TheChek News Flow

In October, on a panel at an investor conference in Riyadh, Saudi Arabia, futurist Peter Diamandis congratulated a technology entrepreneur sitting next to him, Jack Hidary, on the success of his startup, SandboxAQ. “I’ve never seen a company scale revenues as quickly as you have,” said Diamandis, founder of the nonprofit XPrize Foundation.

Also on the panel was Eric Schmidt, former CEO of Google, who has similarly shown his enthusiasm for SandboxAQ—which focuses on two buzzy technologies, quantum computing and artificial intelligence—by becoming one of its most prominent financial backers and its executive chair. Earlier this month, Schmidt participated in a $300 million funding round that valued SandboxAQ at a whopping $5.6 billion, almost double its valuation in a 2022 funding round.

But there are reasons to be cautious about the company’s outlook. According to interviews with more than a dozen current and former SandboxAQ employees, Hidary and the startup have embellished some of its success when communicating with investors.

For example, it said in an investor presentation in recent months that by the end of the year it would reach $50 million in annualized sales—its fourth-quarter sales multiplied by four—and had amassed more than $110 million of signed contracts. In the presentation, it displayed the logos of more than two dozen companies and organizations, including the U.S. Air Force, Wells Fargo, Vodafone and Boeing, alongside text that said SandboxAQ sells “to the largest, most impactful enterprises on the planet.”

Another slide said SandboxAQ, which was selling software for cybersecurity, had “renewed 100% of licensed contracts, with many expansions.”

The implication is that all of the companies mentioned in the presentation are currently customers. But a few of the companies contacted by The Information said they were not customers. Vodafone said it had only tested SandboxAQ’s technology, and a Boeing spokesperson said the company had a “collaboration agreement” with SandboxAQ that gave the company access to its flight test platform, but no longer has a commercial relationship with SandboxAQ. Wells Fargo carried out a proof-of-concept test and no longer uses SandboxAQ’s products, former employees said. The bank declined to comment.

Commercial Struggles

SandboxAQ is also facing difficulties commercializing its technology, according to former employees. The firm develops large quantitative models, numeric-focused alternatives to the large language models that power apps like OpenAI’s ChatGPT. Those LQMs can help pharmaceutical companies discover new drugs or governments to invent new materials, such as plating for military vehicles.

But SandboxAQ has struggled to make its LQMs into one-size-fits-all products for customers, according to former employees. Instead, it has had to painstakingly retool them for specific use cases, such as modeling the effects of a new drug for Parkinson’s disease or discovering materials for a new luxury product made by Hermès, the former employees said. This has made it time-consuming to develop its technology, former and current employees said. Recently, SandboxAQ laid off most of its sales staff.

SandboxAQ’s vice president of product, former Amazon executive Nadia Carlsten, left in August to become CEO of the Danish Center for AI Innovation. Before she left, Carlsten complained to staff that Hidary had exaggerated what the company’s existing products could do for customers, former employees said.

Former employees said SandboxAQ’s internal tool for forecasting potential sales was overly optimistic. The former employees said workers regularly flagged their concerns about the internal projections to Hidary and his direct reports. In recent months, the company has revised those projections, former employees said.

Still, the company has notched some notable successes with customers. It told investors that its largest customer—a laboratory at the University of California, San Francisco, focused on neurodegenerative disease—spent $20 million with SandboxAQ this year. The lab could spend up to $25 million more, depending on the achievement of certain milestones, it said in an investor presentation. And SandboxAQ recently renewed a contract with the U.S. Air Force to develop a new kind of GPS technology.

The situation underscores the inherent uncertainty facing investors trying to assess startups in any sphere of the tech industry, where the underlying technology often is highly complex and understood by few. For that reason, investors often put undue weight on the connections founders have with others, including big-name executives. These issues are exacerbated in AI, where many companies are still struggling to turn cutting-edge developments and world-leading theoretical research into commercial products that can generate a profit.

Whether SandboxAQ can fulfill its ambitions depends largely on Hidary, a well-connected, energetic founder who has dabbled in politics and first made a name for himself during the dot-com boom of the late ’90s. He eventually incubated what became SandboxHQ inside a Google research lab, where he clashed with other staffers, according to people who worked there at the time.

Still, that friction didn’t hurt his relationship with Schmidt, who has stumped for Hidary with prospective investors in recent months. That has helped Hidary attract an all-star list of financial backers including Salesforce CEO Marc Benioff, early Facebook investor Jim Breyer and Meta Platforms AI chief scientist Yann LeCun.

In a video interview earlier this year, Schmidt raved about Hidary. “I’ve been fortunate in my life to work with people much smarter than I—and when I find them I collect them,” he said. “One of them is Jack Hidary.”

In a statement, Breyer said he was “proud to be backing SandboxAQ and its impressive accomplishments,” adding that the startup “is the pacesetter in quantitative AI which impacts large sectors of the economy."

Cutting deals, and spending freely to win them, seems to be another focus for Hidary. His travel and entertainment spending, including bills to wine and dine prospective investors and commercial clients, has been one of the largest items of expenditure at the company, according to former employees and people familiar with the company’s expenses.

