needed, and was approved on the grounds that it had been approved last year. The cost had risen nineteen percent. We were paying for four hundred and fifty licences and using a hundred and ten. And the agreement had quietly acquired, in a schedule appended two renewals earlier, a clause that made the price of leaving considerably highe...
🌍 Dev.to

Keep the Feedback You Disagreed With

There is feedback you were given once that you decided was wrong. Do not throw it away. It happens to everyone. Someone tells you that you are too blunt in reviews, or too slow to commit to a decision, or that you go too deep into detail in meetings. You think about it honestly and you conclude they have misread you. Maybe they had. Maybe they were measuring you against themselves, or they caught you on your worst week, or they were repeating something they heard secondhand. So you set it ...
🌍 Dev.to

The canary that never saw the traffic that would have broken it

We were proud of the canary. Five percent of traffic for fifteen minutes, automated analysis comparing error rate and p95 latency against the baseline pods, promote or roll back with no human in the loop. It had caught four bad releases in six months. Then it promoted a release that corrupted nine hours of the nightly settlement run, and the canary's verdict had been green with wide margins. The release changed how a shared serialiser handled a nullable decimal. The synchronous API almost nev...
🌍 Dev.to

How Does a VPN Work? A Simple Guide

How Does a VPN Work? A Simple Guide You have probably heard about VPNs many times. People use VPNs for privacy, security, remote work, gaming, streaming, and accessing websites that may not be available on their network. But how does a VPN actually work? Let's explain it in simple words. What Happens Without a VPN? When you use the internet normally, your device connects to your internet provider first. For example: Your computer → Internet Provider → Website ...
🌍 Dev.to

The Work Juniors Used to Learn On Is Disappearing

The tasks that used to turn beginners into engineers are now being handed to machines first. Think about how you actually learned. Not the courses. The grind. The small bug nobody senior wanted, the tedious form, the test suite you wrote because somebody had to, the config you fixed by trying things until it worked. That work was boring and badly paid and it was also the entire apprenticeship. You were learning to read errors, to navigate a codebase you did not write, to feel when something w...
🌍 Dev.to

Building a Daily Interactive Map for the PEAK Game

A game map is easy to underestimate. At first glance, it looks like a large image with a few pins placed on top. In practice, a useful daily map has to answer several questions quickly: Which level is active? Which marker matters? Is the objective close to the required route? Can a player inspect the terrain on a phone without losing the map? That problem is what led to Peak Map Today , an independent fan-made map for the PEAK game. The site is designed around a simple rule: a player who sea...
🌍 Dev.to

We Tried to Make a Voice Agent Control a Real Device. Here's What Broke

Most voice-agent demos have a simple loop: The user speaks. The model thinks. The agent replies. That loop works well when the agent only needs to answer a question. It becomes much harder when the agent is also controlling a real device. Aiden can read a phone’s screen and operate it through an external hardware control path. A task might involve several screenshots, tool calls, taps, swipes, and waits for the device to respond. While that is happening, the person may sti...
[Boost]
🌍 Dev.to

[Boost]

I got tired of opening ports just to query my own database Abdullah Kaya Abdullah Kaya Ab...
🌍 Dev.to

Your Guardrail Test Still Passes With the Guardrail Deleted

To test whether an AI agent guardrail is still working, run your labelled corpus twice: once as built, once with the gate replaced by a pass-through. ablation_probe.py sorts each bad input into LOAD_BEARING , SHADOWED or MISSED , and exits 1 when no verdict moved. A rejection shows that something said no. It does not show that the gate did. Here is the shape of the problem. You have a policy gate. You wrote a fixture that it is supposed to reject. CI runs the fixture, the pipeline says ...
HackerNews

I resigned from Anthropic today

https://xcancel.com/hilbertspaess/status/2097476196791709843... Comments URL: https://news.ycombinator.com/item?id=49619227 Points: 473 # Comments: 614
HackerNews

AI Has a Discovery Problem

Article URL: https://mhacevedo.com/posts/the-discovery-problem Comments URL: https://news.ycombinator.com/item?id=49621223 Points: 29 # Comments: 9
HackerNews

The Navier–Stokes Millennium Prize Problem

Article URL: https://simonwillison.net/2026/Sep/8/on-navier-stokes/ Comments URL: https://news.ycombinator.com/item?id=49621697 Points: 151 # Comments: 127
HackerNews

"Tweet" and the bird logo apparently enter the public domain

Article URL: https://blog.ericgoldman.org/archives/2026/09/tweet-and-the-bird-logo-apparently-enter-the-public-domain-but-x-maintains-its-grip-on-the-twitter-mark-for-now-x-v-project-bluebird.htm Comments URL: https://news.ycombinator.com/item?id=49621751 Points: 33 # Comments: 4