The Business of Tech

Forget soy, pea, or lab-grown meat – the next frontier in sustainable food might just be hiding in plain sight. Specifically, in the leaves of everyday plants growing across New Zealand’s farmland. In the latest episode of The Business of Tech podcast, I talk to Ross Milne, the CEO of Rolleston-based startup Leaft Foods, which has pioneered a breakthrough technique to extract and process Rubisco, a naturally occurring protein found in every green leaf.  Scientists have long known Rubisco’s potential, calling it the “utopia protein” for its rich nutritional profile and low environmental footprint. What’s been missing until now is a practical way to isolate it for use in food and animal feed. From milk to leafy greens Milne, a former process engineer who worked for some of the world’s largest food companies, saw an opening for innovation back home in New Zealand, teaming up with Leaft Food founders John Penno (Synlait Milk co-founder) and Maury Leyland Penno. Leaft’s approach promises not just a powerful alternative to traditional protein sources, but a clever circular system where farmers can use the high-protein byproducts as feed supplements, boosting productivity while cutting emissions. The engineering ingenuity allowing Leaft to extract sufficient quantities of Rubisco is what caught the eye of global investors in 2020, when Leaft raised US$15 million in a Series A round. Among Leaft’s backers is Khosla Ventures, the legendary Silicon Valley venture capital firm known for betting early on world-changing green technologies.  Low-impact protein Working mainly with alfalfa crops in the Canterbury region, Leaft harvests and processes the leaves, extracting the protein which is sold to food suppliers and which features in Leaft Blade, the company’s line of nutritional products. The leftover leaves are used by farmers for supplementary feed. “The interesting thing about [Alfalfa] from a grower point of view is it regrows straight away,” Milne told me.  “So about six weeks later, for example, we're back in that same paddock harvesting it again. And we just constantly do that. It's a perennial which stays in the ground for multiple years.” As the world races to find scalable, low-impact protein sources, Leaft’s innovation could place Canterbury at the center of the solution. For Milne, it’s a mission to transform the food system from scratch. Listen to the entire conversation on The Business of Tech podcast to find out how this Kiwi startup is redefining what we eat, how we grow it, and why the leaves in your backyard might hold the key to feeding the future. Streaming on iHeartRadio, Apple, Spotify or wherever you get your podcasts. Show notes Developing a new plant-based protein - Science Learning Hub An IPO could be on the cards for Leaft Foods one day - BusinessDesk NZ green protein producer sprouts new deal with Asian food giant - RNZ Leaft protein boosted from ground up - Farmers WeeklySee omnystudio.com/listener for privacy information.

New Zealand drivers are about to discover a whole new way of paying to use the roads – and for most, it will be a shock.  For decades, petrol and diesel motorists have funded the transport network through fuel excise quietly folded into every litre at the pump - currently a 70c tax.  Soon, that largely invisible tax will give way to something much more visible – paying per kilometre under an expanded road user charge (RUC) system. On the latest episode of The Business of Tech, I talk to Dunedin-based entrepreneur Adam Johnston about what may be the biggest shake-up to transport funding in 50 years. Light vehicle owners who have never had to think about RUC before will be pulled into a regime that currently applies to heavy vehicles, diesel cars and electric vehicles. Petrol vehicles currently make up around 55% of the national fleet. Instead of passively paying when you fill up, you’ll be actively buying distance in advance, tracking your odometer, and keeping on the good side of Waka Kotahi. A new marketplace for RUC payments   That sounds like a recipe for confusion and admin overload, especially in a cost-of-living crisis where drivers are already stressed about the price of petrol and diesel. But this shift is also opening the door to a wave of innovation. As the government hands more of the RUC system over to private providers, a new marketplace will emerge around how you pay to drive. It will likely be in the form of apps that let you buy and manage your RUC from your phone, real-time dashboards that show how much you’ve used, and even telematics devices that automate the whole process by reporting your mileage in the background. Payment platforms will sit in the middle, clipping the ticket on every transaction. Start-ups and incumbents alike will compete to become your go-to RUC retailer, bundling services and perks to win your attention and loyalty. Johnston and his co-founder, Briyarne Pascoe, both former Delivery Easy workers, are among the entrepreneurs racing to shape this new ecosystem. Building on their RUC Hub project, a free-to-access platform that tells you everything you need to know about road user charges, they saw an opportunity to make a complex system more transparent and user-friendly, while preventing the market from devolving into a cosy oligopoly. The pair plan to become a retail player in the emerging RUC ecosystem. In the episode, Johnston explains the trade-offs between better digital experiences and the extra transaction costs that could quietly inflate what you pay overall. This episode unpacks what’s coming, how your relationship with your car and your wallet is about to change, and the tools that could make surviving the new RUC era a little less painful. Streaming on iHeartRadio, Spotify, Apple or wherever you get your podcasts.See omnystudio.com/listener for privacy information.