Hidary’s willingness to exaggerate SandboxAQ’s successes isn’t uncommon among tech founders, among whom a fake-it-till-you-make-it ethos is commonplace.

“That’s how the deep tech environment often works—you get your narrative out there, make people believe it, raise enough capital to solve enough problems and do that before the chickens come home to roost,” Tanveer Kathawalla, managing partner of Pioneer1890, a venture capital firm that invests in companies catering to government clients.

Representatives for SandboxAQ, Hidary and Schmidt declined to provide comment on the record. A spokesperson for Alphabet, the parent company of Google, declined to comment on the record.

A Mayoral Run

Hidary has said he dropped out early from Columbia University after winning a Stanley Fellowship in Clinical Neuroscience at the National Institutes of Health. A former lab chief for the NIH, Dr David Pickar, said in a statement that while “we usually focus on medical students for this position, we were very pleased to grant this fellowship to Jack.”

Afterwards, Hidary made a dramatic entrance on the technology scene. In 1998, he took EarthWeb, the early internet company he co-founded, public in one of the first and most spectacular listings of that frothy era. But EarthWeb floundered after the dot-com bubble burst in 2000. Hidary sold off an array of the company’s assets and in 2001 agreed to step aside as CEO.

He quickly launched his next venture: Vista Research, a financial research company connecting experts with investors, which McGraw-Hill, a division of Standard & Poor’s, bought in 2005 for a reported $40 million. The company sold the entity at a loss a few years later.

In 2001, Hidary also established a charitable organization, the Jack D. Hidary Foundation, which initially focused on worldwide economic development, according to an archived webpage for the charity, and was later described as having a focus on medical oncology research, according to more recent mentions of the charity by Hidary.

However, a minority of its grants went to cancer-affiliated groups or medical causes over two decades, regulatory filings show. The foundation also ended up giving money to numerous groups associated with Hidary, including $400,000 to OpinionSource, an online opinion aggregation service Hidary founded in 2004.

It also made a $25,000 loan to iAmplify Inc., a now-defunct web-based content publishing company started by Hidary and his brother Murray in 2005. Hidary’s foundation wrote off the loan in 2016, six years after making it, filings show. The foundation, to which Hidary contributed a total of over $2 million over two-plus decades, has reported holding less than $10,000 in assets in each of the past seven years.

In 2013, Hidary parlayed his entrepreneurial reputation and wealth into a run for New York City mayor, with a promise to encourage startup investments in the city. Celebrity economist Nouriel Roubini hosted a penthouse fundraiser for his campaign. One of Hidary’s tongue-in-cheek pledges was to give every New York citizen a pair of Google Glasses, the early smart eyewear from the company, if and when he was elected.

That didn’t happen. Hidary won only around 3,400 votes, coming in fifth place behind the winner, Bill de Blasio.

Moon Shots

Hidary met Schmidt, then Google’s CEO, through their mutual involvement with Diamandis’ XPrize Foundation, a nonprofit that hosts public invention competitions.

Later, Hidary came to know Google co-founder Sergey Brin, who named Hidary as an external adviser in 2016 to help a group of researchers inside X—the Alphabet lab it also refers to as its moon-shot factory—examine the intersection of AI and quantum technologies. Two years later, Alphabet formally hired Hidary to lead what became known as Sandbox@Alphabet, a collection of theoretical experts looking at ways to turn cutting-edge quantum research into products, according to former Google employees.

But Sandbox@Alphabet didn’t quite fit in with the rest of the teams at X. Hidary butted heads with X director Astro Teller over the direction of Sandbox, former employees said.

Sandbox’s interest in working with government entities to develop quantum and cryptography products also put it at odds with the rest of X’s staff, who viewed such military work with suspicion, the former employees said.

Teller didn’t respond to requests for comment.

A former X employee said some members of the Sandbox@Alphabet team were frustrated by the lack of resources coming from Alphabet for Sandbox’s projects. They added that the group also lacked a cohesive plan or strategy, delving into disparate fields such as quantum navigation, drug discovery and chemistry.

One senior X researcher at the time, Guifré Vidal, complained to Hidary about what he saw as Sandbox@Alphabet’s tendency to rush toward creating products before nailing down the science underpinning its projects, two former X employees said.

After several years of overseeing experimentation and research inside X, the frictions inside the group eventually prompted Hidary to strike out on his own. In 2022, he recruited several dozen researchers from Sandbox@Alphabet to join him in a new venture, SandboxAQ. He raised more than $450 million for his startup, a corporate filing shows, valuing it at nearly $3 billion.

Alphabet didn’t invest in the new company, and many employees didn’t follow him to it. But Hidary corralled investments from a coterie of powerful figures, including Schmidt, Breyer and Benioff. T. Rowe Price, its highest-profile institutional investor, put in a relatively small amount of money, a person close to SandboxAQ said.

‘AI or Die’

At SandboxAQ, Hidary has preached a consistent message as he has tried to win business for the company: Even if powerful quantum computers are years away from practical use, it’s not too early for governments and businesses to adopt related technologies.

