The Street

Most chip suppliers spend years trying to win a single major AI customer. MediaTek just landed the biggest one and got a direct investment.

On Aug. 31, 2026, Nvidia (NVDA) said it would put $3.5 billion into convertible bonds issued by Taiwan’s MediaTek.

A convertible bond is debt that can later turn into shares. So Nvidia takes a lender’s position now, with the option to become a shareholder later.

It is Nvidia’s largest direct investment outside the United States.

Investors moved fast. MediaTek shares jumped about 10% to their daily trading limit in Taipei the next session, extending a rally of almost 200% this year.

Why Nvidia is paying to wire MediaTek into its systems

The heart of this deal is a technology called NVLink Fusion. It lets outside chipmakers design their own custom AI chips that still plug directly into Nvidia’s data center systems.

Here is the pressure behind the move. 

Big cloud companies like Amazon, Google, Microsoft, and OpenAI are building their own custom AI chips to rely less on Nvidia’s expensive processors, Yahoo Finance reported. MediaTek is one of the companies that can help them build those chips.

By handing MediaTek NVLink Fusion, Nvidia ensures that even custom chips built to avoid it still connect back into its own systems.

So Nvidia keeps earning from the wiring, the racks, and the software, even when the main chip is not its own.

Nvidia’s $3.5 billion investment pulls MediaTek deeper into its AI hardware roadmap.Bloomberg / Getty Images

What the deal changes for MediaTek stock investors

For years, MediaTek was known for chips inside mid-range smartphones, TVs, and Wi-Fi routers. This deal moves it into a much bigger arena.

MediaTek is the world’s largest smartphone chip supplier by market share and Qualcomm’s (QCOM) main rival, according to CNBC.

Now it steps directly into the custom AI chip market led by Broadcom (AVGO) and Marvell (MRVL). That matters because Nvidia’s backing gives MediaTek instant credibility with the exact customers it wants to win.

More AI Stocks:

Marvell’s $120 billion deal with Google has fine print

Marvell stock fell 10%, but Morgan Stanley still likes it

BMO starts Broadcom at Outperform with $455 target

On its July 31 earnings call, MediaTek said its data center chip business should bring in more than $2 billion in revenue this year.

It also raised its target to capture 15% to 20% of a custom AI chip market it expects to be worth $80 billion by 2027, Reuters reported.

Working with Nvidia lowers engineering costs and shortens the time customers need to get a chip to market, which makes those targets easier to chase.

How the money is structured, and why that helps MediaTek

By using bonds instead of buying stock outright, Nvidia avoids taking a large equity stake or a board seat right away.

That gives MediaTek flexible funding for its balance sheet while Nvidia keeps the option to convert into shares later.

The wider bond issuance totaled about $3.9 billion and also drew Alphabet (GOOGL) as an investor, Reuters reported.

The deal also extends beyond data centers to personal AI computers and car software, CNBC noted.

The risks that could still trip up the stock

Investors should watch a few things.

Dilution. If MediaTek shares keep climbing and Nvidia converts its bonds into stock, current shareholders would own a smaller part of the company.

Thinner margins. Custom chip design usually earns lower profit margins than selling ready-made platforms, so more revenue may not mean the same jump in profit.

A Qualcomm response. MediaTek’s main mobile rival could cut prices or form its own alliances to protect its ground.

Geopolitics. As a Taiwanese firm tied closely to Western AI infrastructure, MediaTek remains exposed to cross-strait tension that can move the stock regardless of results.

Some analysts have raised a separate concern. 

Related: Jim Cramer reveals 6 AI stocks to watch in 2026

Bernstein Research analyst Stacy Rasgon, who has covered semiconductors for years, warned that Nvidia’s recent string of investments deepens worries about “circular” financing across the AI industry, TechTimes reported.

The idea is simple. Nvidia’s own money helps fund customers and partners whose spending then flows back to Nvidia.

Nvidia CEO Jensen Huang pushed back, telling Bloomberg TV the businesses operate independently.

What MediaTek investors should watch next

If you own MediaTek stock or want to buy it, these steps will help you stay realistic.