This week on The Business of Tech, I talk to Inspired Education founder Nadim Nsouli to explore a bold experiment in AI‑driven schooling that will reach Auckland primary students from 2027. Inspired Education operates seven ACG private primary schools in New Zealand, including five in Auckland, focusing on personalised learning. A new learning programme, Inspired Edge Academy, compresses the traditional core curriculum – English, maths, science, computing and languages – into three highly structured, interactive hours each morning.  Afternoons are turned into a lab for real‑world skills: financial literacy, entrepreneurship, public speaking and problem‑solving. Underpinning it all is an adaptive AI learning system that changes questions and pathways in real time, depending on where each child is struggling or racing ahead, making progression based on mastery rather than age. AI can personalise learning Nsouli told me that Inspired has already invested tens of millions of dollars in technology across its 125 schools, using platforms like Century Tech to personalise homework and classwork for 95,000 students. In some subjects, students using these adaptive tools have lifted assessment scores by the equivalent of a GCSE grade boundary in just six weeks. Nsouli walks through what this looks like for an eight‑year‑old: short 20–25 minute learning blocks, small clusters of students regrouped by mastery for each subject, a teacher‑to‑student ratio of about 1:8, and AI dashboards that show educators exactly where to intervene. The philosophy behind the empire Nsouli also tells Inspired's origin story. He left a successful private equity career after a personal tragedy, the death of his daughter Lyla, who died in 2012 at age 3 from a rare, aggressive brain cancer. It was a turning point in Nsouli’s life, inspiring him to build a global group of premium schools that now employ 15,000 staff and educate 95,000 students on six continents. He sees the use of AI as “digitally native but human‑centred”. Smartphones are banned in all Inspired schools globally, teachers remain central, and technology is used where it can clearly outperform paper – in adaptive practice, feedback and assessment. What it means for New Zealand From 2027, the Edge model will appear in Auckland’s Inspired schools, after an early access launch in London, with the potential to spread faster where parent demand is strongest.  We also discuss whether AI‑powered mastery learning will widen the gap between private and state schools or eventually filter through to the public system as costs fall and evidence grows. Listen to the conversation with Nadim Nsouli, streaming on iHeartRadio, Spotify, Apple, or wherever you get your podcasts.See omnystudio.com/listener for privacy information.

Why do some places become tech powerhouses while others, just as smart and connected, stall out?  In the latest episode of The Business of Tech, global innovation expert Mehran Gul, a former policy expert at the World Economic Forum, and the United Nations, dispels the myths about where breakthrough technology happens – and gives his take on what New Zealand should do to improve its innovation game. Gul’s new book, The New Geography of Innovation: The Global Contest for Breakthrough Technologies, comes to a simple conclusion - there’s no fixed map of innovation.  Silicon Valley, with its concentration of startups, tech talent and venture capital, remains an innovation powerhouse. But China went from being dismissed as a copycat to a genuine tech superpower in five to seven years.  Canada, via one small lab at the University of Toronto, helped trigger the entire modern AI boom with AlexNet and a handful of researchers who went on to shape OpenAI and the wider industry. If they can bend the curve that fast, so can so-called “middle powers” and smaller countries like New Zealand – if they stop telling themselves they’re too small, too distant, or too late. Tech makers, not tech takers Gul explains how the ingredients of innovation look radically different from place to place. Silicon Valley has anchor universities, thick venture capital markets and a constant deal engine of big firms buying smaller ones.  In China, the giant tech platforms came first, then research labs and world‑leading developments like ResNet, a neural network architecture. Singapore doesn’t have household‑name tech brands, yet it dominates venture capital in Southeast Asia. Gul’s assessment of New Zealand is that the likes of Rocket Lab have demonstrated our ability to produce successful, innovative companies. It’s that we keep losing them, talent and listings included, to the United States and other major markets. He argues the real test is whether a country can hang on to its winners, build a startup “factory” around breakout successes, and use policy to push founders towards IPOs and broad employee ownership instead of early exits to US tech giants.See omnystudio.com/listener for privacy information.