At the moment, quantum computers are still experimental technologies, but some of the companies investing in the technology are making progress, most notably Alphabet, which announced earlier this month that Willow, its quantum computing chip, had made a significant breakthrough.

Even without yet having quantum computers, SandboxAQ says it uses technologies derived from quantum computing on existing hardware, including chips from Nvidia. Hidary himself has sought to become a public authority on the technologies by writing a textbook on the practical applications of quantum computing in 2019 and a new book, “AI or Die.”

There are signs that SandboxAQ has been overoptimistic in its view of how eager investors are for a piece of the company. While the $300 million it raised was a tidy sum for a young startup, Hidary had originally sought to raise $500 million this year, according to an October update to the company’s corporate charter. Among the new investors in the company was asset manager Fred Alger Management, T. Rowe Price and the sovereign wealth fund of Bahrain.

The $5.6 billion valuation made Hidary—who owns more than 20% of SandboxAQ, a corporate filing indicates—a likely billionaire, at least on paper. The funding round valued SandboxAQ at more than 100 times what it has told investors this year’s annualized sales will be. Its valuation is higher than the combined market capitalizations of Recursion and Schrödinger, two larger, publicly held drug discovery firms.

Current and former employees said poor oversight contributed to a lack of focus at SandboxAQ. They cite a brief effort by researchers—at Hidary’s instigation—to contemplate new lines of work involving AI robotics, which were not the researchers’ areas of expertise and didn’t end up going anywhere. There were other side projects, like AI-powered management tools for distribution of emergency supplies, which former employees said was directed at Covid-19–era health supplies well after the pandemic had subsided.

While Schmidt is chair of the company, its board’s only other members are Hidary and his longtime accountant. That’s a smaller lineup than for other companies that have attained a similar valuation. SandboxAQ also has an advisory board stacked with influential individuals such as Larry Summers and Blythe Masters.

The company appears to have taken steps to keep a tighter lid on costs. In September, SandboxAQ hired Andrew McLaughlin, a former global public policy executive at Google, as its first chief operations officer. One of his first acts was to lay off roughly 10% of SandboxAQ’s staff. That included around 20 employees and leaders focused on securing contracts with government and commercial entities, who represented most of its sales team, according to current and former employees.

The entrepreneur's iconic boats have been coveted by the world's rich and famous for decades.

MILAN — Paolo Vitelli, who founded Azimut Yachts in the late 1960s, has died at age 77, the company confirmed Tuesday. The company said it was not ready to release a statement.

Azimut, which is known for its luxury yachts coveted by the world’s rich and famous, said the entrepreneur died at his home in the hamlet of Mascognaz, located in the northern Italian region of Aosta. He slipped on ice and fatally hit his head on concrete.

In the 1980s, Vitelli acquired Cantieri Benetti, a historic Viareggio-based shipyard and created the Azimut|Benetti Group.

Today, Azimut|Benetti Group is ranked as the number-one firm in the world according to length built by Boat International. Vitelli’s daughter Giovanna has been chairman of the group since March 2023.

In October, Azimut|Benetti Group, which is headquartered in Avigliana, Italy, near Turin, said its revenues rose to 1.3 billion euros in the fiscal period between 2023 and 2024, compared to 1.2 billion euros last year, driven by key markets. At the time, its chief executive officer Marco Valle forecast 15 percent growth by the same period of 2025.

Paolo Vitelli was born in Turin, Italy in 1947 into a family of entrepreneurs. Driven by a passion for boats and yachting, he set up his business, Azimut, at just 21 years old, the company said on its website.

After graduating with a degree in economics and commerce, at the age of 22, he became an importer of boats produced by some of the most prestigious shipyards in Northern Europe.

By the mid-1970s, Azimut had launched a series of boats and later expanded the range from 8 to 10 meters to larger models, up to the iconic Azimut Failaka 105′ (30.8 meters), which was sold to the Onassis family.

More than half of Azimut’s fleet is now made up of low-emission yachts that reduce CO2 emissions by between 20 and 30 percent. Seadeck 6 integrates an innovative system that allows both zero-emissions at anchor and cruising with the generator off, further reducing on-board consumption. This system, called Mild Hybrid Zero Emission Hotel Mode, is based on a 42-kWh lithium battery pack and an alternator connected to one of the three engines. Overall the vessel reduces CO2 emissions by 40 percent over a year of average use, both when cruising and at anchor, compared to a traditional flybridge boat of similar dimensions.

The company said it has invested 20 years of research and development in sustainable innovation and is working toward meeting the goals set by the International Maritime Organization for the overall industry to reduce CO2 emissions by 40 percent by 2030.

‘MiniLife’ research extends growing field of synthetic biology into new territory

European scientists have started work on a project to create simple forms of life from scratch in the lab, capitalising on theoretical and experimental advances in the fast-growing field of synthetic biology.

Starting with inanimate chemicals, the researchers aim to produce metabolically active cells that grow, divide and show “Darwinian evolution” within six years.

The €13mn “MiniLife” project, which is funded by the European Research Council and involves biologists and chemists from several universities, could be the first in the world to reach the minimum criteria for a synthetic living system.