Treat it as a high-growth AI position, not a safe one. The stock has already run hard this year, so size any position to a risk level you are comfortable with.

Track the design wins. The number that matters most is whether major cloud providers officially sign MediaTek to build their custom chips. Listen for that on upcoming earnings calls.

Read the fine print on the bonds. Nvidia has not disclosed the price at which it can convert the bonds into shares. Watch for filings that clarify the price, since it will help set a floor and a ceiling for how big investors value the stock.

The takeaway for MediaTek investors is simple. Nvidia does not commit $3.5 billion without expecting something back.

This deal locks a fast-growing chip designer into Nvidia’s ecosystem before Broadcom or Marvell can offer a better alternative.

For MediaTek, that brings real validation and a higher bar to clear. The company now has to deliver design wins at the pace investors are pricing in.

Tuesday’s 10% jump shows the market is already betting on that outcome. The next several quarters will show whether the bet was right.

Related: Morgan Stanley delivers bold pre-earnings verdict on Broadcom

TheStreet aims to feature only the best products and services. If you buy something via one of our links, we may earn a commission.

Why we love this deal

Dressers are one of the easiest ways to add storage to your home. You can use them in a bedroom, the living room, or even a dining room. However, if you’re limited on space, it can be tricky to work with a more traditional dresser. But that doesn’t mean you have to give up on the useful piece of furniture just yet. Mini dressers provide ample storage, and they can fit into small spaces with ease.

The Homfa 3-Drawer Mini Dresser is a compact storage solution, and it’s on sale at Walmart for as low as $50. With a limited-time Flash deal, you can get up to 58% off the mini dresser that can upgrade storage in almost any room in your home.

Homfa 3-Drawer Mini Dresser, From $50 (was $120) at Walmart

Courtesy of Walmart

Shop at Walmart

Why do shoppers love it?

This three-drawer mini dresser is a fantastic way to add extra storage to a small space. Measuring 17.7 inches long by 15 inches deep by 21.9 inches high, it can easily fit into compact areas, whether it be an entryway, hallway, living room, or bedroom. At just under 22 inches high, it would pair well with low-profile furniture, like sofas and bed frames. The drawers provide ample space for everything from clothing to home essentials. With a handle-free design and a smooth gliding mechanism, each drawer opens and closes with ease.

Its minimalist design makes it easy to blend into a room, and it’s available in six colors. The green and blue colorways will get you the best deal at only $50 each, and a single dresser in black or white can cost up to $70. If you get a set of two, you can also choose gray or wood grain options. 

Related: Amazon has a heavy-duty 4-tier wire shelf that can hold up to 800 pounds on sale for only $29

Details to know

Dimensions: 17.7 inches long by 15 inches deep by 21.9 inches high.

Colors: Green, blue, black, gray, white, and wood grain, with prices and set sizes varying.

Material: Medium-density fiberboard.

There’s a reason why this dresser is a bestseller. Walmart shoppers say they’re happy with the piece of furniture, as it’s versatile and good quality. “I purchased this item as my mini vanity desk, and it fits perfectly in my small room. The manual/instructions were simple and easy to follow, and the screws were complete, with some extra pieces,” a reviewer said. 

“I eventually got four of these,” another customer said. They shared that they used two in a closet and the other two in front of a window to provide extra drawer space and storage in a kids’ room.

Shop more deals

Eunos 2-Drawer Mini Dresser with Charging Station, $31 (was $35) at Walmart

HomeYoobure End Table with Charging Station, $47 (was $80) at Walmart

Negylim 2-Drawer Farmhouse Mini Dresser, $50 (was $92) at Walmart

The Homfa 3-Drawer Mini Dresser is on sale for as low as $50, and it’s a small-space storage solution that can upgrade any room in your home.

Every chief executive officer gets one free sentence on the way in. It becomes the line people quote back three years later, usually with a scoreboard attached.

Most of them know this and say nothing memorable on purpose. Vagueness is cheap. Specificity is a liability you carry into every earnings call that follows.

Apple (AAPL) has been unusually good at that discipline. Under Tim Cook, who ran the company for 15 years and moved into the executive chairman role on Sept. 1, the corporate voice stayed flat by design.