When global giants argue about which AI models are safest, smartest and ready for prime time, a New Zealander in Redmond, Washington is one of the people making the call. Steve Sweetman leads the team at Microsoft's global headquarters deciding which AI models make it onto Microsoft Foundry – the platform that now offers more than 10,000 models from OpenAI, Anthropic, Meta, DeepSeek, Microsoft’s own labs and others, to customers around the world.  Sweetman studied architecture at Unitec in Auckland and fell in love with technology as the industry shifted from manual drawing to computer-assisted design. Stints at Wang and Telecom led to Microsoft New Zealand, then on to Redmond, where he’s spent over two decades at the coalface of new tech: HoloLens, early chatbots, and the shift from narrow AI tools to today’s generative AI platforms. Setting the responsible AI rules He helped set up Microsoft’s Office of Responsible AI, turning high‑minded principles into practical policies and tools. That experience now shapes how Foundry works. Before any third‑party model is switched on for customers, Microsoft runs its own evaluations and has, Sweetman reveals, rejected popular models that didn’t meet its safety bar.  Sweetman explains why the real value is no longer “just the model” but the data, governance and agent frameworks wrapped around it. We talk through concrete use cases from Alaska Airlines using generative AI to personalise travel planning, to CVS Health applying models to cancer research, and tiny Australian startup Lyrebird building multilingual tools to close gaps between patients and clinicians. For New Zealand, where AI adoption is lagging and talent is thin on the ground, Sweetman is bullish. You don’t need to build your own foundational model, he argues. You can plug into powerful platforms like Foundry, understand the safety guardrails, and start experimenting. If you want an insider’s view of how Microsoft is curating the world’s AI models – and a Kiwi’s take on what that means for local businesses – this is an episode worth queuing up. Streaming on iHeartRadio or wherever you get your podcasts. Thanks to our sponsor 2degrees and to Microsoft for hosting me in Redmond.See omnystudio.com/listener for privacy information.

This week on The Business of Tech, I talk to Neara co‑founder Jack Curtis about how a “physics-based digital twin” of electricity grids is changing the way we plan, build and protect electricity infrastructure – from Taranaki to Texas.  Neara has just raised A$90 million in a Series D round led by US investment firm Technology Crossover Ventures (TCV), which also invested in Netflix, Spotify, Facebook and Xero. That takes total funding in Neara to about A$180 million, as some of the world’s most exposed utilities rush to digitise their networks in the face of extreme weather and the clean‑energy transition. Neara’s origin story isn’t very corporate. Software engineer and Neara co-founder, Daniel Danilatos, hacked together a better power line design tool over a weekend for his wife, a line designer frustrated with clunky legacy software. The prototype spread “organically” in an industry notorious for moving slowly. Within a few years, it had become the basis for a company now modelling around 90% of Australia’s electricity networks and working with most major utilities in Texas and California, and with a roster of New Zealand clients. Predicting when things break Most “digital twins” in utilities have been glorified 3D maps – pretty visualisations that don’t give asset owners enough confidence to make high‑stakes decisions. Neara instead builds behavioural models where every pole, line and substation is infused with real‑world physics: how it bends in a storm, heats up as load rises, or fails when gusts hit a certain speed.  As Jack puts it, if you look at the pole outside your house in a gale, it should behave exactly the same way in Neara’s model – right up to the moment it snaps. We also look at how physics‑based models help solve “good problems” like renewables congestion. Neara simulates how much extra power can safely be pushed through existing lines, where new wind or solar should connect, and how different mixes of generation and load will behave over 10–30 years.  That’s crucial for countries like New Zealand, which sprinted to 80–90% renewable electricity without fully modelling system‑wide side‑effects such as dry‑year risk and fossil‑fuel fallback. I found this chat fascinating and I’m sure you will too if you are interested in how evidence-based digital twins can transform industries.  Streaming on Apple, Spotify, iHeartRadio or wherever you get your podcasts. Thanks to our sponsor 2degrees.​See omnystudio.com/listener for privacy information.

A half-tonne metal “donut” silently floating in a vacuum chamber in Wellington might sound like science fiction. But as you’ll hear in the latest episode of The Business of Tech, it’s very real – and it could reshape New Zealand’s role in the global race for nuclear fusion. This week, I sit down with BusinessDesk journalist Greg Hurrell to unpack OpenStar’s dramatic new milestone: levitating a superconducting dipole in a near-perfect vacuum and firing superheated plasma around it.  It’s a key proof point for the Wellington startup’s radically different approach to fusion, one that flips the dominant tokamak design inside out. Instead of surrounding the plasma with giant magnets, OpenStar suspends a powerful magnet in the centre of the chamber and uses Earth-like magnetic fields to confine the plasma. ​Big ambitions, big interest Greg and I were in the room at Open Star last week as Prime Minister Christopher Luxon, Regional Development Minister Shane Jones, Infrastructure minister Chris Bishop, investors, scientists and even a representative from the United Arab Emirates watched the demonstration.  For a company that has only raised a modest $10 million Series A round, hitting this milestone matters. it shows OpenStar can deliver on ambitious engineering promises, exactly what global venture capital wants to see. The government has granted OpenStar a $35 million loan via the Rural Infrastructure Fund to build a bigger prototype – Tahi. Shane Jones, Chris Bishop, and Christopher Luxon listen to OpenStar CEO and co-founder Ratu Mataira explain the plasma firing experiment. Inside-out design We dig into why this “inside-out” design could be simpler to build and maintain than giant international projects, how New Zealand-grown intellectual property in high‑temperature superconductors and flux pumps gives OpenStar a potential edge, and what comes next with its larger Tahi and Maui machines aimed at real fusion and, eventually, commercial-scale power.  We also tackle the hard questions: tritium supply, neutron damage to reactor components, and whether a relatively small team in Wellington can compete with well-funded overseas rivals and decades of tokamak momentum. If you are interested in energy, climate, deeptech or New Zealand’s science system, this episode goes deep on a genuine moonshot as it crosses from lab experiment into serious industrial ambition.  Streaming on Spotify, Apple, iHeartRadio, or wherever you get your podcasts. Thanks to our sponsor 2degrees. Show notes OpenStar Plasma Showcase Event - Youtube OpenStar completes critical step on nuclear fusion path - BusinessDesk Openstar says $35m Government loan will help it stay in NZ - BusinessDesk New Zealand fusion startup claims major advance in New Zealand trial - Bloomberg Nuclear fusion seems hot right now — but how close is fusion power? - CBC Chinese nuclear fusion reactor pushes plasma past crucial limit: what happens next - NatureSee omnystudio.com/listener for privacy information.

