If you’ve been following the AI space lately, you’ll know it’s moving at a cracking pace. Just this month, xAI dropped Grok 4 Fast—a model that’s all about strong reasoning, but at a fraction of the cost. We’re talking 98% cheaper than its predecessor, Grok 4, for the same benchmark results. That’s not just a tweak, that’s a game-changer for anyone watching the bottom line.
What’s clever about Grok 4 Fast? It can handle both quick-fire answers and deep-dive reasoning, all in one system. So whether you want a snappy reply or a detailed breakdown, it’s got you covered—without burning through tokens (or your budget).
NZ Punches Above Its Weight in AI
Here’s something to be proud of: According to Anthropic’s latest research, New Zealand ranks fourth in the world for AI use, per capita. Israel’s leading the pack, then Singapore, Australia, and we’re right up there too. It’s a testament to how quickly Kiwi businesses are picking up these tools and running with them.
Lawsuits, Laws, and the AI Wild West
Of course, it’s not all smooth sailing. The owner of Rolling Stone, Billboard, and Variety has just sued Google, claiming their journalism is being used in AI summaries without permission. This is the first big US publisher to take on Google’s “AI Overviews”—those summaries you see at the top of search results.
Publishers reckon these overviews are stealing their traffic, which means less ad and subscription revenue. Watch this space—whatever happens could shape how AI and media play together for years to come.
Meanwhile, Anthropic has officially backed SB 53, a California bill that could set the first state-level rules for major AI companies like OpenAI, Google, Anthropic, and xAI. If it passes, big AI labs will need to:
- Publish public safety reports before launching powerful models
- Set up internal safety frameworks
- Protect whistleblowers who raise concerns about risks
The focus is on preventing “catastrophic risks”—think AI being used for biological weapons or billion-dollar cyberattacks. It doesn’t cover smaller stuff like deepfakes, but it’s a start.
China’s Brain-Like AI—A New Approach
Not to be outdone, Chinese researchers have announced SpikingBrain1.0, the world’s first “brain-like” large language model. Built by the Chinese Academy of Sciences, it’s designed to use less energy and run without Nvidia chips (which are in short supply these days). Instead of processing everything at once, it only activates the “neurons” it needs—just like our brains do. The upshot? Lower energy use, less reliance on expensive hardware, and potentially faster, cheaper AI for long documents.
What Does It All Mean for Kiwi Businesses?
So, what’s the takeaway for us here in NZ? I reckon it’s this: AI is getting faster, cheaper, and more accessible by the day. But with that comes new risks, new rules, and a need to keep our eyes open—especially as the legal and ethical landscape shifts.
If you’re thinking about how to use AI in your business, now’s the time to get curious, get informed, and maybe even get a little bit excited. The future’s arriving fast—and we’d love to help you make the most of it.