The (now Japanese-owned) company that runs car parks across the UK has fallen into administration over its unsupportable debts. CC-licensed photo by Elliott Brown on Flickr.
You can sign up to receive each day’s Start Up post by email. You’ll need to click a confirmation link, so no spam.
A selection of 10 links for you. Reverse out. I’m @charlesarthur on Twitter. On Threads: charles_arthur. On Mastodon: https://newsie.social/@charlesarthur. On Bluesky: @charlesarthur.bsky.social. Observations and links welcome.
Car park firm NCP collapses with nearly 700 jobs at risk • BBC News
Mitchell Labiak and Emer Moreau:
»
National Car Parks (NCP) has gone into administration, putting 682 jobs at risk.
The administrator, PwC, said demand for parking had not recovered to pre-Covid levels, pointing to “shifts in commuting and customer driving patterns”. It said NCP could no longer afford to pay its creditors after consistently losing money and was unable to scrap “long-term, inflexible” leases on loss-making sites.
PwC is looking to sell the business as the “best outcome” for those NCP owes money to. “All sites are open, staff remain in post, and trading continues as normal,” PwC added. “We will be engaging with landlords, employees, and other stakeholders as we explore all options.”
NCP was founded in 1931 and is one of the biggest car park operators in the UK. It runs 340 car parks across the country, including in airports, hospitals and train stations. The firm’s debts were £305m greater than the value of its assets, as of 30 September last year, according to a filing from its parent company.
Among the options to secure NCP’s future is to sell some or all of the company.
PwC said NCP had a “high concentration” of inflexible leases that prevented it from reducing costs or scrapping unprofitable car parks. Zelf Hussain, joint administrator and PwC partner, said the company had faced “a challenging trading environment” in recent years.
“Our priority on appointment is to ensure continuity of service while we undertake a detailed review of the business.”
NCP’s parent company, Park24, which is Japanese, said higher energy prices as a result of the outbreak of war in Ukraine in 2022 also put pressure on the business, adding to its operating costs.«
When you look at the financials, it actually made a slim operating profit (£7.6m on £233m) last year, and in 2024, but a hefty loss in 2023. But the leases and other costs dragged it into consistent losses.
NCP going bust is an eventuality I’d never have expected, based on the cost of parking there. But the trend to Work From Home has murdered it.
unique link to this extract
Changes to UK working habits push car park group into administration • Financial Times
Megan Snaith, with some fabulous historical detail about NCP:
»
The company was founded by Colonel Frederick Lucas in 1931 with a small car park in London.
It grew rapidly after it was bought by Donald Gosling and Ronald Hobson who combined it with their business, Central Car Parks, in 1959. They had spotted the opportunity of converting city bomb sites into car parks, having paid £200 for their first bomb site in Holborn, central London in 1948. NCP was sold to US company Cendant for £800m nearly 40 years later.
That deal almost broke down at the last minute after Hobson objected on the basis that he wanted to keep a single parking space for himself in central London off Oxford Street. He wanted to maintain his weekend habit of packing a flask of tea and driving there in his Bentley to try to predict where people would park.
NCP has changed hands multiple times since then. In 2007, it was bought by Macquarie, which later sold it to its current owners, Japan’s Park24 in 2017.
«
That detail about bomb sites as potential car parks always stuck for me. That’s a real “lemonade from lemons” insight.
unique link to this extract
Gamblers trying to win a bet on Polymarket are vowing to kill me if I don’t rewrite an Iran missile story • The Times of Israel
Emanuel Fabian:
»
on X, I saw a user reply to a recent tweet of mine: “There are people saying that they have received word from you that the missile strike in Beit Shemesh on March 10th was in fact intercepted, is this true or did no such interaction occur?”
Another X user responded to my post with the video showing the Iranian ballistic missile impact in Beit Shemesh with: “was there any video of the actual impact.” (Clearly, he didn’t watch the video.)
Checking those X accounts, both appeared to be involved in gambling on the Polymarket betting site.
As far as I now understand, the emails I [had previously] received were intended to confirm whether or not a missile had hit Israel on March 10 in order to resolve a prediction on Polymarket.
Polymarket is one of the largest prediction markets in the world, where users can wager their money on the likelihood of future events, using cryptocurrency, debit or credit cards, and bank transfers. However, there are accusations that the site has been plagued by manipulation and insider trading.