OpenClaw is the moment AI stops feeling like a clever chatbot and starts behaving like something closer to a digital co-worker. In the latest episode of The Business of Tech, you’ll hear exactly why. Veteran software developer and AI entrepreneurMike Hall joins me to break down what OpenClaw actually is in plain language. It’s not another prompt-and-response assistant, but a particularly smart type of AI agent that can wake itself up on a schedule, scan your data and tools, and decide for itself whether there’s work to be done. If it needs to write code to perform a task for you, it will do that too. Mike explains how that simple “heartbeat” loop, asking “Should I do something?” every minute, is the key shift that turns AI from reactive to proactive, and why that’s such a big deal compared with the chatbots most people have used so far. Skills and the hive mind We dig into how OpenClaw goes far beyond the current crop of AI agents baked into office suites and CRM platforms. OpenClaw is designed to live on your own infrastructure, plug into email, files and SaaS tools, and then act autonomously rather than waiting to be told what to do.  OpenClaw has sparked a surge of interest from developers, an explosion of “skills” that any OpenClaw instance can download. The hive mind model central to OpenClaw is unlike anything we’ve seen in commercial agent products. That’s probably why OpenAI has snapped up OpenClaw founder Peter Steinberger and will put him to work developing the next generation of AI agents for the creator of ChatGPT. Sandboxes essential We also cover the risks OpenClaw raises. Running OpenClaw on your personal machine can expose your entire digital life, which is why sandboxes and strict permissioning are essential. What happens when you let agents install community-built skills that might contain malware?  Then there’s Moltbook, the social platform where OpenClaw-powered agents post and argue with each other, and what that experiment tells us about a near future flooded with AI personas. If you’ve heard the noise about OpenClaw and “agentic AI” but still aren’t clear on what’s genuinely new here – and why it matters for your business, your data and your job – this conversation will get you there.  Streaming on iHeartRadio or your favourite podcast platform. Thanks to our sponsor, 2degrees. Show notes Mike Hall, CEO Ab0t.com OpenClaw: The AI Assistant That Actually Does Things - Turing College OpenClaw, OpenAI and the future - Peter Steinberger Meta and Other Tech Companies Ban OpenClaw Over Cybersecurity Concerns - Wired OpenAI hires OpenClaw founder Peter Steinberger - FT What OpenAI’s OpenClaw hire says about the future of AI agents - Fortune Is OpenClaw Closed? - Hackster OpenClaw threats: assessing the risks, and how to handle shadow AI - KaperskySee omnystudio.com/listener for privacy information.

New Zealand loves to boast about its clean, green energy story. With around 80 to 90% of grid electricity coming from renewable sources like hydro, wind and geothermal, we look like one of the world’s quiet success cases on decarbonisation. But beneath that headline number lies a much more precarious reality. When lake levels fall and gas supplies tighten, our energy system starts to look very exposed.​ In the latest episode of The Business of Tech, I sit down with Melissa Reynolds‑Clarke and Daniel Gnoth from Ara Ake, the national centre for energy innovation, to explore how we can lean on innovation to navigate this emerging energy crunch. The conversation ranges from process heat in dairy factories and meat plants that still run on coal and gas, to the growing risk that international customers will turn away from products that are not backed by genuinely low‑emissions energy.​ Ara Ake sits in the “valley of death” for new technology – that tough space between promising lab results and commercial deployment. Daniel explains how the organisation supports everything from fusion “moonshots” and hydrogen‑electric aircraft trials, to more grounded projects like battery storage at Wellington’s CentrePort, rural microgrids, and ultra‑cheap hot water control that effectively turns our cylinders into a giant, flexible battery. Melissa, drawing on decades in the rural sector and on energy company boards, highlights the brutal realities facing farmers and manufacturers who need affordable, reliable energy today, even as they’re pushed to decarbonise for tomorrow’s markets.​ We dig into some of the most promising levers for fast impact – smarter use of flexibility on the grid, re‑using old oil and gas wells in Taranaki for deep geothermal heat, and new business models that make technologies like biodigesters and community batteries actually stack up in a country of small, dispersed farms and towns. We also talk frankly about the capital gap that still exists between startup and scale‑up, and why system‑wide thinking across regulation, networks, and markets, matters just as much as shiny new tech.​ If you want to understand what New Zealand’s energy transition really looks like on the ground, and where innovation can genuinely move the dial, this episode is for you. Streaming on iHeartRadio or wherever you get your podcasts.  Show notes Who builds NZ’s LNG terminal? The two names being floated - BusinessDesk New liquefied natural gas terminal: 'Vital' or 'bonkers'? - RNZ Why the new LNG terminal could raise, not lower, your power bill - Newsroom Ryan Bridge: The Taranaki LNG terminal is a good idea, depending on who you ask - NewstalkZB Second interim boss appointed at Ara Ake as work continues to find next CEO - The Post Energy research centre Ara Ake secures $70 million in funding to support innovation - Stuff Ara Ake Impact Report 2025 - Ara Ake  See omnystudio.com/listener for privacy information.

