Latest publications
AI agents are changing office work: five steps to avoid being left behind
AI agents are already being integrated into corporate processes — and instead of fear, experts offer five concrete steps to help employees preserve their position and increase their value.
Thomson Reuters lays out four rules for AI agents that businesses can trust
Thomson Reuters explained how companies can build AI agents for real work: with clear metrics, expert review, a shared language across teams, and a foundation of trusted tools.
1Password launches Unified Access to protect logins, tokens, and secrets used by AI agents
1Password introduced Unified Access, a platform that helps companies find vulnerable credentials, grant access to AI agents on demand, and track exactly who used secrets.
John McNeill: AI will accelerate IT hiring through demand for infrastructure and architecture
Former Tesla president John McNeill believes AI will not shrink the tech labor market, but will increase demand for infrastructure engineers, network engineers, and architects of complex systems.
Gemini helped restore YouTube email notifications for comments with Python in an hour
After YouTube disabled email notifications for comments in late June 2025, the author used Gemini to build a Python script on the YouTube Data API v3 and restore alerts in an hour.
OpenAI added a file library to ChatGPT for storage, search, and reuse
ChatGPT now has Library: OpenAI automatically saves uploaded and created files so they can be searched, reattached, and not lost along with chats.
Cisco introduced DefenseClaw — an orchestration and blocking layer for securing AI agents
Cisco released DefenseClaw, an open control layer for AI agents that scans skills and code, monitors actions at runtime, and instantly blocks dangerous operations.
Microsoft Copilot, open source and cloud: five ways to adopt AI on a limited budget
Experts from Ricoh, Thomson Reuters, Booking.com and Toyota advise against building AI from scratch: it is cheaper to use services you have already bought, open source and flexible cloud services.
Why ChatGPT and Other Chatbots Are No Place for Secrets: 5 Risks and What to Do After a Leak
Chatbots like ChatGPT, Claude, and Gemini encourage openness, but those conversations can be stored, reviewed by people, and used outside your control.
NiCE Cognigy showed how AI agents and humans will manage customer service together
At Nexus 2026, NiCE Cognigy presented a platform where AI agents, operators, and AI assistants work in a single environment, turning the contact center into a CX orchestration layer.
Project Nomad turns a Debian PC into an offline knowledge hub with local AI, maps, and Wikipedia
Project NOMAD is an open system for Debian and Ubuntu that, once installed, provides offline access to local AI, Wikipedia, maps, educational courses, and reference materials without internet.
Google introduces TurboQuant: how the new compression lowers the cost of local AI
Google’s new algorithm compresses the KV cache of language models in real time, promising up to 6x memory savings and faster inference, but it does not end the costly AI infrastructure race.
Noi brings ChatGPT, Claude, and Gemini into one desktop app without extra switching
Noi combines popular AI services and local models in one desktop interface so users can work with ChatGPT, Claude, Gemini, and Ollama without constantly switching tabs and windows.
MIT report: AI will reshape text-based work gradually rather than eliminate professions all at once
MIT research shows that AI is getting better at text tasks, but its impact on the labor market will build step by step, giving companies and workers time to adapt.
ZDNet: BlueOptima showed that AI for code in production is noticeably weaker than promised
ZDNet breaks down BlueOptima's findings: the best AI models for code succeed in fewer than 23% of production tasks, and the main adoption risk lies in inflated expectations and weak expertise.
Stanford’s Erik Brynjolfsson: AI won’t eliminate programming jobs, it will expand the market
Stanford professor Erik Brynjolfsson believes AI will not devastate the labor market for developers: it will speed up work, create new roles, and bring many more people into software development.
ZDNet compared ChatGPT and Claude: OpenAI stronger on features, Anthropic on writing and logic
ZDNet tested ChatGPT and Claude across ten work and everyday scenarios and concluded that Claude writes and reasons better, while ChatGPT wins on voice, images, and integrations.
Google showed how Gmail with Gemini turns months of correspondence into one report in 10 minutes
A new AI search in Gmail showed practical value: with three prompts, Gemini compiled statuses for five vendors, found the right contacts, and drafted an email instead of requiring a manual review of threads.
Raspberry Pi 5 gets much more expensive amid the AI boom: 16 GB boards now cost as much as a laptop
Raspberry Pi said the latest board price hike was driven by a sevenfold rise in LPDDR4 prices: memory is being diverted to AI data centers, and higher-end Pi 5 versions are already losing their budget-device status.
AI assistants help develop applications by voice and question the role of traditional IDEs
A keyboard-free development experiment showed that AI assistants already let users move real projects forward with voice and mouse, although it is still too early for them to fully replace IDEs.
Google Photos launches AI Enhance on Android: auto-cropping, lighting and color — with some misses
Google has begun the global rollout of AI Enhance in Google Photos for Android: the one-tap tool adjusts lighting, color and cropping, but sometimes crops photos too aggressively.
ChatGPT can now search for movies and TV shows on Tubi: a free catalog of 300,000 titles
ChatGPT has received an integration with Tubi and can now recommend movies and TV shows from the streaming service’s free catalog, which includes more than 300,000 titles and episodes.
How to share Kindle books with friends: Amazon’s instructions and limitations
Amazon lets you temporarily lend Kindle books to another user for 14 days — but the feature has non-obvious limitations worth knowing in advance.
Google updated AI Mode in Search: results now open alongside and reduce tab chaos
Google changed AI Mode in Search: results users open are now shown next to the query window so they can read sources without extra switching and dozens of new tabs.
Linus Torvalds approved AI code rules for Linux: humans bear responsibility
Linux has officially allowed AI assistance in kernel development, but barred models from signing patches and placed all legal and technical responsibility on humans.
Google released Gemini for Mac: the native app is better than the website because it can access window contents
Gemini’s native Mac app turned out to be more useful than the web version: it analyzes the contents of a window shared from the desktop and answers right away based on the current document, email, spreadsheet, or note.
Cisco, PwC and Wyndham showed how they are actually retraining employees to work with AI
Amid talk of layoffs due to AI, some companies are betting not on cuts but on retraining: Cisco is rolling out mandatory courses, PwC is redefining roles, and Wyndham is speeding up service.
Hostinger became the leader among AI website builders, and the market finally matured in 2026
ZDNet updated its test of AI website builders and for the first time found them suitable for real work: Hostinger took the lead, and GoDaddy learned to edit CSS.
Google added Skills to Chrome for saving AI prompts and launching them in one click
Google launched Skills in Chrome, a Gemini feature that saves effective AI prompts, runs them on the current page and selected tabs, and lets users edit them for their own tasks.
Google released a Windows desktop app with quick access to Gemini and Lens
Google has made the Google app for desktop available to all Windows users: via Alt + Space, they can launch Gemini, Lens, web search, and file search without a browser.