AI
Practical explainers on AI tools, models, and how teams actually use them — written for developers and business owners.
Google's MedGemma: Open Medical AI You Can Run Yourself
Google's MedGemma lets hospitals and developers run medical AI on their own servers, so patient data stays in-house. Here's what it can do by Google's own numbers, what its licence forbids, what hardware it needs, and ho...
Jev AI by TypeSafe Explained: How It Works, Real Use Cases, Cost and Benchmarks (2026)
Jev AI from TypeSafe is not a chatbot. It is a decision model that returns typed answers with probabilities for $0.042 per million input tokens. How it works, use cases, cost and sourced benchmarks.
Grok 4.7: What Actually Changed
xAI's Grok 4.7 keeps Grok 4.6's $2 / $6 price but brings a bigger base model, a 500K context window, and big jumps in agent coding. Here is what actually changed, and who should switch.
Claude Opus 5.5: What Actually Changed
Anthropic’s Claude Opus 5.5 claims Fable-level work at ~40% less cost than Opus 5. Here is what actually changed for coding agents, price, safety routing, Copilot, and whether you should switch.
GPT-6 Astra vs Claude Fable 5.1 vs Gemini 3.8 Flash: Which AI Model Should You Use in 2026?
A simple, detailed comparison of GPT-6 Astra, Claude Fable 5.1 and Gemini 3.8 Flash — price, coding agents, computer use, documents, speed and safety — with a clear pick for your work.
GPT-6 Astra: What Actually Changed
OpenAI's GPT-6 Astra is not just a bigger chat model. Here is what actually changed for computer use, coding, cyber risk, pricing, and whether your team should switch.
Best AI News of August 2026
A short briefing on the AI stories that actually moved in August 2026 — the chip pileup, the trillion-dollar cheques, the jailbreaks, and what they mean if you build or buy software.
Stop Pasting Secrets Into AI
A new 2026 jailbreak can trick AI assistants into leaking chats and API keys. What not to paste, and a safer browser-side workflow.