The event that these people had bet on was “Iran strikes Israel on…?” More than 14 million dollars had been wagered on March 10.
The rules of the bet state: “This market will resolve to ‘Yes’ if Iran initiates a drone, missile, or air strike on Israel’s soil on the listed date in Israel Time (GMT+2). Otherwise, this market will resolve to ‘No’.”
However, there is a clause: “Missiles or drones that are intercepted… will not be sufficient for a ‘Yes’ resolution, regardless of whether they land on Israeli territory or cause damage.”
My minor report on a missile striking an open area was now in the middle of a betting war, with those who had bet “No” on an Iranian strike on Israel on March 10 demanding I change my article to ensure they would win big.
More emails arrived in my inbox.
«
I thought that Polymarket didn’t allow betting on wars. But of course it does. And whereas we used to think prediction markets would be a great way to guess at the future – on the “wisdom of crowds” principle – it turns out that giving people a financial interest in bending truth to match their prediction will distort them rather than give us the reading we wanted.
unique link to this extract
Google scraps AI search feature that crowdsourced amateur medical advice • The Guardian
Andrew Gregory:
»
Google has dropped a new artificial intelligence search feature that gave users crowdsourced health advice from amateurs around the world.
The company had said its launch of “What People Suggest”, which provided tips from strangers, showed “the potential of AI to transform health outcomes across the globe”.
But Google has since quietly removed the feature, according to three people familiar with the decision. A Google spokesperson confirmed “What People Suggest” had been scrapped. The move came as part of a “broader simplification” of its search page and had nothing to do with the quality or safety of the new feature, the spokesperson said.
The revelation comes as the company faces mounting scrutiny over its use of AI to provide millions of users with health information and advice. In January, a Guardian investigation found people were being put at risk of harm by false and misleading health information in Google AI Overviews. The AI-generated summaries are shown to 2 billion people a month, and appear above traditional search results on the world’s most visited website.
Google initially sought to downplay the Guardian’s findings. The AI Overviews that alarmed independent experts linked to reputable sources and recommended seeking expert advice, the company said. Days later, Google removed AI Overviews for some but not all medical queries.
In March last year at an event in New York, Google said it planned to expand medical-related AI summaries in search. The company said it was adding a new feature, “What People Suggest”, which aimed to provide users with information from people with similar lived medical experiences.
On the day of “The Check Up” event, Karen DeSalvo, then Google’s chief health officer, wrote a blog post outlining why the company was launching the new feature, and how it would help users.
“While people come to search to find reliable medical information from experts, they also value hearing from others who have similar experiences,” wrote DeSalvo. “That’s why we’re making it even easier to find this type of information on Search with a new feature labelled ‘What People Suggest’.
«
Is it worth pointing out that this was always an astonishingly bad idea, but Google doesn’t seem to learn from its mistakes?
unique link to this extract
Encyclopedia Britannica is suing OpenAI for allegedly “memorizing” its content with ChatGPT • The Verge
Stevie Bonifield:
»
On Friday, Encyclopedia Britannica and dictionary publisher Merriam-Webster filed a lawsuit against OpenAI alleging that it used their copyrighted content to train its AI, then generated responses that were “substantially similar” to their content, as previously reported by Reuters.
According to Britannica, OpenAI repeatedly copied its content without permission, stating, “GPT-4 itself has ‘memorized’ much of Britannica’s copyrighted content and will output near-verbatim copies of significant portions on demand. The memorized examples are unauthorized copies that [OpenAI] used to train their models, including GPT-4.”
The lawsuit goes on to include examples of responses from OpenAI’s models side by side with Britannica’s text, in which entire passages appear to match word for word. Britannica also claims that OpenAI has been “cannibalizing” its web traffic by generating responses that “substitute, or directly compete” with Britannica’s content, rather than directing users to its website the way a traditional search engine would.
«
Join the queue, Britannica, though its complaint does have some merit. The content of an encyclopaedia isn’t as random as the rest of the web’s contents; it’s filled with facts, but those facts are organised in a particular way.
The grumble that OpenAI isn’t sending clicks in the way that a search engine does: welcome to the new paradigm, people.
unique link to this extract
AI agents are taking over contract negotiations • IEEE Spectrum
Michael Dumiak:
»
Some of the world’s largest companies with the biggest supply chains—including Walmart, the global shipping giant Maersk, and the telecom servicer Vodafone—are now using bots powered by artificial intelligence to negotiate and maintain supplier contracts.