“Success would constitute a landmark achievement in basic science,” said Eörs Szathmáry, director of the Centre for the Conceptual Foundations of Science at the Parmenides Foundation in Germany, who is a principal investigator on the ERC grant. “De-novo creation of living systems is a long-standing dream of humanity.”

John Sutherland, who works on the chemistry of early life at the MRC Laboratory of Molecular Biology in Cambridge, said the project joins a growing worldwide effort to “create minimal living systems”.

Sutherland, who is not involved in the MiniLife project, added: “This is driven by the perennial desire to understand how life originated on Earth and whether it could also have originated elsewhere in the observable universe.”

Other artificial life researchers are working with the known building blocks of life on Earth, particularly the nucleotides that make up ribonucleic acid. The ERC project, in contrast, aims truly to start from scratch, without using molecules that are themselves products of evolution.

“We abstract away from known life forms because they are highly evolved creatures,” said Szathmáry, “and simplify so as to arrive at a minimalistic formulation.”

The MiniLife researchers are evaluating four systems that might, individually, or in combination, be developed into a basis for minimal life. All are “autocatalytic”, a property essential for self-replication in which a chemical reaction is catalysed by its own products.

One candidate is the formose reaction. The process, discovered in the 19th century, converts an extremely simple chemical, formaldehyde, into an increasingly diverse and complex series of sugar molecules. As the reaction is fed with formaldehyde, the droplets’ behaviour varies with the composition of sugars inside them.

“Some grow faster and divide more quickly than others,” said Andrew Griffiths, a MiniLife investigator at the École Supérieure de Physique et de Chimie Industrielles in Paris. “We end up with the emergence of something equivalent to fitness in biology, like a mixture of slow-growing and fast-growing bacteria, but in a very simple chemical system.”

The formose-based system must be able to display reliable hereditability — passing on acquired characteristics from one generation to the next — perhaps in conjunction with one of the other systems being evaluated.

The six-year timing is ambitious, said Griffiths, who is optimistic that the project will be able to “demonstrate rudimentary Darwinian evolution”. As a minimum that would involve a system that can switch between two heritable states in different environments, analogous to the famous peppered moth whose wings are white in clean environments and black when it lives in polluted places with dark surfaces.

Sijbren Otto, a professor of systems chemistry at Groningen university and another member of the MiniLife team, said his primary motivation was “fascination with the nature and origin of life. Although the molecules we develop will probably not be the ones from which life started on the prebiotic Earth 3.8bn years ago, the mechanisms we hope to unveil will be very relevant for understanding what happened then.”

Last month an international group of researchers warned of the “unprecedented risks” posed by another area of synthetic biology. They said “mirror life” — manufactured bacteria that are structural reflections of natural microbes — could overwhelm the defences of people, other animals and plants.

Asked about the safety of the MiniLife project, Otto said its creations were “extremely unlikely to have any viability outside very controlled lab conditions” and posed no possible risk to the public.

However, the team is working with experts to develop an ethical framework for the research. “Now is the time to think much further ahead to where the research is likely to lead,” Otto said.

Private credit may end up looking a bit more like a bank

As the industry expands, there are some areas that should start to attract regulatory scrutiny

Banks may find it galling to see lightly regulated private credit making inroads into their business. But by and large, the growth of the sector is good news for the financial system. Rather than turning short-term deposits into long-term loans (a mismatch that has been known to trip up banks), funds tap into genuinely long-term capital, broadly matching the duration of their loans. 

Yet that is only true insofar as private credit sticks to its core mission. As the industry expands, and stretches its tendrils towards new investors and new investment opportunities, there are some areas that should start to attract regulatory scrutiny. 

For one thing, private lending is becoming increasingly enmeshed with the traditional banking system. Banks strike partnership agreements with private credit funds, and lend them money. That is not necessarily concerning given that the debt gets sliced up and banks end up holding the less-risky tranches on their balance sheets. It is also still tiny in the context of banks’ loan books. But the overall exposure — and whether it becomes concentrated with some particular lenders, or some particular private credit managers — is worth keeping an eye on.

Another issue is private credit’s own loan quality. Where it engages in direct lending, it often serves riskier companies — especially now that banks and the syndicated loan market are again on hand for those seeking plain-vanilla debt. Payment-in-kind loans — which allow issuers to defer interest payments — have been rising, partly as a result of straitened companies seeking to refinance.

Whether this additional flexibility will just delay inevitable blow-ups — a drag on private credit groups themselves and the economy at large — or whether it will help fundamentally good companies find a path to profitability, is an open question. But as the private credit industry becomes bigger, regulators may be keen to delve more deeply into loan books.

Most obviously, any effort to attract high net worth, or even retail, money is destined for much hoop-jumping. Just think about the complexities involved in State Street and Apollo’s mooted public-private credit exchange traded fund, for which the companies requested approval in September. There are obvious questions around transparency, pricing and valuation — hard enough to do on a quarterly basis, let alone daily.