Cook built a business on more than 2.5 billion active devices and quarterly revenue above $100 billion, and he almost never described a product before it existed. Restraint was the house style, and investors priced it in.

That is the setting for what happened Sept. 1, when Apple’s longtime hardware chief became its eighth chief executive in a 50-year history.

John Ternus broke the house style on his first morning.

Ternus told employees that “we have a huge launch next week that’s going to be phenomenal,” in a memo first reported by Bloomberg and published in full by 9to5Mac.

What Apple already told Wall Street about this quarter

The memo reads like standard first-day warmth until you put it next to Apple’s own numbers, which is exactly what I did before writing a word of this.

Apple posted June-quarter revenue of $109.4 billion, up 16% from a year earlier, with diluted earnings per share of $2.02 and gross margin of 50.1%, according to the company’s earnings release filed with the Securities and Exchange Commission. Every one of those was a June-quarter record.

Then management gave guidance, and the stock fell more than 6% in extended trading.

Apple projected September-quarter revenue growth of 9% to 11%, short of analyst expectations for 12%, according to CNBC. Gross margin was guided to 47% to 48%.

Related: Apple’s new CEO inherits a fortune and an AI question

Finance chief Kevan Parekh pointed to a 2.5 percentage point currency headwind and supply constraints that would increase from the June quarter. Cook described the component situation as “a 100-year flood” in memory pricing, according to the earnings call transcript published by Investing.com.

So Ternus inherited a company that had just told Wall Street the growth rate was coming down and the margin was coming down with it.

One word in Ternus’s memo carries real risk

“Phenomenal” is not a neutral word. It is a forecast — delivered by the one person at Apple whose forecasts now move the stock — about an event nine days away.

Here is what makes it interesting rather than merely bullish.

The Sept. 9 event sits almost entirely outside the quarter Apple guided down. Apple’s fiscal fourth quarter ends in late September, so new iPhones that reach buyers roughly 10 days after the keynote contribute perhaps a week of sales to it.

The launch Ternus called phenomenal will therefore be graded, in any financial sense, in the December quarter. He has set an expectation now and cannot post a number against it until late January.

That is a long time to hold a word.

What the Sept. 9 launch is walking into:

Apple guided September-quarter revenue growth of 9% to 11%, down from 16% in the June quarter, according to CNBC.

Gross margin is expected to land at 47% to 48%, down from 50.1%, according to Apple’s earnings release filed with the SEC.

Global smartphone shipments are set to fall 16.7% in 2026, the steepest annual decline on record, according to IDC.

Memory prices have climbed more than 300% from a year earlier, according to IDC.

Ternus promises staff a “phenomenal” launch ahead of Apple’s Sept. 9 event.picture alliance / Getty Images

The market has already started pricing an outcome

Apple shares rose about 3% on Ternus’s first day, closing near the upper end of a 52-week range that runs from roughly $226 to $345, with market capitalization around $4.75 trillion. Verify those figures at publication, since they move intraday.

What struck me when I mapped the memo against the guidance is how little room that leaves. A stock this close to its high has already absorbed a good launch. Phenomenal is the price of admission, not the upside.

More Apple News:

Apple’s latest move could change how you use AI every day

Bank of America predicts big changes for Apple after Tim Cook

Wall Street is starting to warm to Apple’s $2,199 foldable gamble

The product widely expected to carry the event is Apple’s first foldable iPhone, a device my colleague Faizan Farooque examined in Apple’s new iPhone strategy comes with a $2,500 question. Research firm IDC projects an average selling price above $2,550 for the category.

A folding phone is also the hardest thing Apple could choose to build during the worst component squeeze in years. Hinges, dual displays, and memory-hungry silicon all land on the same bill of materials that IDC says has been reshaped by memory costs.

Ternus ran hardware engineering for five years before this. He knows precisely how difficult that is, which is part of why the word he chose is worth noticing.

What Sept. 9 actually settles for Apple investors

The genuine test next week is not whether the foldable looks impressive on stage. Apple keynotes are reliably impressive on stage.

