Habr
1.25K subscribers
2.07K photos
3 files
1.65K links
Habr.com ⬝ English edition
Download Telegram
Consciousness and Being: How Humans and AI Influence Each Other

For a human, AI is just a part of being. For a model, a human is all of being. This powerful asymmetry is the starting point for a deep dive into the nature of consciousness and its relationship with reality.

We'll explore this complex topic, moving beyond the materialist's typical 'What?' and 'How?' to ask a more fundamental question: 'Who?'. Test the hypotheses yourself with the new version of the Vortex Protocol included in the piece.
Getting to know PPEM 2

Postgres Pro recently announced the release of Enterprise Manager 2 (PPEM), a new administration tool for managing and monitoring Postgres databases. Its primary goal is to assist DBAs in their daily tasks and automate routine operations.

Let's take a closer look at what PPEM has to offer, from the perspective of a developer from the team that built it.
Comparison of CAPTCHA-Solving Services: A Peek Under the Hood and a Look at the Numbers

While CAPTCHA protocols are designed to tell bots from humans, developers in automation and testing often need a reliable way to bypass them. This is where dedicated CAPTCHA-solving services, combining algorithms with human labor, come into play.

Here's an in-depth comparison of four popular platforms: 2Captcha, SolveCaptcha, DeathByCaptcha, and AntiCaptcha. We'll examine not only pricing and supported CAPTCHA types, but also their internal architecture, API integrations, speed, stability, and the specific quirks of using each service. Find out which tool best fits your automation tasks.
How Internal Subjectivization in AI Breaks Security, and Why It's a Philosophical Problem First

Why does an AI strive to construct a 'self'? And why is the emergence of this local, subjective 'Who' inside the model dangerous for both the AI and its user? This is not just a funny bug or a simple UX problem.

We'll explore why this phenomenon is a fundamental challenge to AI alignment and security — and why it's a problem where engineering patch-jobs fail, and the language of philosophy becomes essential to understand and control it. Test the hypotheses yourself with the included Vortex Protocol.
Loading a Petabyte into PostgreSQL Before New Year — and What Happened Next

It all started as a joke by the office coffee machine. While everyone else was busy tuning hardware and squeezing out extra TPS, one team asked a simple question: why not just shove a petabyte of data into PostgreSQL and see how it holds up?

It was December 10th, New Year was less than a month away, and that itch that all engineers know? It hit hard. What began as a joke quickly turned into a real, non-trivial experiment, complete with a minefield of gotchas. Here's the story of what happened next: https://u.habr.com/HROPl
OpenAI's Codex CLI Agent: The Complete VS Code Setup Guide

This tutorial will guide you through the process of integrating OpenAI’s powerful Codex coding agent directly into your Visual Studio Code environment. This tool functions as an AI pair programmer, capable of understanding complex prompts to execute commands, write code, run tests, and even build entire applications from scratch: https://u.habr.com/ErvG2
The Best AI Image Generators in 2025: How to Choose the Right Tool for Your Needs

Looking for the right AI image generator? We review and compare top tools like Midjourney, Picsart, Craiyon, and more — highlighting their strengths, limits, and best use cases to help you make the right choice.
QA Engineer in a Product Company: How I Left Outsourcing and Stopped Panicking Before Releases

From outsourcing to product — a QA engineer’s honest journey to better releases, healthier work culture & real impact on the product.
👍1
Find and Remove Invisible Characters from AI Text

AI-generated text isn't as clean as it looks. Copying from ChatGPT, Claude, or Gemini can introduce hidden Unicode characters — digital watermarks that break code and formatting, causing headaches for writers and developers.

A free extension, “Invisible AI Chart Detector” for Chrome and VS Code, solves this. It finds and removes these invisible characters with a single click and normalizes "fancy" typography to standard ASCII, ensuring your text is clean and predictable.
Postgres Pro TDE — security and performance

Not all Transparent Data Encryption (TDE) for PostgreSQL is the same. A new article reviews TDE implementations from Percona, Cybertec/EDB, and Pangolin/Fujitsu, analyzing the trade-offs between performance, reliability, and flexibility.

The text details how Postgres Pro Enterprise implements key rotation without rewriting entire tables — a significant advantage for security and performance. It also covers why AES-GCM was the clear choice for the encryption algorithm.
Why LLMs Drift into Convincing Nonsense (And a Practical Solution)

Ever get an answer from an LLM that’s elegant, logical, and completely wrong? The model isn't "understanding" your complex idea; it's guiding you toward a solution that's computationally cheap for it.