The question of liquidity is even more problematic. If more investors want to sell their ETF than there are buyers in the market, Apollo is on hand to buy a slice of the private loans to provide backstop liquidity. It may not need to, especially if the fledgling secondary market for private loans develops. But, as private credit expands, the age-old problem of how to turn short-term money into long-term loans may end up resurfacing in unexpected places. 

Belgium bans disposable vapes as concern grows over e-cigarettes

Experts say crackdowns on single-use products may have limited impact on growing industry

Belgium on Wednesday became the first EU country to ban sales of disposable vapes, as governments across the continent try to reverse growing e-cigarette use among younger people.

England, France and Germany are planning tighter regulations, but experts warn such moves are likely to have only limited impact on the rapidly growing industry. Hungary banned the ElfBar brand of disposable vapes two years ago on the grounds that the product had a high nicotine content and was being marketed to minors.

The looming controls highlight the difficulties policymakers face as they try to manage the health risks of vapes while supporting their positive impact in helping adults wean themselves off tobacco cigarettes. The moves reflect concern over how the pricing and marketing of disposable products have fuelled underage use and attracted young people who may never have smoked cigarettes.

Yet disposable vape curbs were unlikely to have much effect as consumers could easily shift to reusable products such as pod vapes and vape pens, said Caitlin Notley, professor of addiction sciences at the UK’s University of East Anglia.  

“Policies like this [Belgian ban] are important for politicians because they make a clear statement,” Notley said. “But there’s probably going to be a limited impact.”

E-cigarette use has grown sharply over the past decade. The products are less harmful than tobacco cigarettes because they do not produce tar or carbon monoxide, but most still contain the addictive ingredient nicotine. Studies have suggested vape use can damage blood vessels and lung function.

Authorities are also worried about the social and environmental effects of disposable vapes. They are generally made of hard-to-recycle mixed materials, including a lithium-ion battery.

In the past few years, single-use vapes have rapidly attracted a new generation of users who have never previously smoked.

More than half of British vapers aged 11-17 said they used disposable products more frequently than reusables, according to a survey this year by campaign group Action on Smoking and Health. Current vaping rates among the underaged cohort, which includes using the products less than once a month, is 7.2 per cent, Ash added.

Young people drove a rise in the prevalence of vaping by adults in England who had never regularly smoked from 0.5 per cent to 3.5 per cent between 2021 and 2024, according to a study published this year.

“For those who would have taken up smoking, [vaping] is a good thing,” said Sarah Jackson, co-author of the paper and principal research fellow in the tobacco and alcohol research group of UK university UCL. “For people would not have otherwise taken up smoking, starting to vape regularly will expose them to more harm than if they had neither smoked nor vaped.”

Some vaping products use designs that appeal to younger people. Disposable brands such as ElfBar and Lost Mary, both made by China’s Shenzhen iMiracle Technology, deploy tropical fruit flavours and vibrantly coloured packaging.

Vaping industry representatives argue that authorities should respond with stronger enforcement rather than outright bans on disposables. This could include licensing of retailers, as well as steep fines without prosecution for selling illegally to under-aged buyers.

“Just banning the products really isn’t the answer,” said Dan Marchant, managing director of Vape Club, the UK’s largest online vape products retailer. “We need proactive enforcement of the current regulations.”

Banning single-use vapes would drive consumers towards illegal products, industry figures add. Illicit products account for as much as one-third of the vape items sold, the UK’s Chartered Trading Standards Institute has estimated, with infringements including inaccurate tank sizes, excessive nicotine concentration and harmful materials.

All vaping products sold in European countries should be pretested or preregistered to ensure regulatory compliance, said Asli Ertonguc, western Europe area director for British American Tobacco.

“The [disposables] prohibition will create illegal markets, which is our biggest concern,” Ertonguc said. “[It will also encourage] users who are accustomed to these products to go back to smoking.”

Other observers, including the World Health Organization, argue for tougher rules to restrict vaping products’ attractiveness. The UK parliament is considering restrictions on packaging, marketing, display and minimum pricing of vapes, said Nicholas Hopkinson, a professor of respiratory medicine at Imperial College London.

The disposables ban would not reduce usage if companies continued to market other “cheap, colourful and appealing devices that are notionally reusable but in practice treated as disposable”, Hopkinson said.

The disposables bans will be closely watched to gauge consumer behaviour.

Jasmine Khouja, a senior research associate in the tobacco and alcohol research group at the UK’s Bristol university, pointed to a range of potential outcomes in a survey of young UK adults on the upcoming ban, from switching to reusable or other nicotine products to quitting vapes altogether.

“Some even said they would switch to smoking because they thought it might help them quit nicotine, as it’s not as nice as vaping,” she added.

Will weight-loss drugs lead to upheaval in the sugar market?

While many traders have brushed off concerns, the potential impact is clear

The health risks of too much sugar have been made clear, but the billion-dollar global market to supply it is thriving. Sales of sweet treats remain strong, and waistlines keep expanding. Could weight-loss drugs now succeed where governments, scientists and doctors have failed: crushing demand for sugar? 

