But it would be all downhill from there.

Despite some promising, YouTube-friendly test flights, the startup never managed any commercial sales.

Its founder exited in June 2015, saying, “The corporate world isn’t for me”, as he fell out with other investors over the company’s direction.

I saw a Martin Jetpack in person once, at a Morgo event at Waitangi in 2009. After a couple of hours of faffing around, the pilot managed a few feet in the air, with the help of two assistants. His jetpack’s 2-litre, V4 petrol engine sounded as loud as 10,000 lawn mowers.

It would manage more impressive test flights over the next few years, but never any sales (it was priced at US$250,000 in 2016) outside of its shareholders.

It was uninsurable for thrill-seekers and too unstable in strong winds for defence or rescue, where cheap-as-chips drones were coming into favour.

But the real bombshell came in 2017. Investors were looking for a ray of sunlight after Martin Aircraft reported a $24m loss and had run down its cash reserves. Instead, they were told the company’s Rotron RT1200 engine needed a complete stripdown and maintenance after just 10 hours of flight – well short of the 1500 hours that is standard for small aircraft.

Martin would have to start from scratch with a new engine design – as long as investors chipped in a fresh $50m in capital.

None came forward. Martin Aircraft delisted from the ASX in June 2018 and laid off nearly all of its 100 staff. It ceased trading in February 2019.

In late 2021, Martin Jetpack would finally manage its first sales – or at least revenue for its liquidators – as various bits of hardware and several (non-operational) jetpacks were put up for grabs by the firm’s liquidators, including one via a $1 auction on Trade Me, which went to an anonymous private buyer for $158,000, and another bought by Auckland’s Museum of Transport and Technology (Motat) for $37,600.

Retired rescue helicopter pilot Ralf Rodl, who lost all of his A$1.1m in retirement savings as his Martin Aircraft shares went to zero, unsuccessfully tried to launch a class action related to the jetpack maker’s 2014 prospectus that, in his view, overhyped the timeline toward commercial sales.

Then NZ Shareholders Association head John Hawkins said that while he felt for Rodl, “Putting all your eggs in one basket is a high-risk strategy.” More so when said egg basket is strapped to a jetpack.

Lime’s faceplant

The e-scooter ridershare market is no stranger to tech blunders.

In 2019, Lime sought to allay speeding fears with new GPS-enforced speed limits – only for a spate of accidents where riders flew over the handlebars after sudden auto-braking as they entered a speed limit area, defined by a (invisible to the rider) geo-fence. The council ordered the Lime scooters off its streets in February, and the United States-owned firm lost its contract in November.

It would turn out to be a brief pause, however. In 2020, a complex deal brokered by Uber (which had stakes in both companies) saw rideshare scooter firm Jump buy Lime and combine operations. Jump – soon to be rebranded Lime after the pair’s merger – already had a contract with the Auckland Council.

Beam me up

By 2024, e-scooters – formerly branded “the cockroaches of the footpaths” – seemed relatively under control, thanks to speed limits in most areas, more bike lanes, more stable designs and reduced numbers under revised council contracts.

But why did it seem there were actually many more of them? The Auckland Council said Singapore-owned Beam was gaming an app used to monitor numbers and, in fact, exceeding its cap by 40%, putting some 400 more scooters on the street than it was allowed.

Beam said it was merely trying to allow for a few being out of action at any one time, but inter-company Slack messages published by the Australian hinted at a more deliberate plan to beat limits.

“The evidence suggests that these anomalies were intentional, with Beam providing misleading data to appear compliant with cap limits,” the Auckland Council said as it cancelled Beam’s licence and referred the matter to police (who told the Herald it should be pursued as a civil case. The council said it was offered a cash settlement by Beam but declined).

Multiple Australasian cities gave Beam its marching orders. The firm was folded into another Singapore e-scooter rideshare operator, Neuron, in September last year.

You know I can’t see your ghost data centres

It’s breakfast time, September 1, 2025, and a nation falls into total confusion about a mysterious $7.5 billion investment.

