TheChek News Flow

The short answer: a lot. And at least the top executives say it is paying off.

It’s one of the fundamental questions about generative AI: Can companies put it to good use? One answer comes from a new survey by the Wharton School of the University of Pennsylvania, which found that business leaders went all in on generative AI in the past year, and they say it is paying off.

If there was any doubt about corporate leaders’ comfort with AI, the report indicates that doubt is no longer a factor; 82% of the executives say they personally use generative AI once a week, and 46% now use it daily. In 2023, the first year of the survey, fewer than 37% said they used it once a week.

The annual survey, conducted this summer by Wharton and marketing consulting firm GBK Collective, covers 800 business leaders at companies with more than 1,000 employees and $50 million in revenue.

The survey also finds 72% of executives say their companies formally measure their return on investment from generative AI, and roughly 3 in 4 say they see positive returns.

The report also provides an optimistic signal that generative AI may not wipe out the entry-level job. Senior leaders expect generative AI to have its biggest impact on interns and entry-level workers, but substantially more senior leaders expect AI to create more hiring at these levels. For instance, 17% say they expect to hire fewer interns, but 49% expect to hire more interns; 18% expect fewer hires for entry-level or junior positions, while 40% expect more.

That said, senior leaders are at times more enthusiastic about generative AI than the middle managers below them. Almost half of the top leaders—those with titles of “vice president” or higher—say they are seeing significantly positive ROI, while only 27% of middle managers do; middle managers are twice as likely to say it is too early to tell. Senior leaders are also twice as likely as middle managers to believe their own organizations are adopting generative AI much faster than other organizations.

To dive more deeply into what the survey found—and what it means—we talked to Stefano Puntoni, a Wharton professor of marketing who co-authored the report along with Jeremy Korst, partner at GBK Collective, and Prasanna Tambe, a professor of operations, information and decisions at Wharton. Here are edited excerpts of the conversation:

WSJ: Why do senior leaders seem more enthusiastic about implementing AI than middle managers?

STEFANO PUNTONI: Leaders probably have the bigger overview of what is going on in the organization. On the other hand, people who are a little bit below that level, they might feel a bit squished between bottom-up and top-down pressure. They might feel more threatened by the technology than the VP level. The VP, on the other hand, may be overoptimistic, because they are the ones making these decisions about big deployment, and therefore they might try very hard to see the positive.

WSJ: Could it also be that sometimes the senior leaders are more about the vision but it’s up to the lower-level managers to carry out the mission?

PUNTONI: It could be that the people who are at the top tend to be thinking about longer term. They may report a greater level of optimism because they’re projecting two, three, five years out. People who are a bit further down, they have more of a concrete mindset about the thing they need to do next week.

There is also probably the pressure of people who’ve been tasked with implementing a technology feeling like this is actually more creating friction for me right now than giving me benefits. This might make it better later on, but right now, it’s just more work.

WSJ: So much of the conversation around AI revolves around whether it will replace jobs or enhance existing jobs. What was the prevailing sentiment on that in your study?

PUNTONI: We tapped into that kind of question in a bunch of different sections of the report. One question was, to what extent do you believe that this technology is going to complement employee skills or replace employee skills? What we saw in 2023, people reported very high levels of both beliefs about replacement and enhancement. By 2024, we saw that shifting to significantly more people endorsed beliefs about human enhancement. This year we see no change: About 90% believe AI will complement employee skills.

WSJ: What about the concerns that AI will eliminate the need for lots of entry-level jobs?

PUNTONI: Senior people tell us that Gen AI will have more of a net-positive impact on junior positions than on senior people [helping them do their jobs better or create new jobs]. So it seems, looking at our data, like the picture for entry-level positions might not be that bad. That surprised me.

WSJ: Why were you surprised by that?

PUNTONI: The conversations in the media about entry-level positions have been very negative. My son is in high school, and last week, he had a bit of existential anxiety about his job prospects. He was saying, why should I study for the test? I’m never going to get a job anyway. I was trying to explain to him that I don’t think that there will be no opportunities for junior people. It’s just that these opportunities are going to change. So what you need to be able to show an organization is that you can work with this technology. I think he feels better now.

WSJ: How much of leaders’ enthusiasm for, and investments in, AI is looking over your shoulder at competitors and FOMO?

PUNTONI: There is some of that. There is a sense that for some companies, they’re adopting AI faster with a lack of vision or strategy.