The test is whether Ternus repeats the word in front of analysts, or quietly replaces it with the language of supply constraints and staged availability. That substitution, if it comes, will tell you more than any spec sheet.

There is a second thing to watch that has nothing to do with hardware. Cook spent 15 years building a reputation for underpromising, and that reputation is a real asset priced into a $4.7 trillion company.

If Ternus turns out to be a chief executive who talks bigger than he guides, the market will eventually charge him for it. If he delivers a December quarter that makes “phenomenal” look conservative, he will have bought himself years of benefit of the doubt in a single sentence.

New chief executives are usually judged on their first product. Ternus will be judged on whether he meant what he wrote on a Tuesday morning before anyone had seen it.

More on Apple & its stock: 

History of Apple: Company timeline and facts

Who owns Apple? Institutional holdings & executives’ shares

Does Apple pay dividends? A history of rewarding shareholders

Apple’s stock split history: Everything you need to know

John Ternus’s net worth as Apple’s next CEO

McDonald’s franchises nearly all of its restaurants in the United States, which means franchise operators have a significant say in operational changes. In some cases, for example, they can reject deal pricing.

Usually, when McDonald’s rolls out a promotion like $2.99 Snack Wraps or $4 breakfast meals, only select franchise operators will not adopt the deal. Maybe franchisees operating in airports, rest stops, and other expensive real estate will pass on the deal, but most franchisees generally come on board.

That seemed to be what was happening with the chain’s new $3 Value Menu.

CFO Ian Borden noted during the chain’s first-quarter earnings call that franchisees overwhelmingly supported adding the promotion.

“With unanimous approval through the franchisee field votes, we launched the revamped McValue platform in mid-April. The new under $3 menu features well-known a la carte items available throughout the day,” he said.

Unanimous sounded encouraging, but in reality, something else happened that put customers in a challenging position.

McDonald’s franchisees passed on the $3 menu

As a retail and restaurant writer who has been covering these industries for over 30 years, I generally expect heavily-promoted deals to actually be on the menu when I visit a chain. That has not always been the case for McDonald’s $3 Every Day Affordable Price or EDAP menu.

CEO Christopher Kempczinski explained the menu during the Q2 earnings call.

“That is the EDAP menu. You could call that 10 items for under $3. That was sort of the last piece that we felt like we needed to get done in the U.S., and that was what McValue 2.0 was, as we referred to it. That’s what we introduced in April of this year,” he said.

The $3 menu, he noted, “has not delivered against our expectation,” and the CEO shared the reason for that.

“Part of that was due to the fact that we’re getting really inconsistent execution. Only about, call it, 60% to 65% of our system is currently executing the recommended pricing architecture with the 10 items for under $3,” he added.

That’s a large number of McDonald’s not delivering on a promise that the chain has made to its customers.

McDonald’s actually pulled other deals

While Kempczinski was careful to not blame franchise operators for the failure of the $3 menu, he did share that the chain had pulled other promotions to clear the way for its adoption.

“We compounded that unintentionally by our system pulling off a lot of digital offers. And digital offers for us is something that is core to kind of our loyalty program. It’s something that’s valued by our most loyal customers. And so that ended up being a bad trade,” he said.

More Restaurants:

52-year-old international restaurant chain closing all locations

46-year-old casual dining chain closes underperforming locations

Classic burger chain has closed down all its restaurants

He blamed much of the company’s weak 0.8% Q2 U.S. same-store sales growth.

“Putting in an EDAP program that didn’t deliver and taking away a lot of digital offers and the Buy One, Add One program that was the point I referenced or Ian referenced in the call, which is 2/3 of our miss in the quarter was related to that bad trade,” he shared.

Some McDonald’s owners opt out of certain promotions. Shutterstock

McDonald’s needs buy-in from franchise operators

The chain’s franchise operators do not have to use promotions created by corporate.

“McDonald’s prices vary by location. Ninety percent of McDonald’s restaurants are independently owned and operated by franchisees, who have the ability to set their own prices,” the company shared on its website.

That’s something the chain has to manage actively, according to RTM Nexus CEO Dominick Miserandino.