That these sophisticated AI systems were designed and built by a startup in Estonia is interesting; it’s even more notable that bots now routinely engage in automated contract negotiations for sprawling global enterprises. But what’s really eye-opening is that these AI agents aim to work autonomously. Which prompts a question: What will happen if the AIs start to haggle amongst themselves?
“In the future I can imagine all sorts of agents in the real physical world negotiating with one another,” says Tim Baarslag, a senior researcher in intelligent and autonomous systems at the Centrum Wiskunde & Informatica in Amsterdam. “Letting these bots run completely wild, I think, requires more research.”
Baarslag has wrestled with negotiation bot concepts for years (one of his peers has a running project called Pocket Negotiator). In 2017 he and his colleagues published “When Will Negotiation Agents Be Able to Represent Us?” They drew a sharp line between automated and autonomous negotiation. The difference is the freedom to negotiate independently.
The five-year-old Estonian startup Pactum is clearly marketing its bot as an autonomous agent. In addition to Maersk and Walmart, its client list now includes a wire and cable supplier and an electrical supply wholesaler (once part of Westinghouse). The startup landed a US $20m venture capital investment in July from backers including Maersk itself.
…Pactum calls its agent an autonomous negotiation suite. The system’s machine learning can analyze a massive set of complex contractual terms using historical and market data from both within and outside the company. It can send its analysis to a human user, such as a buyer or procurement officer—or, on its own, it can produce and forward a set of contract options to a vendor (mostly based on price, delivery dates, and billing cycles). The bot can take counteroffers and respond.
«
Companies House vulnerability enabled company hijacking • Tax Policy
Dan Neidle:
»
A major vulnerability in the Companies House website gave unauthorised access to the private dashboard of any of the five million registered companies for five months. It exposed directors’ home addresses and email addresses, and appears to have enabled attackers to change company and director details – and even file accounts.
This article sets out what we know, what we don’t, and what businesses should be doing to protect themselves.
What we confirmed: unauthorised access to any company’s dashboard; exposure of non-public personal data; submission of a change that generated a confirmation number.
What remains unconfirmed: whether the change was actually processed; how long the vulnerability existed; whether it was exploited by criminals; whether Companies House can identify affected companies.
The vulnerability is incredibly simple, and involves just pressing the “back” key at a particular time.
The vulnerability was discovered on Thursday 12 March by John Hewitt at Ghost Mail, a corporate services provider. He tried to contact Companies House immediately, but didn’t get a response – so he contacted us.
…John used the vulnerability to view the private Companies House dashboard of ClarityDW Ltd, a digital communications consultancy owned by Jonathan Phillips. Jonathan kindly gave us permission to do this.
John then used it to view the dashboard of a company I own, and to modify my own registered address. That appeared to work, as it generated a confirmation number. As you will hear, I was incredulous at what John showed me.
I then spoke to computer security specialists. To rule out the possibility that it was something specific to John’s computer, network or account, I tested the vulnerability myself… This shows the exploit revealing private information that’s not published by Companies House, such as personal email addresses and full dates of birth (and you can see that in the video, with Jonathan’s personal information masked).
«
On Monday, Companies House released a statement acknowledging that “It may also have been possible for unauthorised filings — such as accounts or changes of director — to have been made on another company’s record.” So now every company just has to go through all those records and check they haven’t changed. The demand will probably overload the Companies House servers.
unique link to this extract
Behind the curtain: Trump’s escalation trap • Axios
Jim VandeHei and Mike Allen:
»
Trump could pull out tomorrow. But the Iranians could keep the Strait of Hormuz closed and push oil prices so high that America would have to re-engage.
The Iranians have made it clear in private and in public that even if Trump decides to end the war, they could continue shooting missiles and rockets until they get guarantees that this is the end of the war, not just a temporary ceasefire.
Behind the scenes: Trump has grown accustomed to doing what he wants and then quickly improvising if things go south. But this time, some in his inner circle have what one official called “buyer’s remorse” — growing fears that attacking Iran was a mistake.A source close to the administration said some key officials around Trump were reluctant or wanted more time. “He ended up saying, ‘I just want to do it,'” the source said. “He grossly overestimated his ability to topple the regime short of sending in ground troops.”