One issue that I have with a lot of the conversations in business about generative AI is that we’ve been focusing so much on productivity. And when you talk about productivity, it’s almost like a code word for cost-cutting.

Companies ought to be as efficient as possible. But I think in the long run, even in the medium run, actually, this technology is only going to give results where we are imagining new things to do. I’m a marketing professor, and I get excited about productivity, but I get more excited about what kind of new customer experiences we can create. What kind of new products and services can we generate? What kind of better, stronger customer relationships can we forge using this technology? To me, that is the kind of thing that, in the end, is going to lead to sustained growth, rather than simply throwing money at AI because everyone else is doing it and they don’t want to get left behind.

WSJ: What did you learn about general usage?

PUNTONI: We switched to tracking daily use from weekly use because basically so many people are using it. A large amount of people are building comfort, familiarity and certainly a degree of expertise. This technology is permeating a workplace to a very high degree.

The top-rated use cases tend to be general productivity tools, like data analysis, meeting summation and writing. That’s the bread and butter of knowledge work. The fact that those are the top three tells you this is not a technology that is going to have a narrow impact on a particular function of a particular industry, but it’s going to have a wide impact across the economy.

The company satisfied regulators, accountants and investors in its quest to raise billions but keep its financials unscathed.

It took a year and some financial acrobatics, but Meta Platforms and a small army of bankers and lawyers were able to secure funding for a $27 billion data center project without doing any immediate damage to the company’s finances. A private letter from the Securities and Exchange Commission helped seal the deal.

How Meta and Morgan Stanley pulled off the fundraising, and whether other companies racing for dominance in artificial intelligence will be able to match their feat, will potentially shape one of the biggest investment booms in history. Meta’s deal is one of several expected in the coming months to raise tens of billions of dollars for giant data center projects.

Investors this month snapped up the bonds used to fund Meta’s massive Louisiana data center project, Hyperion, which is a big part of the company’s AI ambitions. Bond giant Pimco and BlackRock, the world’s largest asset manager, were among the buyers of the $27.3 billion debt offering, issued by an entity called Beignet Investor. Investment firm Blue Owl is putting up $2.5 billion in equity.

To make the deal work, Meta placed the Hyperion project into a special purpose vehicle. Blue Owl will own an 80% stake in the SPV and control its board, allowing Meta to keep the debt off its balance sheet. Meta further limited the financial impact by breaking up its leases into four-year chunks so rating agencies wouldn’t consider them debt.

To get the deal done the way it wanted, Meta walked a tightrope through accounting rules, bond rating agencies and financial regulators. Its goal was to limit its long-term liabilities, a key measure of a company’s financial health, said analysts and people with knowledge of the process. Meta obtained a letter from the SEC effectively approving its proposed accounting treatment for the project, the people said.

A Meta spokesperson declined to comment. The company previously said the deal “was designed to support the speed and flexibility required for Meta’s data center projects and long-term AI ambitions.” An SEC spokesperson said the agency was unable to respond to many press inquiries during the government shutdown.

Meta also reached out to credit rating agencies Moody’s and Standard and Poor’s to ensure the deal wouldn’t harm its strong investment grade rating, one of the people said. Both agencies said last week the transaction shouldn’t have any immediate impact on Meta’s credit rating, though both also highlighted risks with the deal.

The result is that Meta will get to develop its biggest-ever data center project without taking on any debt, while also freeing up cash as CEO Mark Zuckerberg embarks on an expensive bid to become a major player in AI. Meta has dramatically ramped up its capital expenditures on data centers, spending as much as $72 billion this year and about $100 billion in 2026, according to analysts. It has historically funded the spending with cash but is now turning to outside investors while trying to maintain its strong balance sheet.

To give Meta further flexibility, the deal will give it the option to walk away from the project sooner than it would under similar data center leasing agreements, providing an exit route if it no longer needs the facilities.

“As long as the capital markets are willing to do this, why not?” said Gil Luria, head of technology research at D.A. Davidson.

Bankers were closely watching the Meta deal as a test of whether traditional bond investors would embrace an unconventional structure for funding massive data centers. Elon Musk’s xAI is using a similar arrangement to fund the chips going into its Colossus 2 data center, through a deal arranged by investment firm Valor Equity Partners.

Meta shareholders are expecting to learn more about the exact financial impact of the Hyperion deal when the company reports earnings Wednesday. It cut about 600 employees from its AI organization last week, a further sign it’s trying to balance the costs of competing to build powerful AI.