So-called glucagon-like peptide-1 receptor agonists (GLP-1s) contained in such drugs as Wegovy, Mounjaro and Ozempic curb users’ appetites and are being hailed as game changers for tackling obesity and potentially a range of other conditions, from diabetes to addiction. They could also lead to an upheaval in sugar markets.

Fears that Americans on GLP-1s will stop buying treats have already spooked businesses and investors. Mondelez and PepsiCo stocks took a hit after Walmart chief executive John Furner reported that customers on these drugs were buying fewer groceries. Hershey has also acknowledged experiencing a “mild impact” on sales, attributed to the growing use of GLP-1 medications.

Sugar traders, for now, are brushing off concerns about weight-loss drugs’ potential to dent demand. Perhaps they’re battle-hardened — decades of sugar-is-bad campaigns have not made a dent in global consumption, which has quadrupled in the past 60 years, according to Professor Paul Behrens of the British Academy. Sweet treats still fly off the shelves in most markets, and until recently sugar prices have been riding high on weather woes and rising production costs.

There are other reasons for traders’ nonchalance. So far, Ozempic and other drugs are pricey and only available to a small segment of wealthy consumers in developed countries. Even if these appetite-suppressing drugs do start reducing demand, the thinking goes, it’ll be a slow burn, giving markets and sugar producers plenty of time to adjust.

But could this confidence be misplaced? Weight-loss drugs do work in curbing appetites. That makes them far more likely to reshape sugar consumption — and with it, the sugar market.

Stephen Geldart of London-based sugar merchant Czarnikow believes the GLP-1 drugs are already on track to disrupt demand, starting with wealthy nations. “I think we’re just in the early stages of it, partly because these drugs are still quite expensive,” he told me. “So they’re not widespread — yet.” 

That “yet” looms large. In the UK, for example, the government plans to roll out Mounjaro on the NHS. Prices are likely to drop elsewhere too, especially as pharmaceutical companies race to sell compounded versions of drugs, circumventing patents.

If prices fall and access broadens, the ripple effects could reach middle-income and even developing markets. The obesity epidemic and slew of health conditions that come with it are not limited to rich western nations. In India, the world’s biggest sugar consumer, rates of diabetes and obesity are soaring. Indians consume a staggering 29mn tonnes of sugar annually — 15 per cent of global demand. Even a modest uptake of GLP-1s there could shake the market in ways that traders might find hard to ignore.

Geldart highlights another phenomenon that should give sugar market participants pause: heavy buyers of products with high sugar levels such as ice cream, cookies and chocolate account for a disproportionate amount of sales. If such users cut back, there could be a big impact.

Tracking sugar consumption is notoriously tricky, though. There’s a real risk, therefore, that these trends could develop under the radar before the industry wakes up to what’s happening.

However, a quiet drop in demand could be magnified by rising supply. High prices over the past year have spurred a wave of sugarcane investments, particularly in Brazil. If production ramps up while consumption falters, the market could face a supply glut that drags prices lower and forces a painful adjustment. Already, prices have fallen back recently on reduced production worries. Raw sugar futures in New York have dropped from their November 2023 peak of 28 cents a pound, plunging this month to less than 19.50 cents.

Sugar’s adaptability — it can be repurposed into ethanol for fuel or bioplastics — might provide a cushion, especially as demand for low-carbon fuels and renewable materials grows. Yet this transition won’t happen overnight.

Until now, the sugar market has not attracted much speculative interest. But how long before short sellers seize an opportunity to get a slice of the Ozempic pie? 

Tech group emerges as crucial backer of start-ups seeking to gain from the AI revolution its chips are powering

Nvidia invested $1bn in artificial intelligence companies in 2024, as it emerged as a crucial backer of start-ups trying to gain from the AI revolution the big tech group’s chips are powering.

The semiconductor giant, which surpassed a $3tn market capitalisation in June on the back of huge demand for its high-performing graphics processing units (GPUs), has pumped ever greater sums into some of its own customers in the burgeoning sector.

According to corporate filings and Dealroom research, Nvidia spent a total of $1bn across 50 start-up funding rounds and several corporate deals in 2024, compared with 2023, which saw 39 start-up rounds and $872mn in spending.

The vast majority of deals were with “core AI” companies with high computing infrastructure demands, and so in some cases also buyers of its own chips.

Tech companies have spent tens of billions of dollars on Nvidia’s chips over the past year since the debut of ChatGPT two years ago kick-started an unprecedented surge of investment in AI.

Nvidia’s uptick in deals comes after it amassed a $9bn war chest of cash with its GPUs becoming one of the world’s hottest commodities.

The company’s shares rose more than 170 per cent in 2024, as it and other tech giants helped power the S&P 500 index to its best two-year run this century.

Nvidia’s $1bn worth of investments in “non-affiliated entities” in the first nine months last year includes both its venture and corporate investment arms. According to company filings, that sum was 15 per cent more than in 2023 and more than 10 times as much as it invested in 2022.

Some of Nvidia’s largest customers, such as Microsoft, Amazon and Google, are actively working to reduce their reliance on its GPUs by developing their own custom chips. Such a development could make smaller AI companies a more important generator of revenues for Nvidia in the future.

