Kia ora! Welcome to New Zealand’s weekly roundup of AI news, events, jobs and education.
“AI will solve the world’s biggest problems”. That’s the hype.
The reality is teenagers are minting cash making AI deepfakes of world leaders arm-wrestling on yachts 😂.
But while memes go viral, the serious money is moving.
On September 4th, Trump turned a White House dinner into Silicon Valley’s priciest handshake.
Mark Zuckerberg, Tim Cook, Sundar Pichai, Satya Nadella, Bill Gates, and Sam Altman lined the table.
The offer: fast-track permits and power for AI data centres.
The response was eye watering: Meta and Apple: $600B each. Google: $250B. Microsoft: $75–80B annually.
All earmarked for AI infrastructure chips, data centres, grid capacity—under the White House’s new AI Action Plan.
Behind the photo ops sits the real play: America isn’t just betting on AI apps. It’s buying the rails, and it wants to lock in global dominance.
This week’s highlights
New section: 📚️ AI for Business Education
🤖 Bella: $100M Farm Research Goes AI
📉 Redesign Work: The Curiosity Divide
🛠️ Custom GPTs: Still the Secret Weapon
📊 NZ AI: Real Impact or Dress-Up?
🦺 Safety Tech: Tane van der Boon’s Inviol
🌍 Global AI Moves: Big Tech’s Next Play
Happy reading and listening ✌️
Did someone forward you this? Sign up!
HEADLINE STORY
📊 Kiwi farmers get an AI assistant Bella
Logo of the Bella app.
📣 Word On The Street: Beef + Lamb NZ launches AI assistant putting $100M research at farmers' fingertips via smartphone.
🔍 Zooming In:
Bella processes 4,600 pages plus 220 hours of verified NZ farming content instantly.
Over 100 farmers tested the AI across 15 specialist areas during two-month trials.
Multi-language support includes te reo Māori, works via voice commands anywhere
🏘️ Our Take: Beef + Lamb New Zealand just accidentally proved why agricultural expertise could be our country's biggest export opportunity.
While everyone celebrates Bella the AI assistant, the real breakthrough is the verified farming knowledge now sitting in machine-readable format. That's $100 million worth of research transformed into New Zealand's most comprehensive agricultural intelligence platform. Every agtech startup, university researcher, and government department will want access to this verified, NZ-specific farming knowledge. Beef + Lamb NZ have built not only a helpful chatbot, but also the foundation for every future agricultural AI system in the country.
The path to commercialising this goldmine runs through farmer adoption at scale, not just technical brilliance. Agricultural consultants currently charge $150-300 per hour for expertise that Bella delivers instantly. Farm management plans costing thousands can now be generated through voice commands. But the real value emerges when thousands of farmers start using the system daily, creating usage patterns and insights that improve the AI model. Each query becomes training data. Each successful recommendation strengthens the algorithm. The winner won't be whoever builds the best AI - it'll be whoever controls the feedback loop between farmers and the system. That's how we transform from commodity exporters to knowledge exporters, selling our agricultural intelligence to farming communities worldwide.
🐝 Save hours of work starting today.
Try Autohive free and build AI agents in minutes that tackle your most repetitive tasks - secure, private, and ready when you are.
💼 Business & Industry
Wellington mum tackles the invisible mental load crisis with AI app. PAM (Personal Admin Manager) hit #2 in Apple's lifestyle rankings after soft launching in February. The Wellington startup's AI reads everything from school emails to birthday invites, automatically creating tasks and reminders for family members. Over 6,500 users have signed up - 75% through word-of-mouth alone. More than a quarter report 70-90% stress reduction in family admin.
3-min read.
Our take: The organic growth metrics here are striking, when stressed parents become your marketing team, you've found product-market fit. This could be the productivity category nobody knew they needed.
Amazon's recycled announcement still delivers genuine regional transformation for AI. Tuesday's announcement repackaged existing data centres as New Zealand's first "infrastructure region", but the substance still delivers. Local businesses can now train AI models and run machine learning workloads without offshore latency that previously slowed real-time applications by hundreds of milliseconds.
6-min read.