In the latest episode of The Business of Tech, we look at the rise and fall of MethaneSAT, the $30 million national space project that was supposed to cement New Zealand as a serious spacefaring nation.  Instead, it became a case study in governance failure, misaligned incentives and lost opportunity.  Launched in March 2024 and lost in June 2025 after persistent spacecraft glitches, MethaneSAT’s methane-sniffing science payload worked but the rest of the system carrying it in space failed. Working in space is risky, and satellites do fail. But as this week’s guest on The Business of Tech, University of Auckland physics professor Richard Easther points out, New Zealand’s involvement in the international MethaneSAT project raised questions from the start.  “What happened… is that we found this opportunity and then we found reasons to do the opportunity,” he told me.  “If someone had come to us in 2018 and said, here’s $30 million, I want you to develop things that will lead to startups, things that will provide the workforce… we could have come up with a plan and it would have been much, much better than MethaneSAT.” Picking winners: "A terrible job" Easther is careful not to scapegoat individual scientists or engineers. His critique is aimed squarely at how New Zealand chooses its science priorities and partners.  “We do a terrible job of choosing science priorities in New Zealand,” Easther said.  “And the people who pushed MethaneSAT were not scientists and do not have visible track records of testing proposals for excellence and competence.”  From governance issues to the gap between what officials were told privately and what the public heard, Easther argues MethaneSAT exposed deep problems in how we govern high‑risk, high‑cost science. But this isn’t just a post‑mortem of a failed satellite. Easther draws a direct line from MethaneSAT to today’s multi‑million‑dollar bets on AI and quantum, warning that without transparent, contestable processes – of the kind used in US “decadal reviews” – New Zealand risks repeating the same mistakes at even larger scale.  The Government yesterday announced another significant science investment, committing $35 million from the Regional Infrastructure Fund to help start-up OpenStar Technologies develop a new, specialised facility for its new fusion machine. Easther says major science investments shouldn’t come at the cost of long‑term, curiosity‑driven funding, pointing to world‑leading local strengths in high‑temperature superconductors and quantum devices that were quietly underwritten by the Marsden Fund decades ago. Tune in to The Business of Tech to hear Professor Richard Easther on what MethaneSAT got wrong, and what we should learn from it. Streaming on iHeartRadio or wherever you get your podcasts. Thanks to our sponsor 2degrees. Show notes An eye in the sky to detect methane emissions - RNZ Taxpayer-funded climate satellite MethaneSAT finally reveals what's behind delays - RNZ Taxpayer-funded satellite had 'deep-seated problems' from launch - RNZ MethaneSAT Report: Advancing space capability and climate science - MBIE Government pulls back from full membership of Square Kilometre Array - RNZNew Zealand pulls out of the Square Kilometre Array after benefits questioned - Physics Today  See omnystudio.com/listener for privacy information.

