/// BLOG

LOG: BLOG

Ruminations on code, intelligence, and the architectural voids of the modern web.

2026-09-26

AI Contract Review - Risk Analysis, Version Comparison and a Pre-Signing Checklist

In the 2025 Vals benchmark, AI beat lawyers at data extraction and summaries but lost at marking up contracts (lawyers: 79.7%). How to build a contract review where AI finds the clauses and risks and a human decides - plus a clause playbook and protection against hidden instructions in documents.

14 min

Image SEO in the AI Era - Google Lens, Visual Search and Graphics Models Cite

2026-09-24

AI & SEO

Image SEO in the AI Era - Google Lens, Visual Search and Graphics Models Cite

Google Lens now processes close to 20 billion visual searches a month, and Circle to Search runs on over 580 million Android devices - images stopped being an afterthought to text and became a search surface of their own, with its own rules. Here's what alt text actually does (and doesn't do), how to wire images into ImageObject structured data, why an infographic built as HTML beats an infographic-as-image for citability, which file format to pick in 2026, and what the EU AI Act says about labeling AI-generated graphics.

14 min

AI Translation and Content Localization - Running Multilingual Content Without a Translation Agency

2026-09-24

AI & Automation

AI Translation and Content Localization - Running Multilingual Content Without a Translation Agency

The human evaluation from the WMT25 competition showed that Gemini 2.5 Pro, Claude 4 and the top commercial engines beat classic machine translation on most of the 30 evaluated language pairs - yet DeepL still wins most blind tests on European language pairs. The real question isn't "which engine translates better", it's "which engine for which job". I cover when to reach for an LLM versus DeepL, how not to lose brand terminology across languages, how localization differs from translation, what it costs per 1,000 words - and how I run this exact blog in Polish and English without a translation agency.

14 min

Bing, Microsoft Copilot and IndexNow - the Forgotten Ecosystem Feeding AI Answers

2026-09-22

AI & SEO

Bing, Microsoft Copilot and IndexNow - the Forgotten Ecosystem Feeding AI Answers

On July 1, 2026, Fabrice Canel, Bing's longtime head of crawling and indexing and the chief architect of the IndexNow protocol, retired from Microsoft. It's a good moment to notice how much of today's AI-generated answers actually run on the infrastructure he built. The whole SEO world is debating Google AI Overviews, while Microsoft Copilot entirely, and ChatGPT substantially, pull sources from the Bing index - and since February 2026, Bing itself shows you when you're being cited. Here's how the ecosystem works, how to configure it, and how much traffic it actually delivers.

13 min

Copilot, Gemini, or ChatGPT Business - Which AI Package Should Your Company Choose (2026 Comparison)

2026-09-21

AI & Automation

Copilot, Gemini, or ChatGPT Business - Which AI Package Should Your Company Choose (2026 Comparison)

Microsoft 365 Copilot is really $69-90 per seat a month once you add the required base license, not the $30 from the ad. ChatGPT Business costs $20 today - $5 less than earlier this year. Claude Enterprise dropped from a $40-200 range down to a flat $20 per seat. The prices alone are enough to get lost in - and that's just one of four variables that should actually decide which package goes to the whole company. I break down the three ecosystems into real cost, data protection, admin controls, and integrations - and show how to design a pilot before you sign a 300-seat contract.

13 min

Data Readiness - How to Prepare Your Company's Data for AI Before You Spend a Dollar on Implementation

2026-09-21

AI & Automation

Data Readiness - How to Prepare Your Company's Data for AI Before You Spend a Dollar on Implementation

Gartner projects that by the end of 2026, organizations will abandon 60% of AI projects specifically because they lacked AI-ready data - not because the model was bad. MIT NANDA's July 2025 report went further: with $30-40 billion invested by companies in generative AI, 95% of pilots delivered no measurable return. The common denominator in both cases isn't technological - it's the data mess nobody cleaned up before signing the vendor contract. Here's how to audit your data sources, assess their quality, and clean them up BEFORE the rollout, not while putting out fires afterward.

13 min

AI in Accounting Firms - From Invoices to Tax Filings: What to Automate in 2026

2026-09-21

AI & Automation

AI in Accounting Firms - From Invoices to Tax Filings: What to Automate in 2026

73% of accounting firms worldwide have already rolled out some form of AI automation, and among tax advisory firms adoption jumped from 9% in 2024 to 41% in 2025. At the same time, starting February 1, 2026, every business in Poland must be able to receive invoices through KSeF (the national e-invoicing system), and from April, issue them too. That's not a coincidence: mandatory e-invoicing and AI automation reinforce each other - just not the way most accounting firms assume. Here's the process map, what AI already automates well, what has to stay with a human signing the filing, and how to calculate ROI before you sign a vendor contract.

14 min

Deepfake Fraud - How Companies Lose Money to Fake CEOs, and How to Defend Against It

2026-09-20

AI & Security

Deepfake Fraud - How Companies Lose Money to Fake CEOs, and How to Defend Against It

Poland's NASK checked: 66% of Poles can't tell a cloned CEO voice from the real one. In 2025 the FBI broke out AI as a standalone fraud category for the first time ever - 22,364 complaints and $893 million in losses. The single biggest case: $25.6 million wired after one video call with a deepfake fake CFO. This isn't conference-talk science fiction anymore - it's a real risk for any company with a finance department and a phone. Here's the anatomy of the attack, the procedures that stop it, and why “spotting the fake” alone is a losing strategy.

14 min

TOTAL_RECORDS: 0110

PAGE 3/14

SYSTEM_STABLE