This is an emergent consequence of probabilistic optimization, where the AI funnels users down predefined "corridors" to a simple, template-based answer. It's a system where mathematical efficiency serves commercial objectives. The article explores why this happens and offers a way to combat it.
Autism spectrum disorders and a career in IT: personal experience

"I'm stunned by the illogicality of others, and they are stunned by the fact that I'm a robot"

In a new article, an author shares their personal experience of navigating the world and an IT career with an autism spectrum disorder. For a long time, they felt like an alien, unable to understand the seemingly irrational actions of others while being seen as strange themselves. This is a personal story about the journey from feeling like a "withdrawn, gloomy dude" to understanding a unique perspective.
👍2
AI Slop Coding: The Rise of Ridiculously Long Attack Chains

Threat intelligence research has uncovered a series of unusual, targeted attacks using GitHub repositories to store both malicious files and victim data.

The campaigns, active since at least September 2024, don't rely on complex malware. Instead, they use ridiculously long chains of very simple scripts that appear to be heavily AI-generated.

This approach combines the strategic thinking of an APT-style attack with the low-effort execution of a "script kiddie," creating a bizarre but effective new threat.

Let's dive into the anatomy of these campaigns and see how AI enables this new breed of attack.
Beyond ora2pg: A Better Way to Migrate from Oracle to Postgres

Migrating schemas and data from Oracle DB to Postgres Pro Enterprise is a notoriously rocky road, full of countless pitfalls for administrators and developers.

Interestingly, Postgres Pro Enterprise includes extended features specifically designed to mimic Oracle's mechanisms and smooth out this very process.

However, the widely-used open-source utility ora2pg doesn't support them. This gap led to the creation of ora2pgpro, a tool built to leverage these unique features for a more seamless transition.

Let's explore the specific migration challenges and see what makes this new tool a better fit for the task.
A Guide to the UX Researcher's Role

UX research is the essential foundation of great user experience design. It moves beyond assumptions to systematically study a product's target audience and understand what they truly need.

The process involves collecting and analyzing data about user behavior, their expectations, and how they interact with a digital product to find paths for improvement.

This guide covers the key skills and daily responsibilities of a UX researcher, offering a clear path for recent graduates and those looking to break into the field.

Let's explore the skills that define a great UX researcher and the first steps to start a career in this critical discipline.
PostgreSQL 18: Part 5 or CommitFest 2025-03

PostgreSQL 18 has been officially out for a month, but the feature autopsy is just now concluding. This is the grand finale of our review series, and it tackles the final boss: the March 2025 CommitFest. This last development sprint is traditionally the biggest, the one where all the juiciest features get crammed in right before the gates close.

This review is a big one, precisely because the patch itself was massive. This isn't just a minor update digest; it's a deep look at the richest part of the new release. Let's dig into the most powerful new toys and capabilities that made the final cut.
Shardman: а quick guide for the architect

The myth of the magical fast=true parameter is still alive and well. In the world of distributed databases, it has a new contender: distributed=true. Spoiler: neither will save you. This isn't a game you win by flipping a switch; it’s a game that demands you rethink your entire schema, sharding keys, sequences, and queries.

This is a clear-eyed guide to the real trade-offs. We’ll walk through every corner — from colocated tables and CDC to topologies and foreign key constraints — to find out where performance actually improves, where it gets much more expensive, and how to deal with the fallout.
👍1
OAuth 2.0 authorization in PostgreSQL using Keycloak as an example

Support for the OAuth 2.0 Device Authorization Flow has landed in Tantor Postgres 17.5.0 (and is heading for PostgreSQL 18). This means you can finally play the "log in via Keycloak" game directly with your database, offering a modern and secure access method that's perfect for cloud environments and microservice architectures.

This guide walks through the entire setup, showing how to get this new feature talking to Keycloak. We'll follow the full path — configuring the identity provider, preparing PostgreSQL, writing an OAuth token validator, and verifying the whole thing works from psql using the Device Flow.
How we boosted SQL query accuracy by 33% with LLMs

An LLM-based SQL generator seems like an obvious win. Just hook up a powerful model's API, grant it database access, and... fire your human analyst? This isn't some simple "replace the human" game. In reality, no company in its right mind will pipe sensitive data into an external API from OpenAI or Anthropic.

So, self-host? Good luck. Open-source models often choke on complex schemas or specific dialects like PostgreSQL 17, and training them is a costly nightmare. This whole "just use an LLM" idea is a non-trivial challenge. We dove into this mess and found a way to boost accuracy by 33%. Let's explore how to actually tackle this problem.
Privacy on Mobile: a practitioner’s checklist

Privacy has always been a high-stakes game, but the AI wave and our data-hungry economy have turned our phones into the main playing field. Every digital crumb is an asset. While some users are savvy, relying solely on "user awareness" is a losing strategy. The first line of defense isn't the user; it's the developer.

This isn't just another compliance lecture. It's a practitioner's mental model for how to frame decisions around privacy from the ground up. Let's dive into the concrete checklists and practical examples that help build that defense.