Two Kiwi engineers who helped build the future of self‑driving cars in Silicon Valley are now quietly laying the foundations for the next great tech wave: physical AI.  In the latest episode of The Business of Tech, I talk to Harry Mellsop, co‑founder of simulation startup Antioch, and Adrian Macneil, co‑founder and CEO of data platform Foxglove, for a fast‑paced tour of where robotics is really at as Elon Musk talks up his Optimus humanoid robots. Both founders cut their teeth at the pointy end of autonomy. Harry worked on Tesla’s Autopilot, watching first‑hand how much time and money is burned putting robots into the real world safely. Adrian led key parts of Cruise’s self‑driving infrastructure and developer tooling, helping build the internal platforms that let engineers understand what a robot “saw”, “thought” and did on the streets of San Francisco. Big dollars for physical AI startups  That experience has now crystallised into two companies sitting at the infrastructure layer of physical AI – and investors are paying attention. Foxglove has raised US$40 million US in Series B funding, led by Bessemer Venture Partners with Icehouse Ventures on the cap table, to build the data and observability backbone for robotics teams. Antioch has secured US$4.2 million US dollars in pre‑seed funding, with Icehouse Ventures again involved, to bring Tesla‑grade cloud simulation to any robotics startup that wants to test thousands of edge cases virtually before a robot ever leaves the lab. Integration testing for atoms We explore how these platforms turn messy real‑world sensor feeds into structured insights, shorten development cycles from weeks to hours, and dramatically reduce the risks of unleashing autonomous machines into warehouses, construction sites and farms. Harry explains why “integration testing for atoms” is the missing link in robotics, and how simulation can slash the cost of safety validation. Adrian unpacks the idea of a data flywheel for robots – logging everything, surfacing the rare but dangerous failures, and feeding that back into better models and better code. If you want to know where AI goes next, why humanoids are still relatively clunky despite the viral demo videos, and how New Zealand founders are quietly shaping the infrastructure every serious robotics company will rely on, tune into episode 133 of The Business of Tech, streaming on iHeartRadio or wherever you get your podcasts. Thanks to our sponsor, 2degrees. Show notes Kiwi Harry Mellsop raises $7.3m for his physical-world AI start-up Antioch - NZ Herald Physical infrastructure AI firm Foxglove, headed by Kiwi Adrian Macneil, raises US$40m - NZ Herald  The Missing Infrastructure Holding Robotics Back with Adrian Macneil - The Machine Minds Show Rise of the robots: the promise of physical AI - AFP Physical AI: robotics are poised to revolutionise business - FT Humanoid robots take over CES in Las Vegas as tech industry touts future of AI - CNBSee omnystudio.com/listener for privacy information.

New Zealand likes to see itself as an agile, innovative tech nation. But when it comes to artificial intelligence, the story is more sobering than triumphant.  A new survey of 4,000 business leaders from around the world by research group IDC has revealed that just 8% of companies in Australia and New Zealand can be classified as “frontier firms” when it comes to their uptake of AI. That compares to the global average of 22%. Around half of our firms are classed as “laggards” and risk falling behind. Are we shrewdly waiting on the sidelines for AI to really prove its worth. Or are we merely dabbling with the tech, ill-equipped to embed it in our businesses? AI can be more than “fancy Google” In the first episode of season 4 of the The Business of Tech, I sit down with Sarah Carney Microsoft’s national chief technology officer, to explore an uncomfortable question: are Kiwi companies quietly locking in a decade of underperformance by moving too slowly on artificial intelligence? Carney has a front‑row seat to how “frontier firms” around the world are using AI to rewire their businesses, not just write better emails. In this episode, Carney, a ten-year veteran of Microsoft, spells out why that matters for jobs, growth and competitiveness. She also challenges some of the myths holding local leaders back. Is AI really a threat to entry‑level roles, or could it create better ones? Is governance a brake on innovation, or actually the catalyst that lets people take bolder bets? And why is our national cynicism becoming a liability in a world where experimentation is the new survival skill? If you’re a founder, executive, policymaker or just trying to work out what AI really means for your job and your business in 2026, this is an episode you do not want to miss. Streaming on iHeartRadio or wherever you get your podcasts. Thanks to our sponsor 2degrees.See omnystudio.com/listener for privacy information.