Our take: Beyond the headlines, this accelerates New Zealand’s AI ecosystem by shrinking iteration cycles so local developers can prototype and launch faster, lowering the entry barrier for startups that no longer need to lean on costly offshore infrastructure, and turning the country into a more compelling base for global talent and firms who see real capability onshore.
New Zealand's SMEs are falling behind in the global AI race. Icehouse CEO confirms what many suspected after 100+ business leaders highlighted strategic AI adoption gaps. The new A.icehouse programme in partnership with Mark Laurence from Ten Past Tomorrow joins forces to deliver 24-month AI roadmaps for Kiwi SMEs. The programme blends strategic leadership guidance with hands-on operational training.
3-min read.
Our take: SMEs don’t need to outspend corporates, they just need to out-execute them. With $20-a-month tools and the right roadmap, small businesses can move faster than big incumbents weighed down by bureaucracy, but this starts with Leader buy-in to ensure the right top-down culture is in-place.
Queenstown Airport just deployed AI-powered LiDAR across five departure zones to slash passenger wait times. The system tracks queue lengths and predicts congestion without collecting personal data, helping staff redirect passengers before bottlenecks form. Real-time insights let operators open new counters and adjust schedules instantly during peak travel periods.
4-min read.
Our take: With LiDAR mapping crowd movements, staff can be deployed where they’re needed most instead of reacting after bottlenecks form. Predictive modelling means the airport can scale seamlessly during peak ski or holiday seasons. Efficiency like this protects margins while lifting service standards.
💼 Government & Legal
Twenty AI experts are urging the government to regulate AI, warning that New Zealand’s “she’ll be right” stance risks deepfakes, fraud and threats to democracy. Dr Andrew Lentzen of Victoria University proposes a three-tier approach: ban high-risk uses like state surveillance, regulate medium-risk tools such as automated loans, and leave low-risk apps alone. Businesses need AI for productivity but unclear rules stall investment. The government promises $76b in GDP gains, though Lentzen questions the Microsoft-sourced figure.
5-min listen.
Our take: The real danger is less about heavy-handed rules, it’s stalled productivity. Clear guardrails give businesses the confidence to adopt AI at speed, while building the public trust needed for widespread use. If New Zealand wants to climb from 40th in readiness, regulation is not just compliance, it’s a chance to compete globally.
Hutt City Council just won innovation gold for saving $900,000 annually with AI assistants. The Lower Hutt council deployed 15 custom AI tools and 300 licences across operations, reclaiming 44,000 staff hours per year. Their AI-Volution initiative beat sector-wide competition at Te Hapai Hapori Awards, marking rare local government recognition.
3-min read.
Our take: Local government is often seen as outdated and disconnected from residents’ daily lives. By embracing generative AI, Hutt positions itself as a council that speaks the same digital language as its citizens, showing leadership that builds trust. When council staff spend less time on admin and more time on core services, residents feel it through faster permits, quicker responses, and services that actually keep pace with demand.
🎙️ The AI Corner podcast
This week’s guest is Tane Van der Boon, the founder of Inviol.
Hear how AI and computer vision are being deployed to tackle complacency and prevent injuries in New Zealand’s most dangerous workplaces. Discover how Inviol builds AI that detects unsafe behaviour in real time, why edge computing matters for latency and reliability, and how startups can partner with legacy industries to scale.
📚️ AI for Business Education
Helping leaders and teams adapt, learn, and scale with AI.
Rethinking How We Work
1️⃣ Branching Conversations in ChatGPT
A simple new feature with big implications for structured thinking. Leaders can now branch threads to compare outputs, run A/B tests, and keep strategy discussions clean.
80-sec watch.
2️⃣ The Curiosity Divide: Redesign Work for the AI Era
Stanford data shows entry-level jobs in AI-exposed industries are down 13%. Compliance gets automated first. Firms must redesign junior roles around curiosity and AI fluency, or risk losing their future leaders.
6-min read.
Building with AI
3️⃣ Mastering Custom GPTs
Most businesses underestimate Custom GPTs. Treated properly (with clear roles, sharp prompts, and Actions) they become reliable assistants and scalable assets.
7-min read.