“Right now Nvidia wants there to be more competition and it makes sense for them to have these new players in the mix,” said a fund manager with a stake in a number of companies it had invested in.

In 2024, Nvidia struck more deals than Microsoft and Amazon, although Google remains far more active, according to Dealroom.

Such prolific dealmaking has raised concerns about Nvidia’s grip over the AI industry, at a time when it is facing heightened antitrust scrutiny in the US, Europe and China.

Bill Kovacic, former chair of the US Federal Trade Commission, said competition watchdogs were “keen” to investigate a “dominant enterprise making these big investments” to see if buying company stakes was aimed at “achieving exclusivity”, although he said investments in a customer base could prove beneficial.

Nvidia strongly rejects the idea that it connects funding with any requirement to use its technology. The company said it was “working to grow our ecosystem, support great companies and enhance our platform for everyone. We compete and win on merit, independent of any investments we make.”

It added: “Every company should be free to make independent technological choices that best suit their needs and strategies.”

The Silicon Valley group’s most recent start-up deal was a strategic investment in Elon Musk’s xAI, alongside rival chipmaker AMD.

Other significant 2024 investments included its participation in funding rounds for OpenAI, Cohere, Mistral and Perplexity, some of the most prominent AI model providers.

Nvidia also has a start-up incubator, Inception, which separately has helped the early evolution of thousands of fledgling companies. The Inception programme offers start-ups “preferred pricing” on hardware, as well as cloud credits from Nvidia’s partners.

There has been an uptick in Nvidia’s acquisitions, including a takeover of Run:ai, an Israeli AI workload management platform. The deal closed this week after coming under scrutiny from the EU’s antitrust regulator, which ultimately cleared the transaction. The US Department of Justice was also looking at the deal, according to Politico.

Nvidia also bought AI software groups Nebulon, OctoAI, Brev.dev, Shoreline.io and Deci. Collectively it has made more acquisitions in 2024 than the previous four years combined, according to Dealroom.

The company is investing widely, pouring millions of dollars into AI groups involved in medical technology, search engines, gaming, drones, chips, traffic management, logistics, data storage and generation, natural language processing and humanoid robots.

Its portfolio includes a number of start-ups whose valuations have soared to billions of dollars. CoreWeave, an AI cloud computing service provider and significant purchaser of Nvidia chips, is preparing to float early this year at a valuation as high as $35bn — increasing from about $7bn a year ago.

Nvidia invested $100mn in CoreWeave in early 2023, and participated in a $1bn equity fundraising round by the company in May.

Another group, Applied Digital, was facing a plunging share price in 2024, with revenue misses and considerable debt obligations, before a group of investors led by Nvidia provided $160mn of equity capital in September, prompting a 65 per cent surge in its share price.

“Nvidia is using their massive market cap and huge cash flow to keep purchasers alive,” said Nate Koppikar, a short seller at Orso Partners. “If Applied Digital had died, that’s [a large volume] of sales that would have died with it.”

Elon Musk had an eventful 2024, what with his hard push into right-wing politics and leveraging his X social media platform and considerable influence to help Donald Trump win the presidential election.

Putting politics aside for a moment, it’s worth focusing on another Musk pastime: making promises for Tesla.

The Tesla CEO’s promises — and repeated missed deadlines — have become a primary thread in the Tesla storyline. And they’ve always had an eye-popping quality that captured the imagination of investors and helped drive up the value of Tesla stock, giving the company a valuation of $1.3 trillion. There was his claim in 2015 that Tesla vehicles would be self-driving in two years, that cross-country driverless road-trips would be enabled by the end of 2017, and that Tesla owners would be able to earn money via a massive driverless ride-hailing network in 2020.

And while Tesla has cemented its place in history books thanks to building and selling millions of electric vehicles, none of the above promises (nor many others) have come to pass.

But that hasn’t stopped Musk from making more promises in 2024. Here’s an accounting of those promises and when he expects them to become reality.

The $25,000 EV

In 2024 alone, Musk pledged to unveil a $25,000 EV, then scrapped it in April to prioritize a robotaxi prototype — a decision that led to mass layoffs as Tesla pursued its “next phase of growth.”

Musk has flip-flopped on whether the affordable EV would ever come to market. But during Tesla’s third-quarter earnings call, Musk said the idea of building a $25,000 car with a steering wheel and pedals is “pointless” and “silly.” He said the only car he’d sell at that price point would be the Cybercab.

During the call, an analyst asked if Tesla would make a lower-cost EV that’s not the Cybercab, and Musk replied that all of the company’s cars moving forward would be autonomous. He also said that of the 7 million vehicles Tesla has built to date, the “vast majority” are “capable of autonomy,” and that Tesla is “currently making on the order of 35,000 autonomous vehicles a week.” Musk is clearly using a loose definition of autonomy here, because Tesla still does not produce vehicles that are safe to use without a human behind the wheel.

(Flashback: Musk originally promised in 2016, in a since-deleted post on Tesla’s website, that “All Tesla Cars Being Produced Now Have Full Self-Driving Hardware,” and that only a software update would be needed to turn regular old Teslas into self-driving cars. That didn’t happen, and Tesla has had to upgrade cars with older hardware.)