“Never a dull moment” is how Wellington-based veteran consumer tech reviewer and commentator Pat Pilcher describes the year in tech after relentless product launches, an “utterly insane” Black Friday sales season and the “enshitification” of the internet, thanks in large part to AI. In our final episode of The Business of Tech for 2025, Pilcher joins the show to break down the biggest trends of 2025 and what’s coming in 2026, from AI agents and smart rings to humanoid robots and the debut of solid‑state batteries.​ Apple, AI and the year of the fold Pilcher starts with the elephant not in the room: Apple’s slow play on generative AI.  “Every tech player and their pet poodle had an AI offering except Apple,” he said. “This is just crazy. This is a company that sets the trends that everyone slavishly follows, and they missed the bus on the biggest AI trend probably of the decade.”  Yet he thinks there is method in the apparent madness, arguing that “stepping back… until they get a mature offering” may prove “quite sensible” in such a fast‑moving space.​ That patience, he predicts, will collide with hardware in 2026. Pilcher is convinced 2026 is going to be the year of the iPhone fold, following in the wake of foldables leader Samsung.     AI slop, deepfakes and the S-curve of tech adoption AI dominated 2025, working its way along the classic S‑curve of technology adoption. While an enthusiastic user of generative AI tools, Pilcher is blunt about the downsides, from “AI slop” filling Facebook, X and LinkedIn to academics “pulling their hair out” as students outsource learning to chatbot tools.​ With hyper‑realistic video models like Sora3and an election year looming, Pilcher says “the general public needs to be a lot more critical, a lot more sceptical – and they’re not”. Pilcher chooses Cory Doctorow’s famous term “enshittification” to sum up a key, regressive trend of 2025. “You subscribe to a service, it sounds fantastic and it’s only $5 a month. Three months later, it’s $25 a month, does less, requires more of your information and they can’t guarantee your privacy and by the way, your password’s been stolen,” he said. Pilcher sees this as evidence that the business model underpinning AI is dubious, with companies investing “billions and billions of dollars in massive data centres” in a period of “geopolitical instabilities and macroeconomic instabilities”. Silicon became “the new global currency” in 2025, from Nvidia’s dominance to Google’s Tensor processing units (TPUs) and China’s push to go beyond 40nm (nanometers) under US export bans.  Smart glasses, smart rings and genuinely smart homes If 2025 was AI’s year, Pilcher also thinks it was when home and wearable tech quietly levelled up. He rates Meta’s new Ray‑Ban smart glasses, which can describe what you’re looking at and translate signs on command. Future prototypes, he notes, combine wristbands that track “tendon movements” for hand‑gesture interfaces with augmented reality (AR) overlays that could do everything from lie detection in negotiations to live 3D navigation in unfamiliar cities.​ Smart rings are another sleeper hit, with Pilcher praising rings for being “unobtrusive” and “tiny” while monitoring health stats well enough to “tell you proactively when you’re coming down with a cold or a flu a week before you start noticing symptoms”. In his own testing, backed by a blood‑pressure cuff and digital thermometer, a smart ring delivered accurate results. On the home front, Pilcher says the long‑promised smart home is finally here, thanks to the Matter standard, which means new gadgets “will basically work regardless if you have an Alexa, Apple, Siri or… Google Home”.  EVs, robots and the 2026 futures Pilcher also covers the post‑rebate slump in EV sales, the rise of value‑packed Chinese brands like BYD, and the misinformation around EV fire risks, pointing out how a petrol vehicle, not a battery, was to blame in a widely shared bus fire incident.  Putting his futurist hat on, Pilcher talks about smart contact lenses with built‑in displays and gesture‑tracking bracelets that could make smartphones “look as quaint as a Model T Ford”, always‑on access to AR shopping lists and navigation, and the first serious wave of humanoid robots.  With cheaper AI silicon and compact models, he “wouldn’t be surprised if in late 2026… humanoid robots become the next must‑have consumer electronics category for the well‑heeled”.  He also expects to see the debut of solid-state batteries as an alternative to Lithium-ion batteries tha

Drones have gone from hobbyist toys to decisive tools of war and essential infrastructure for industry. Few people have had a better vantage point on that shift than FenixUAS founder Dr Andrew Shelley.  In the latest episode of The Business of Tech podcast, the economist and aviation specialist explains how a decade of incremental innovation has transformed uncrewed aircraft into platforms that can reshape modern warfare, agritech and even search and rescue.​ From DIY quadcopters to smart weapons New Zealand’s first drone rules arrived ten years ago, when the technology was still rudimentary and often home‑built.  “Pretty much every part of drone technology has improved,” Shelley said.  Better batteries and lighter and stronger materials have almost doubled flight time, while mass‑manufactured airframes have brought the price of drones down. and far more capable sensors and onboard software. Other advances, such as sensor technology and onboard software, have flowed into features many consumers now take for granted, such as obstacle avoidance, rock‑solid position hold and follow‑me modes, as well as increasingly autonomous flight profiles.​ The Ukraine war, now approaching four years in duration, has been characterised by the use of drones by both Ukrainian and Russian forces.  The changing face of warfare Shelley recalled watching footage of a small first‑person‑view drone in Ukraine flying straight past a Russian electronic warfare vehicle “festooned with antennas” and striking the armoured vehicle ahead of it. The drone was trailing a hair-thin fibre-optic cable, allowing it to avoid radio jamming systems. “To a certain extent, what we’re seeing in Ukraine is that the old is new again,” said Shelley, pointing out that the current generation of drones echo some of the cruise‑missile tactics from the early 1990s.​ Shelley traces a clear line from ISIS workshops that assembled drones from AliExpress parts, through Turkey’s TB2 Bayraktar successes and Russia’s use of DJI’s Aeroscope detection tools, to today’s battlefields where consumer‑grade quadcopters handle intelligence, surveillance, reconnaissance and precision strikes.  The West, he argues, has been complacent: “Turkey was leading the way with its Bayraktar TB2, Iran is clearly leading the way with its Shahed series drones and we are playing catch-up,” he said, pointing out that the US is now reverse‑engineering an Iranian drone rather than setting the pace.​ Artificial intelligence is only beginning to make its mark in commercial uses in New Zealand, but Shelley says the leading edge is already visible in applications like Christchurch‑based SPS Automation’s large agricultural drones. These systems can autonomously identify wilding pines and apply “a small amount of chemical herbicide” to individual plants, an approach he argues could transform conservation economics by reaching areas that are “almost impossible on foot” or too expensive to service with crewed aircraft.​ Agritech, data and the search and rescue gap If the military implications dominate headlines, Shelley sees at least as much untapped potential in agritech and emergency response. He cites spray drones that can drop slug bait on vulnerable crops in muddy conditions where tractors would churn up soil and helicopters are cost‑prohibitive, turning marginal blocks into productive land.  Pasture management is another frontier. Instead of consultants walking paddocks with pasture meters or towing instruments behind quad bikes, he expects drones to fly automated grids soon to map grass cover and optimise feed wedges across entire farms, backed by “clever software” to interpret the imagery.​ Search and rescue, he argues, is “one of the things we haven’t done well with”, despite New Zealand’s vast coastline, mountains and national parks. Shelley believes agencies need to change their mindset and accept that in bad weather or hazardous terrain, “we have to move into a mindset where we’re happy to lose the technology,” risking a $100,000 drone instead of a multi‑million‑dollar helicopter and its crew to find people in distress.​ Building a drone industry – and workforce FenixUAS sits at the centre of the fledgling drone ecosystem, training over a thousand civilian and government operators a year, including the New Zealand Defence Force, and certifying many of the country’s advanced drone operators. That gives Shelley what he calls a broader overview of what everyone’s doing with drones than perhaps anyone else in the country, from agritech to infrastructure inspection.  While firms like Tauranga-based Syos, and SPS Automation point to a growing UAV scene, he says the real bottleneck is software talent, with drone companies crying out for mechatronics and software engineers