“McDonald’s corporate can spend millions on national ad campaigns promoting $5 meal deals, but the franchisee owns the register — and that’s where the strategy falls apart,” he told TheStreet.

Franchisees, however, don’t have to listen, and many have good reasons not to.

“Corporate can set a recommended price, but these operators are independent business owners fighting local wage spikes and rising food costs. When a franchisee looks at a low-margin national promo and realizes it eats into their bottom line, they simply opt out or jack up prices elsewhere on the menu to offset it,” he added.

Consumers, however, don’t always understand why the commercials they see don’t match the reality in their local McDonald’s.

“That creates a massive disconnect for consumers. You see a dollar deal on TV, drive up to the window, and get charged full price because the local owner refused to take the margin hit. McDonald’s biggest pricing battle isn’t with food inflation — it’s with its own franchisees protecting their unit economics,” Miserandino shared.

ALSO READ: After closing 39 locations, 76-year-old Mexican chain has 1 left

Broadcom Inc. is slated to report earnings after the closing bell on on Sept. 2, 2026, previewing how the “Anti-Nvidia” is holding up amid a boomtimes in the AI infrastructure sector.

Here are the figures that analysts polled by LSEG are looking for in today’s report:

Revenue: $29.519 billion

Earnings per share (adj): $3.25

Updates will be posted here as they become available. This page will refresh automatically as updates are posted.

JetBlue Airways (JBLU) is heading into a more significant phase of its comeback without one of the activist investors who helped alter its board.

Carl Icahn sharply cut his JetBlue position, selling roughly 8.13 million shares for about $40.4 million between Aug. 17 and Aug. 20, regulatory filings show. The transactions reduced the Icahn Group’s interest to around 3.32%, or about 12.5 million shares, from over 10% when the activist investor initially revealed his position in 2024.

That’s more than some spring cleaning.

Icahn’s smaller stake fell below the amount needed to retain board presence at JetBlue. Jesse Lynn and Steven Miller, directors supported by Icahn, resigned both effective Aug. 24 after the group’s ownership fell below the appropriate level, the airline said in an Aug. 27 filing with the SEC.

It is interesting timing for investors. JetBlue is aiming to show that its JetForward turnaround will restore sustainable profitability after years of strategic missteps, excessive expenses, and a failed attempt to combine with Spirit Airlines.

Icahn isn’t necessarily suggesting turnaround is going to fail.

But he’s pouring a lot less money and influence into it.

Carl Icahn has cut most of his original JetBlue position

Icahn originally disclosed an almost 10% stake in JetBlue in February 2024, referring to the airline at the time as undervalued.

The investment finally gave his business a significant say in running the corporation. JetBlue agreed to appoint Lynn and Miller to its board as Icahn designees.

Icahn Group had around 33.6 million shares as of the period referenced in JetBlue’s 2026 proxy statement, making it the airline’s third biggest stakeholder behind BlackRock Financial Management and Vladimir and Angelica Galkin.

Icahn’s stake is now roughly 12.5 million shares after the recent transactions.

That implies the activist has cut his investment by around 21.1 million shares, or nearly 63%, from the 33.6 million-share stake revealed previously.

The Icahn Group also dropped under the 5% ownership barrier on Aug. 18. That’s important because stockholders holding more than 5% of a public firm are normally subject to extra SEC beneficial-ownership reporting requirements.

Below that barrier, investors may have less instant insight into any changes in Icahn’s JetBlue holdings.

In practice, the market is now looking at an activist investor who has drastically decreased both his economic exposure to and formal power over the airline.

JetBlue is trying to make its turnaround numbers work

The pullback comes as JetBlue paints investors a rosier financial picture.

The airline recorded operational revenue of $2.7 billion for the second quarter of 2026, up 14.5% year-over-year, driven by a 10.9% increase in revenue per available seat mile.

System capacity was up 3.2%.

JetBlue also said it recovered about 50% of increased fuel costs in the quarter, better than management had projected.

The gains were substantial enough for JetBlue to reinstate its full-year 2026 estimate and create a longer-term aim of at least $1 in profits per share in 2028.