4️⃣ AI in the Trenches
Matteo Castiello explains why AI success isn’t flashy demos, but repeatable workflows that people actually use.
2-min read.
5️⃣ Tools of the Trade
From research to content to automation, here’s a practical look at the AI stack driving productivity today. A useful benchmark for leaders building internal toolkits.
2-min read.
Lessons from NZ
6️⃣ Are We Just Playing Dress-Up?
NZ’s latest AI Productivity Report reveals adoption gaps. The question for businesses: are you applying AI with impact, or just experimenting at the edges?
6-min read.
🌍 Tech Updates From Global
OpenAI: Announced Jobs Platform and Certification Program to connect AI-skilled workers with employers, aiming to certify 10M Americans by 2030; enhanced GPT-Realtime voice agents with reduced latency and emotional nuance, piloted by T-Mobile; launched Real-Time Voice API and a new speech-to-speech model for natural production-grade agents; rolling out parental controls including age rules, memory off, and distress alerts, plus new teen supervision features; abandoned AI detector software due to inaccuracy; rolls out feature to branch conversations in ChatGPT; introduced ChatGPT Projects and customisation tools to free tier users; launched OpenAI for Science using GPT-5 to accelerate discovery; developing first proprietary AI chip to reduce reliance on NVIDIA.
Google: Gemini “Nano Banana” model topped recent rankings for image editing performance; introduced AI-powered Google Finance beta for real-time market Q&A; upgraded NotebookLM with new audio modes (Brief, Critique, Debate) and “choose-your-own-tone” presets; launched updates to Google Vids for non-creators and Photos’ animated clip features; Gemini AI integrated into TCL QM9K TVs for natural home control; partnered with Apple to enhance Siri with Gemini-based web summaries by 2026 while continuing AI search integration tests; committed $150M in AI education grants within a $1B initiative.
Anthropic: Raised $13B Series F at $183B valuation, tripling revenue run-rate from $1B to $5B in under a year; expanding capacity and safety research with funds; halting sales of AI services to Chinese-owned groups amid geopolitical restrictions.
Microsoft: Signed major agreement with U.S. General Services Administration offering Microsoft 365 Copilot free for up to 12 months to millions of G5 users, projecting $3B savings; launched VibeVoice for multi-voice generation and long podcasts as open-source; accelerating AI integration across federal agencies.
Meta: Facing ethical scrutiny after unauthorised celebrity-impersonating bots generated explicit content; internal turmoil deepens with AI team departures despite heavy investments; former Apple AI lead Ruoming Pang joined Meta with a $200M package; Scale AI (backed by Meta) sued rival Mercor and ex-employee for alleged trade secret theft.
xAI: Suing former engineer Xuechen Li for allegedly stealing Grok trade secrets before joining OpenAI.
Amazon: Launched Lens Live with “buy” button for all visible items; introduced Alexa+ with generative AI to handle multi-step autonomous requests.
Apple: Quietly deployed Asa AI chatbot for retail staff ahead of iPhone 17; testing Siri enhancements using Gemini AI with rollout by 2026.
Perplexity: Launched Comet Plus subscription sharing 80% of revenue with publishers; released AI-first browser Comet for students.
Mistral: Nearing €2B raise at ~$14B valuation, becoming one of Europe’s most valuable AI startups; cited in Stanford study linking AI automation to 13% decline in entry-level IT jobs over three years.
📅 AI Events in New Zealand
Spring has sprung, and so have AI events! With 24 AI events happening across Aotearoa this week, this is the most we have seen in a hot minute.
This week’s featured event:
M2 AI Summit, Thursday - Christchurch: Understand how to use AI & Technology to build better products & increase efficiency.
📅 Promote your event with us. Reply to let us know.
💼 AI Roles Around Aotearoa
Picklist of 🌶️ HOT 🌶️ new roles in AI this week.
Mechatronics Engineer, Rocket Lab: Auckland
💼 Promote your job with us. Reply to let us know.
🤦 ️ AI Fail Of The Week
We all love AI, but it’s certainly far from perfect 🤔 …
A dream and a nightmare all in one for Erin
👋 Mike & Erin