In the creaky world of financial advising, where compliance paperwork devours hours and clunky software feels like a relic from the dial-up era, a New Zealand startup is deploying AI to free advisers from the drudgery.  Marloo, co-founded by Hardy Michel, who cut his teeth as head of operations at Wellington-based share trading platform Sharesies, isn't building robo-advisers to supplant humans. Instead, he is using artificial intelligence to free up advisors so they can focus on the trust-building conversations that truly matter to their clients. In the latest episode of The Business of Tech podcast, Michel shares how his London-based venture is already winning paying customers across four countries, proving New Zealand fintech can scale globally from day one. Relocating to London in 2022, Michel joined Estonian-founded investing platform Lightyear, helping it launch across 22 European countries amid regulatory mazes far more complex than New Zealand's.  "I felt like I'd really rounded out probably the missing piece of my knowledge and learning, which was kind of how do you build the machine at scale?" Michel told me. Freeing advisors from low-value admin That experience, combined with angel investing via Blackbird Ventures, convinced him to co-found Marloo with fellow Sharesies alum Shakeel Lala. Marloo’s mission? Financial advice is potentially transformational but inaccessible, the founders realised. Advisers spend 70% of their time on low-value admin, from anti-money laundering checks to 50-page suitability reports that gather dust.  Existing tools are clunky, with Michel describing the "Windows 95-esque" systems financial advisors had to choose from before Marloo arrived on the scene. Marloo offers a hyper-specialised AI note-taker for client meetings. Unlike transcription tools, it sifts through hours of chit-chat to extract the 5% that counts – goals, risk tolerance, fees – and structures it for compliance or client follow-ups. From there, the AI evolves into a full operating system, turning advisors into "reviewers, not doers", Michel said. Finish a meeting, and Marloo drafts an annual review letter in two minutes, 95% ready for a quick edit.  "You no longer have to take notes after the meeting, have a second person in the meeting taking notes for them, or rely on anything else other than our product," Michel explained.  The result? Advisors onboard more clients without burnout, firms cut outsourcing costs, and the human element, crucial for navigating life's emotional money milestones like retirement or inheritance, stays front and centre. Giving robo-advice a wide berth This augmentation ethos sets Marloo apart from robo-advice hype. "If robo-advice was kind of as good as it was cracked up to be, we'd all be using it right now. And the reality is we're not," Michel told me.  He predicts regulators will be reluctant to green-light fully AI-driven advice, given the trust factor. Instead, Marloo aims to overhaul unit economics: lower fees, drop minimum balances (now often $500,000+), and make quality guidance available to more than just the wealthy.  "The mission is [to] transform the underlying [profit and loss]  in the unit economics of what it means to deliver advice to a customer so that we can actually reverse that," he said Marloo recently raised NZ$4.6 million in pre-seed funding to accelerate development.  "We're going to raise a little bit of money to answer a true false question in 12 months... that we are confident we can spend the next 10 years working on this and it's going to be a massive business," he recounts of the Blackbird pitch. As AI bubbles inflate, Michel warns against shiny tech without substance.  "It's never been easier to build... [but] also... to deliver a really shitty product experience," he said.  Marloo, he added, prioritises delight – a consumer-grade user experience in a B2B world. For an industry pricing out everyday clients amid rising fees (up 6% in the UK last year), this could be the reset financial advice needs. Tune into episode 129 of The Business of Tech, powered by 2degrees, for the full conversation, where Michel dives deeper into Estonia's entrepreneurial edge, Sharesies' early battles, and why financial advice must stay human-powered. Available now on all major podcast platforms. See omnystudio.com/listener for privacy information.