More Airlines:

Another airline cancels flights until August, offers some refunds

Another airline will be dissolved, all flights canceled

Airline shuts down, all flights grounded after accident

But the expense side of JetBlue’s financial statement still illustrates why the turnaround remains a tough one to pull off.

Related: JetBlue’s earnings beat hides $407 million warning

Second-quarter operating expense per available seat mile rose 17%. JetBlue’s average fuel price rose $1.83, or 76%, to $4.23 per gallon. Even without fuel and other items, unit costs rose 2.4%. That simplifies investor equations.

JetBlue is seeing greater revenue growth and demand, but it still has to translate those gains into consistently higher margins and profitability.

Carl Icahn just removed himself from JetBlue’s turnaround storyBloomberg / Getty Images

Icahn is giving up more than shares

What happened to Icahn’s board participation is one of the most illuminating elements of the narrative.

The company said Lynn and Miller resigned from the JetBlue board as of Aug. 24.

Lynn also resigned from the audit, governance and nominating, and finance committees, while Miller resigned from the audit and finance committees.

Neither departure included any issue with JetBlue about its operations, rules or procedures, the firm said.

The two directors were praised by JetBlue CEO Joanna Geraghty for their efforts as the airline grew and started implementing JetForward.

Icahn also struck a cordial tone. “We appreciate the constructive partnership with JetBlue over the years.”

JetBlue stated its board would have 11 directors, 10 of whom are independent, after their departures.

The phrase indicates a polite withdrawal, not a public conflict.

But for shareholders, the economic signal still outweighs the tone.

Activist investors often leverage their positions by holding enough shares to interest management and other shareholders in their next moves.

Icahn currently controls a far smaller piece of JetBlue than he did when he first bought into the company, and he has no representation sitting inside the boardroom.

The biggest question is whether Icahn sees better uses for his capital

There are a number of ways investors might take the sales.

Icahn may have just been shifting funds after owning JetBlue for more than two years.

He also could feel the turnaround at the airline has progressed far enough that board presence is no longer needed.

But there is one interpretation that is difficult for shareholders to shrug off. One of JetBlue’s most notable activist investors could consider the risk-reward equation as less compelling than it was in early 2024.

Although management pursues its JetForward strategy, JetBlue faces high operating costs, grounded aircraft, and fuel price pressure. The company wants $850 million to $950 million of incremental EBIT by 2027 and $1 per share of earnings in 2028.

The aims make the next several quarters essential.

JetBlue needs more than just improved demand.

It has to prove that increased income can surpass the expenses that have continually snaked their way back to sustainable profitability.

JetBlue now has to prove the turnaround without Icahn at the table

And it is the best part of the retreat of Icahn for the average investor.

His first investment in JetBlue sent an unspoken message: The airline was worth more than the market was giving it, and activist intervention might help unlock that value.

Now, two years later, JetBlue is still trying to get that reward.

The revenue is looking up. Management set a sound earnings objective for 2028. JetForward is making meaningful progress.

The investor, who formerly held about 10% of the firm, has cut his share to only 3.3% and given up his seats at the board table.

That’s not to say Icahn has lost trust in JetBlue.

But it does alter the burden of evidence.

The next chapter of JetBlue’s recovery will be assessed less on what an activist investor could push management to do and more on what the airline can actually achieve on the revenue, cost, and profitability fronts.

And with Icahn now holding roughly 63% fewer shares than his earlier 33.6 million-share position, his own capital is sending a message worth watching.

Related: JetBlue Airways makes an onboard dining change

Large companies eventually learn that the type of number it chooses to publicize matters more than the number itself.

Pick the right comparison, and an uncomfortable fact turns reassuring. Pick the wrong one, and you have handed your critics a slogan they will repeat for years.

Wisconsin has become one of the busiest testing grounds for that lesson. The state pulled in tech money for the same reasons it once pulled in paper mills and breweries. Those reasons are cheap land, a stable grid, and Lake Michigan sitting right there.

More than 40 data centers now operate in the state with several more proposed, according to the Wisconsin Policy Forum.

