Kinetics News

Does Generative AI Still Needs a Human Eye ?

Does Generative AI Still Needs a Human Eye ?

It’s no secret that generative AI tools have become powerful assistants in modern work. From drafting emails and reports to generating insights and summarising complex documents, the productivity gains are real.  But there’s a catch. A big one.  Generative AI needs a...

July’s AI Headlines: Progress, Problems & Public Concerns

July’s AI Headlines: Progress, Problems & Public Concerns

AI re-Investment causes job cuts at Microsoft:  Microsoft has announced plans to reduce its workforce by up to 9,000 employees, representing approximately 4% of its global staff, as part of the latest round of job reductions this year. While the company has not...

5 Quick IT Wins That Cut Costs in 90 Days

5 Quick IT Wins That Cut Costs in 90 Days

5 Quick IT Wins That Cut Costs in 90 Days A survival guide for NZ professional services, charities & manufacturers under pressure.Cheaper Biscuits & No Colour Printing? Yep, times are tough. The NZ economy is putting pressure on every dollar. The lunchroom...

Can Copilot tame your Outlook Inbox?

Can Copilot tame your Outlook Inbox?

Is your email inbox overflowing? You’re not alone. Managing a busy inbox can be time-consuming and draining. The average professional deals with around 100+ emails a day, and over half of those could be spam or trivial notifications2. Every new email interrupting your...

Meet our upgraded Remote Engineering Team

Meet our upgraded Remote Engineering Team

Listening to your feedback You've told us that we need to step up our initial call contacts.  We understand.  These days, responsive and efficient support is not just a value-add—it is a necessity.  As organisations grow, so too does the complexity of their...

Device Code Phishing: A Dangerous New Scam You Need to Know About

Device Code Phishing: A Dangerous New Scam You Need to Know About

Device code phishing is a sneaky new way hackers are stealing people's online accounts. In device code phishing, threat actors exploit the device code authentication flow to capture authentication tokens, which they then use to access target accounts, and further gain...

June AI Digest

June AI Digest

AI wants to stay 'on'. Hot on the heels of Anthropic's blackmail experiment, OpenAi had a similar problem.  Researchers have observed that their newest model, o3, does not consistently follow commands, particularly in relation to shutting down. In a series of...

Did AI really try to blackmail its operator?

Did AI really try to blackmail its operator?

AI attempts blackmail to avoid being turned off. During pre-release testing in May 2025, Anthropic (maker of claude.ai) conducted extensive safety evaluations using controlled scenarios to test the model's behaviour under extreme conditions.  The company embedded...