It’s also worth noting that during that earnings call, Musk said that he expects vehicle growth to reach 20% to 30% in 2025 due to “lower cost vehicles” and the “advent of autonomy.”

Start production on the Cybercab by 2025 or 2026

Tesla unveiled 20 Cybercab prototypes at a flashy Hollywood event in October, and Musk took the opportunity to share some plans for the vehicles, as well as Tesla’s so-called Full Self-Driving (FSD) software. FSD is Tesla’s advanced driver assistance system that can perform many automated driving tasks, but still requires a human to remain attentive behind the wheel and take over if needed.

Musk told customers they would be able to one day buy a robotaxi — a two-door, two-seat vehicle with no steering wheel or pedals — for less than $30,000. He also said that the average operating cost of the Cybercab will reduce over time to only $0.20 per mile. And he said Tesla would start production on the purpose-built AVs in 2025 or 2026.

(Flashback: Musk said in 2022 that Tesla would mass-produce robotaxis by 2024. Before that, in 2019, he said that Tesla would have a million robotaxis on roads by 2020. Musk has promised that Tesla would solve full self-driving “next year” since at least 2016.)

A few weeks later, during Tesla’s third-quarter earnings call, Musk said Tesla would reach “volume production in ’26,” and that the company was eventually “aiming for at least 2 million units a year of Cybercab.”

Current federal regulations that require vehicles to be built with certain safety standards, like manual human controls, could be roadblocks for Tesla to mass produce its Cybercabs. The National Highway Traffic Safety Administration recently proposed new rules that would fast-track exemptions for such vehicles, but they would require companies to share more data with the agency, like crash reporting. Today, automakers are required to report crashes when ADAS or autonomous driving tech is engaged.

Musk has railed against the rule before, and President-elect Donald Trump’s transition team is reportedly exploring scrapping it.

Robovan in development

During Tesla’s robotaxi reveal, the company also showed investors a Robovan prototype. At the time Musk didn’t share any concrete plans for the vehicle, but in November he posted on X that the Robovan is in development, along with “some other things.”

(Flashback: In 2016, Musk said Tesla would start building a minibus using the Model X chassis in two to three years.)

“Unsupervised FSD” and autonomous ride-hail in 2025

At the event, Musk also promised that Model 3 and Model Y owners would be able to use an “unsupervised” version of FSD in California and Texas in 2025.

It wasn’t, and still isn’t, clear what Musk meant by “unsupervised” FSD. Today, Tesla’s FSD is still not fully autonomous, and to dispel any confusion, Tesla this year began referring to the software as “supervised FSD.” Removing the supervision could mean that Tesla plans to remove the driver, or it could mean that Tesla plans to offer a Level 3 autonomous system that allows drivers to go hands-off, eyes-off for a portion of their ride.

During Tesla’s third-quarter earnings call, Musk took the promise of unsupervised FSD a step further. He said he hopes to launch a service that will let people hail self-driving Teslas in California and Texas in 2025. He also claimed Tesla had begun testing the service in the Bay Area with employees.

(Context: There are several tiers of permits required to test and deploy autonomous vehicles in California. Tesla has held a permit to test AVs with a safety driver in the front seat since 2015, but the Department of Motor Vehicles told TechCrunch in October that Tesla last reported using that permit in 2019.)

It’s not clear if Tesla plans to launch this ride-hail service with its robotaxis or with existing Tesla Model 3 and Model Y owners. Tesla’s first-quarter earnings presentation included a mock-up of a future Tesla ride-hail app, and the company has for years teased the idea of a ride-hailing network using Teslas that had been updated to drive fully autonomously. The idea is similar to Uber, only Tesla owners would add their properly equipped self-driving vehicles to the automaker’s ride-hailing app to make extra cash when the cars are not in use. Tesla would take 25% to 30% of the revenue.

Finally, during Tesla’s first-quarter earnings call, Musk said Tesla was in talks with a “major automaker” to license FSD, but so far has not announced any such deals.

“Over a thousand Optimus robots working at Tesla” in 2025

Musk has made some out-there promises about Optimus, Tesla’s humanoid robot. During Tesla’s annual shareholder meeting, Musk promised that Tesla would move into “limited production” of the Optimus next year, with “over 1,000, or a few thousand, Optimus robots working at Tesla” in 2025. He also said he expects Optimus to be on sale by 2026. In a post on X, the billionaire executive added that he expects to have more Optimus bots available for use by other companies in 2026.

He did not specify if the bots would be working fully autonomously, or if they would be remotely controlled by humans, as they were during Tesla’s robotaxi reveal event in October.

(Context: While humanoid robots are improving to the point of being able to autonomously complete specific tasks, many experts say generalized robotics is still years away due to a lack of training data.)

Aside from manufacturing promises, Musk also predicted that Optimus could one day lift Tesla’s market cap to $25 trillion. That’s roughly seven times the current market caps of Apple and Nvidia. At the end of December, Tesla’s market cap was around $1.42 trillion, which is almost a 160% increase from the $550 billion market cap before Trump won the presidential election.