The public has not warmed to them. Seventy percent of Wisconsinites believe the costs of large data centers outweigh the benefits, and “water impacts were cited most often as the reason,” said Tressie Kamp, assistant director of the University of Wisconsin-Milwaukee’s Center for Water Policy, in an interview with WPR.

That poll result is the backdrop for a comparison Microsoft (MSFT) has been making about its Mount Pleasant campus, and for the pushback that comparison is now drawing.

Why data-center water became a Wisconsin flashpoint

Data centers consume water in two ways, and only one of them shows up on a local utility bill.

The visible one is cooling. Servers throw off heat, and older facilities push water through the building to carry that heat away, losing a share of it to evaporation every hot afternoon.

This is the number that lands in permit filings and city council meetings, and it is the one Microsoft has spent two decades driving down.

More Technology:

Marvell’s $120B AI deal came with an unexpected catch

Two U.S. tech giants power Saudi’s AI vision

Apple’s new CEO inherits a fortune and an AI question

The invisible one is electricity. Power plants burn through enormous volumes of water to generate the power a data center draws, and that water gets withdrawn somewhere else entirely, usually far from the campus and never on its permit.

Roughly 70% of Wisconsin’s water withdrawals are already attributable to power generation, according to Clean Wisconsin.

That split matters more than it sounds. A facility can look almost waterless at the fence line while adding real strain to a watershed 100 miles away.

The transparency gap is the reason the argument keeps going. Quantifying offsite water use is difficult because developers, tech companies, and public utilities disclose so little, and without knowing the fuel and cooling system at each generating station, the number cannot be pinned down, according to Clean Wisconsin.

When I have covered permitting fights before, this is the pattern that decides them.

Neither side is lying about its own number. They are simply measuring different things and talking past each other in public.

Microsoft compares its Wisconsin AI campus to a restaurant, but critics cite 211 billion gallons.Anadolu / Getty Images

What Microsoft says its Wisconsin campus actually uses

Microsoft completed the first building at its Fairwater campus in Mount Pleasant in June 2026, describing it as the world’s most powerful supercomputer. It also won approval in January 2026 for 15 additional buildings on the site, according to Data Center Dynamics.

The company’s cooling design is the reason it can make the claim. The Mount Pleasant facility runs a closed loop that circulates liquid directly to each chip. It’s filled once rather than continuously topped up, which eliminates evaporation entirely.

Related: Palantir CEO escalates Microsoft’s AI warning 

The result is the line now circulating online. The Mount Pleasant datacenter “uses about as much water as a typical restaurant each year,” according to Microsoft.

I went looking for the underlying numbers rather than the analogy, and the record is more textured than the one-liner suggests.

Microsoft’s data centers used 0.27 liters of water per kilowatt-hour last year, roughly three times better than the industry average, according to GeekWire.

City of Racine records show the first Mount Pleasant phase permitted for up to 2,814,000 gallons in 2026, rising toward 8.4 million gallons as the campus expands, reported WTMJ.

Internal company forecasts put annual water needs across roughly 100 campuses at about 18 billion liters by 2030, up 150% from 2020, according to The Register.

Those three figures are all true at once. The efficiency is real, the local draw is modest by industrial standards, and the total keeps climbing anyway.

The number that Microsoft’s restaurant comparison leaves out

Here is where my analysis parts ways with the press release.

U.S. data centers consumed an estimated 17 billion gallons of water directly in 2023, and the indirect figure, the water consumed offsite generating their electricity, was 211 billion gallons, Kamp told WPR, citing Lawrence Berkeley National Laboratory research.

Direct cooling is therefore about 7% of the total. The restaurant comparison describes that 7% and stops.

The irony is that closed-loop cooling can make the offsite share worse. Sealed systems use more electricity to run than evaporative ones, and in a grid where most power still comes from fossil fuel and nuclear plants, more electricity means more water withdrawn at the generating station.

Wisconsin sits inside the MISO grid, where roughly 70% of electricity comes from fossil fuels and 14% from nuclear, according to The Daily Reporter.

So the campus that evaporates nothing on site can still carry a heavier total water footprint than the one it replaced. That is the critics’ actual argument, and the restaurant line does not engage it.