Zuckerberg in July said Hyperion could eventually use up to 5 gigawatts of power after its completion four years from now, making it one of the largest data center projects on record. Meta expects the project to cost $27 billion to construct.

Meta’s partner in the project, Blue Owl, brought big debt investors on board by offering a rich yield for a relatively low-risk bond.

The bonds for the Hyperion data center priced with a coupon of almost 6.6%, roughly a percentage point higher than Meta’s outstanding corporate bonds and in line with the average junk bond. That’s a higher yield than investors would expect given that S&P rated the Hyperion bonds A+, safely within the investment-grade spectrum.

That made the bonds attractive to risk-averse investors such as insurance companies that usually stick to investment-grade bonds. Investors bid up the price of the bonds following the offering, pushing down their yield.

Pimco purchased about $18 billion of the bonds, and BlackRock also made a multibillion-dollar purchase, said people familiar with the offering. Another large chunk went to funds managing money for insurers, said one person with knowledge of the deal.

The bonds come due in 2049, a relatively long maturity for corporate debt. That made them even more attractive to buyers such as insurance companies and pension funds that need to fund long-term liabilities.

Before it approached investors, Meta spent months working with Morgan Stanley bankers to devise a financial structure that would allow it to quickly build the massive campus without harming its financial strength. Meta hired Latham & Watkins to advise on issues ranging from corporate finance to energy regulation, tapping 14 partners in total, according to the law firm.

Though SPVs aren’t a new invention, Meta is the first tech giant to use the financial structure to fund AI data centers for its own use. Companies including Google and Microsoft have turned to other methods for offloading the costs of the AI data center buildout, such as working with upstart cloud and data center providers and having them raise funding.

Meta also appeared to find a way around accounting rules that would normally treat its lease as a long-term liability, similar to debt. Instead of signing a 20-year lease for the data centers, Meta agreed to a series of renewable four-year leases at the 11 individual buildings. That could allow the company to treat them as operating leases instead of financial leases, which must be reported as long-term liabilities.

The company is promising a payout to investors, known as a residual value guarantee, if it decides to walk away from the campus before 2049. That potential payment won’t immediately count against Meta’s debt levels, according to S&P.

“At the end of the day, they were willing to pay up for the flexibility to walk away from this,” said Naveen Sarma, managing director at S&P.

State-owned lender absorbs Ayandeh Bank, owned by one of country’s richest families

Iran has mounted one of the largest interventions in the country’s financial system in decades after taking over a major privately owned lender that had been pushed to the brink of collapse by years of spiralling debt and alleged mismanagement.

The central bank has reassured the 7.6mn depositors with Ayandeh Bank that their funds remain safe, seeking to calm a financial system already under severe strain. Ayandeh has been folded into state-owned Bank Melli Iran (BMI), whose white banners now cover its branches.

Founded in 2012 by one of Iran’s richest families, Ayandeh offered the highest interest rates on the market. The rates attracted vast deposits that it could service only by borrowing heavily from the central bank, which printed money to prevent the collapse of an institution deemed too big to fail, analysts say.

The central bank’s approach stoked Iran’s chronic inflation and ultimately failed to keep Ayandeh afloat, analysts say. It took over Ayandeh using a mechanism designed to manage a bank failure without disrupting the wider financial system.

Ayandeh owed the central bank an estimated 5,000tn rials ($4.6bn), while customers have missed payments on more than 97 per cent of Ayandeh’s total loans, according to the semi-official Tasnim News Agency.

The central bank, BMI and Ayandeh did not immediately respond to requests for comment.

BMI is taking on Ayandeh’s liquid assets and liabilities. Ayandeh’s deposits were worth 2,500tn rials ($2.3bn) and its registered capital was only 16tn rials ($14.8mn). BMI has pledged to keep employees in their jobs and pay depositors’ existing interest rates until their contracts mature.

“The central bank’s intervention is a good move in principle,” said a Tehran-based banking analyst, who asked to remain anonymous to discuss the sensitive issue. But the analyst said balance sheet issues and mismanagement were a major concern elsewhere in the sector.

Ayandeh and its founders, the influential Ansari family, have long been accused by politicians and analysts of channelling depositors’ money into speculative ventures and affiliated companies. Between 40 and 44 per cent of Ayandeh’s loans went to the bank’s own subsidiaries, according to Tasnim.

Ayandeh founder Ali Ansari disputed the allegations on Friday, saying “the truth will eventually come out” but that he did not want to raise any “hue and cry” at present. He insisted that the bank had “shone” but was impeded by outside actors, saying Ayandeh was “a symbol of smart efforts manifested in huge projects”.