On that day, Prime Minister Christopher Luxon breathlessly relayed some news to ZB host Mike Hosking, shortly after 7am:

“Mike, this is an incredible story, right? Amazon, one of the biggest companies in the world, probably three and a half times the size of New Zealand – AWS, which is their web services, they do all the cloud storage, data centres and all that stuff; they’re investing $7.5 billion in New Zealand. That will create about 1000 jobs. And it’s about an $11 billion boost to our GDP.”

(For those keeping score, New Zealand’s GDP was $435b for the year to June 2025, according to Stats NZ. Amazon’s market capitalisation was recently US$2.38 trillion.)

The size of the apparent investment wasn’t news.

As various people – including keen readers of the Herald’s Tech Insider column – would soon point out, Amazon, after meetings with then PM Jacinda Ardern – first announced back in 2021 that it would spend $7.5b over 15 years building and then running a cluster of three data centres, creating 1000 jobs and, according to a study commissioned by Amazon, adding $10.8b to NZ’s GDP over the period.

But that wasn’t actually what Amazon was trying to announce on September 1.

Instead, the tech giant was trying to convey that its Auckland data centre “region” (AWS-speak for a cluster of at least three independently powered data centres) was, in the company’s words, “now open”.

That is, Amazon had fulfilled its pledge to have it up and running by the end of 2025 (after missing its original target of the end of 2024, after drainage issues held up consent). It was live.

But … where was it?

Where was the photo op of Luxon with 1000 new workers, in front of a glistening new data centre?

The groundwork had been laid, literally. As Tech Insider reported in February, property records show Amazon had acquired three parcels of adjoining land in Westgate, northwest Auckland, totalling 41,774 square metres, or 4.18 hectares – that’s the size of four rugby fields.

The last transaction was only approved by the Overseas Investment Office in January this year. But after preliminary groundwork, the Naylor Love project stalled. The Auckland Council said it wasn’t a drainage issue, which had been resolved. There were no outstanding consent issues. Each time the Herald visited the giant site during the year, it only got emptier, until Naylor Love dismantled its cabins (the construction firm said it couldn’t comment, citing an NDA).

It seemed likely Amazon had abandoned the build in favour of leasing space at data centres already built in northwest Auckland by third-party operators CDC (half-owned by Infratil) and DCI (owned by an erstwhile Infratil partner, Canada’s Brookfield).

Amazon refused to say – or even confirm or deny – if it had bought then carried out earthworks at the huge Westgate site (notwithstanding that everyone from various public agencies to a local developer to Google Maps had publicised the address).

The tech giant did say its original claims – a $7.5b spend, 1000 jobs and $10.8b added to NZ’s GDP – all still stood.

That left me scratching my head, along with the rest of the media, many people, and, I’m guessing, the PM.

Microsoft’s giant data centre at Westgate cost “only” $1b.

CDC’s entire revenue for the year to June 2025, from 20 data centres across New Zealand and Australia (its major market) was A$533m ($612m).

Data centres are highly automated, requiring only a few dozen staff. The Microsoft and DCI data centres at Westgate, and the giant CDC facilities at nearby Hobsonville and Silverdale (which cost around $300m a piece), only ever have a few cars parked outside.

And while there are speed and (for some) legal advantages to local data centres – rather than being served by the tech giant’s facilities in Sydney and Melbourne – it’s contestable whether that’s a $10.8b advantage.

Some AWS users say they don’t give a hoot if Amazon built its own data centre or co-located with rivals, as long as they enjoy the benefits that come with a local AWS region.

And the firm’s power supplier, Mercury, confirms Amazon is chugging tens of megawatts of power, indicating it has leased a significant amount of space from its peers.

But as Luxon discovered, in political terms, it’s more photo-flop than photo-op.

Chris Keall is an Auckland-based member of the Herald’s business team. He joined the Herald in 2018 and is the technology editor and a senior business writer.