The source said Trump was “high on his own supply” after last summer’s quick strikes in Iran and January’s abduction of Venezuelan President Nicolás Maduro: “He saw multiple decisive quick victories with extraordinary military competence.”
Reality check: Trump’s war of choice certainly looks like a military success so far. Iran’s missile and drone launches have greatly decreased, indicating it’s running out of weapons or the ability to fire them.
«
I’m not sure about the reality check. Iran has an unknown number of drones, and mines, which it can use to harass transport in the Strait.
unique link to this extract
Why Hormuz will haunt us long after this war ends • Financial Times
Gideon Rachman:
»
The closure of the Strait of Hormuz is one of the most foreseen “unforeseen problems” in history. For decades, academics and game theorists have speculated about the possibility that, in wartime, Iran could choke off the narrow waterway through which 20% of the world’s oil exports pass.
Donald Trump was warned of the danger to the strait as America and Israel prepared to attack Iran. But the US president waved away these concerns, predicting instead that the Islamic republic would swiftly capitulate.
A conflict with Iran that started with vague war aims now has one clear and overriding objective: reopen the Strait of Hormuz. Ironically and infuriatingly, the only reason the strait is closed is because the US and Israel went to war in the first place.
It is not in Trump’s power to reopen this vital sea passage by declaring victory and walking away. Instead his war with Iran — and the particular issue of the Strait of Hormuz — will define the rest of his presidency and may haunt his successors.
That is because the strait’s closure creates both an immediate crisis and a long-term strategic quandary. The current problem is that the longer it is closed, the greater the threat of a global recession. The future dilemma is that Iran now knows that control of the Strait of Hormuz gives it a stranglehold over the world economy. Even if it relaxes its grip in the short term, it can tighten it again in future.
The difficulties of reopening the strait are already very apparent. Iran does not have to sink or impede every tanker that tries to pass through. The spate of attacks already carried out — and the threat of new ones — has been enough to persuade ship owners, crews and insurers to steer clear.
…Trump is now asking America’s allies to send their navies to break the Iranian chokehold on the strait. He has even appealed to Beijing. The UK, the EU and China do have a real interest in reopening the Strait of Hormuz. But they will be understandably reluctant to put their own forces at risk to solve a problem that they did not create and that the US navy cannot fix on its own.
A year of tariffs, threats and insults from the Trump administration towards its European allies has also burned through goodwill towards Washington. They also know that any navy operating in the Strait of Hormuz would be very vulnerable to Iranian attacks — and might have to keep the operation up over many months.
«
The Curse of the Long Boom • WIRED
David Karpf:
»
Back in February I posted some reflections on what I’m looking for in the WIRED back catalog. Here’s how I phrased it then: “All of my best thinking comes from getting stuff wrong. That’s the angle from which I approach all social science research questions.” The whole point of making predictions, from this perspective, is to help yourself identify the limits of your own knowledge, leading to harder questions that improve your understanding of the world.
Credit to Dan Davies—his coinage [“if you don’t make predictions, you never know what to be surprised by”] is so much pithier.
He also provides a corollary: “If you don’t make recommendations, you won’t know what to be disappointed by.”
Let me offer a second corollary: “If you retrofit your predictions to insist they were right after all, you’ll never learn a single damn thing.”
I mention this because, as you might imagine, I run into a lot of incorrect predictions as I read through the WIRED archive. Back in the ’90s, WIRED was chock-full of a very particular style of futurism—one that has not aged especially well.
In keeping with Davies’ Law, this presents a lovely opportunity. I’m rereading the entire magazine back catalog to understand how emerging technologies looked, take stock of where people thought the world was headed, and draw lessons from the resulting surprises.
The thing that sets me back on my heels, though, is that a lot of those old WIRED techno-optimists are still out there making predictions today. And to hear them tell it, they were right all along.
Huh?
«
The folks at Wired have been cheating to protect their Long Boom story for ages. Karpf has been detailing all Wired’s mad but also self-correcting predictions on his Substack, but has ceased writing there. (Thanks Lloyd W for the link.)
unique link to this extract
| • Why do social networks drive us a little mad? • Why does angry content seem to dominate what we see? • How much of a role do algorithms play in affecting what we see and do online? • What can we do about it? • Did Facebook have any inkling of what was coming in Myanmar in 2016? Read Social Warming, my latest book, and find answers – and more. |
Errata, corrigenda and ai no corrida: none notified
The Overspill: when there's more that I want to say 