The bank’s largest project was the construction of Iran Mall in Tehran, one of the Middle East’s biggest shopping complexes.

Its non-performing loans, according to the central bank, amount to at least 1,400tn rials ($1.3bn). Analysts say many of these loans were issued either without collateral or against overvalued assets.

Iran’s financial sector has long had low capital requirements and light regulation, which analysts say have led to banks holding bad debts and balance sheet mismatches. Iran’s isolation from the global banking network under US sanctions has compounded the problem.

Both hardline and reformist politicians had been lobbying for Ayandeh’s closure for years, arguing that the central bank’s support for the bank was fanning inflation and weakening the rial. 

In a bid to stay alive, Ayandeh’s board changed its management team on October 18 and offered the central bank a 2,000tn rial ($1.8bn) capital increase.

The next day, President Masoud Pezeshkian met Chief Justice Gholam-Hossein Mohseni-Ejei and parliamentary Speaker Mohammad Bagher Ghalibaf, who rejected proposals to recapitalise or restructure the bank, according to Iranian bankers.

Mohseni-Ejei then publicly warned last Tuesday that the judiciary would intervene if the central bank failed to act. On Thursday, central bank governor Mohammad-Reza Farzin announced the takeover.

“The announcement came as a shock,” said one of the newly-hired managers at Ayandeh. “We all expected our capital increase offer to be accepted by the central bank.”

Ayandeh’s non-liquid assets, including large real estate holdings, have been handed to a government body that insures bank deposits, to be sold over the next two to three years. The proceeds will be used to settle Ayandeh’s debts with the central bank and other creditors.

The central bank has previously intervened in similar cases. State-owned Bank Sepah, which was compelled several years ago to absorb three failing banks, is still struggling with their debt, analysts say. 

Authorities insist this time will be different, as the economy contracts and a public wearied by eight years of average inflation at 40 per cent bristles at paying another private company’s debts.

Farshad Mohammadpour, the central bank’s deputy for supervision, explained to local media that “our red line is that not even a single rial of Bank Ayandeh’s financial imbalance will be transferred to BMI”. He acknowledged it would be a challenge to repay Ayandeh’s liabilities: “If selling assets like Iran Mall were easy, the bank would not have reached this point.”

The bank’s depositors are nervous. “We put all our savings in Ayandeh because of its higher interest rates,” said Shirin, a 55-year-old in Tehran who earned 28 to 30 per cent annually, compared with the 23 to 26 per cent offered by other banks. “They tell us our money is safe, but we’ll see when our contract ends in five months.”

Substrate plans to use particle accelerators to lower cost of chip manufacturing

Silicon Valley investors, including Peter Thiel’s Founders Fund, have invested more than $100mn in a secretive US start-up with an ambitious plan to challenge the dominance of Taiwan Semiconductor Manufacturing Company and ASML in cutting-edge chipmaking.

San Francisco-based Substrate, founded in 2022 by James Proud and his younger brother Oliver, plans to use particle accelerators to make chips much more cheaply than today’s state-of-the-art equipment. The pair has no previous experience in semiconductor manufacturing.

The start-up has attracted funding from big-name investors including Founders Fund, General Catalyst and Valor Equity Partners, despite the huge logistical and financial challenges Substrate faces to succeed in one of the most technically complex industries in the world.

Its fundraising of more than $100mn, completed last year but not previously disclosed, valued the company at more than $1bn, Substrate said.

Proud told the Financial Times he wanted to boost homegrown US chip manufacturing capabilities at a time when American tech groups, including Apple and Nvidia, had become overwhelmingly reliant on foreign companies for cutting-edge semiconductor production.

Proud said he planned to build alternatives to Netherlands-based ASML’s advanced lithography machines, needed to etch minuscule transistors on to silicon wafers, and TSMC’s giant fabrication plants — ambitions that he said would cost “many billions” of dollars to fulfil.

He added that his previous lack of semiconductor experience was “definitely a positive, because if I had, I would have thought this is impossible”.

Semiconductor foundries cost tens of billions of dollars to build, with ASML’s most advanced machines costing hundreds of millions each. Proud claims his approach can reduce the cost of producing a leading-edge wafer from a projected $100,000 to “closer to $10,000” by the end of this decade. He hopes to have production up and running by 2028.

Proud, who was born in London but is now an American citizen, said he was driven by a desire to help the US beat China in the race for global technological dominance.

“The company itself is very ideological,” Proud said. “The US needs to have advanced semiconductor production and we need to do it at high volume.”

Proud said “foreign monopolies” holding “the two main choke points” on chip production was a “glaringly scary dependence”. Substrate is aiming to produce chips “with a cost structure that enables [the US] to compete against China”, he added.

Substrate’s emergence from three years developing its technology in secret comes as the US government is working to secure the future of struggling chipmaker Intel, including taking a stake in the Silicon Valley group, and China threatening to throttle exports of the rare earth minerals that are vital to the global tech supply chain.

“These are points in history that are really important to get right,” said Trae Stephens, a Founders Fund partner who led the firm’s investment in Substrate. “There’s a lot of pressure right now for the [US] government to figure out how to ensure that we have a semiconductor supply chain that is reliable and resilient . . . That is the moment that James is capturing right now.”

Substrate’s technology involves using particle accelerators as the light source for an X-ray based lithography system, using even shorter wavelengths of light than ASML’s latest “extreme ultraviolet” machines to imprint circuits on to silicon.

The system, which Substrate has demonstrated at US National Laboratories, has created complex patterns on a wafer at a resolution Proud claims is comparable to ASML’s latest “High NA” EUV tools. That would make it suitable for producing chips at the 2nm process node, he said, which is the semiconductor industry’s current state-of-the-art technology. Substrate also plans to use existing chipmaking tools.

Proud said building a “much more vertically integrated” foundry would require raising “many billions — at some point tens, maybe even hundreds of billions of dollars” — as the business scales to multiple facilities.

“The ambition to go and build the fabs themselves is huge and will require a lot of capital,” Stephens said, requiring a mix of strategic investment, debt and government funding. “This will be an all-hands-on-deck moment.”

Proud moved to Silicon Valley aged 19 after becoming an early recipient of Thiel’s $100,000 fellowship, which pays youngsters to become entrepreneurs instead of going to college. The best known of Proud’s previous ventures, health tech start-up Hello, raised $40mn in 2015 but shut down two years later.

Stephens said Proud had assembled an “incredibly impressive” team at Substrate, including scientists and engineers from US National Laboratories, TSMC, Applied Materials, AMD, Google and Qualcomm.

“[Substrate] is obviously a much bigger bite of the apple” than other chip start-ups, Stephens said. “But if they win, they win really, really big.”

Registrations for Tesla fell 10.5% in Europe

BYD logged a nearly fivefold surge in European sales last month, a sign of how customers in the region are increasingly warming to the Chinese auto giant’s lineup of electric and hybrid vehicles.

New-car registrations for BYD models, a reflection of sales, almost quintupled on year to 24,963 vehicles in Europe last month, according to the European Automobile Manufacturers’ Association, an industry body also known as ACEA. The figure includes data from the European Union as well as the U.K., Iceland, Liechtenstein, Norway and Switzerland. In the EU alone, BYD registrations more than tripled to 13,221 vehicles.

The jump in sales, the third since July after ACEA began including China’s largest automaker in its monthly data, underscores the appeal of BYD’s lineup in Europe, even though absolute sales remain far below those of established domestic carmakers like Germany’s Volkswagen or Stellantis, the owner of the Jeep and Dodge brands. Both companies respectively sold 317,432 and 165,457 vehicles in Europe last month, according to ACEA.

BYD has been expanding aggressively in the continent, leveraging its relatively cheap and varied lineups to gain market share from rivals. In late 2023, the Chinese company announced plans to build a new passenger car factory in Hungary, with production expected to begin in the coming months.

Meanwhile, registrations for Elon Musk’s Tesla fell 10.5% in Europe and nearly 19% in the EU alone last month, according to ACEA data, adding to a streak of disappointing monthly sales for the company this year.

Aside from Chinese and European competition, Tesla has been dealing with the fallout from Musk’s involvement with the Trump administration that came to an end a few months ago. The group has been reviewing its lineup lately as it seeks to increase its appeal and recently released pared-down versions of its popular Model Y SUV and Model 3 sedan.

ACEA said the EV market in the EU had continued to improve in September. Sales of battery-electric vehicles increased 20% on year, with a 32% jump in Germany. Registrations of hybrid-electric cars increased 16%, while plug-in-hybrid models logged 65% growth.

ACEA said EU passenger-car registrations increased 10% in September to 888,672 vehicles, with sales up 13% in Germany, 1% in France and 4.2% in Italy.