What the water fight means for AI infrastructure spending

This is not an environmental story wearing a business costume. It is a permitting story, and permitting is now a line item.

Microsoft’s capital expenditures and finance leases reached about $175 billion in fiscal 2026, and management has guided to roughly the same figure for fiscal 2027, according to CNBC. Spending at that scale only works if sites get approved on schedule.

Every month a campus sits in front of a village board is a month of idle capital. Microsoft already walked away from a Caledonia site after community opposition, which is the clearest possible signal that local sentiment now has pricing power over the buildout.

For anyone holding MSFT, or holding an index fund that holds it, that is the transmission mechanism worth watching. Not the water itself, but the delay risk the water debate creates. 

Wisconsin utilities have already approved roughly $2 billion of new power plants with close to $6 billion more proposed, much of it to serve data center demand, according to Marquette University Law School. Those costs land in a rate base that households share.

That is the quiet part of the story for anyone who lives near one of these campuses. The water argument is loud, but the electric bill is the thing most residents will actually feel first, and it is the thing utility regulators can act on fastest.

The fix is not a better analogy. It is disclosing the offsite number, which no hyperscaler currently does at the site level, and which Clean Wisconsin has said it cannot calculate without that transparency.

Whoever publishes it first will end this argument. Until someone does, the restaurant line will keep being quoted and getting picked apart, because the people asking the question already know it answers a different one.

Related: Microsoft makes a controversial decision that changes its AI story

Artificial intelligence agents are being built to do more than answer questions and generate content.

The next generation of AI systems is increasingly equipped to take actions: buying computing power, accessing data, hiring software services, negotiating with other systems, and completing transactions without a human approving every individual step.

That raises a question with significant implications for investors and the companies building the infrastructure behind AI. What happens when machines begin participating in the economy at a scale that existing financial systems were never designed to handle?

Tom Lee, co-founder and head of research at Fundstrat Global Advisors, has raised a provocative possibility. If autonomous AI agents eventually conduct enormous numbers of transactions and traditional payment infrastructure proves too slow or restrictive, machines could gravitate toward alternative systems for exchanging value, CoinDesk reported.

Lee has also argued that AI agents could eventually cut humans out of economic activity entirely if financial infrastructure does not evolve to keep them accountable.

The bigger investment question is whether the AI trade could eventually extend beyond chips, data centers, and models to the systems that allow autonomous software to actually operate in the economy.

Why the financial system was built for people, not machines

Today’s payment infrastructure reflects the needs of human users. Consumers make purchases. Businesses pay suppliers. Banks identify account holders and monitor transactions. Friction and human oversight are features rather than bugs because they help prevent fraud and provide accountability.

AI agents could operate very differently. An autonomous system managing a complex task may need to purchase small amounts of computing power, pay for individual API calls, acquire data or compensate other agents for services. Those transactions could happen continuously and in volumes that make traditional payment processes impractical.

More AI:

Nvidia just made a move Wall Street wasn’t ready for

Microsoft just took sides in AI policy fight

OpenAI just disclosed something genuinely alarming

“The real gap is not just speed, but a machine-readable framework for trust and authorization,” Logan Xie, leader of KuCoin AI Lab, told TheStreet in an interview.

A person might authorize an agent to spend up to a certain amount and establish rules governing what it can buy. The agent then executes thousands or millions of transactions within those boundaries. Traditional financial systems can automate some of that activity, but they were not designed around software entities making continuous microtransactions on behalf of users.

“Card economics put a floor of a few cents under every transaction, so a payment of a fifth of a cent simply cannot exist on those rails at any fee level,” Mark Zalan, CEO of GoMining, told TheStreet. “The machine economy runs on exactly those payments: compute, data, API calls, bought continuously in tiny increments.”

Would AI agents actually create their own currency?

Lee’s suggestion that AI agents could develop alternative systems for exchanging value is easy to interpret as a prediction that machines will create their own money. The experts interviewed for this story see a more complicated outcome.

Creating a token is not the same thing as creating a functioning monetary system. Money does not work simply because software creates something and labels it currency. A functioning monetary system depends on trust, acceptance, liquidity, and its connection to the broader economy.

“If traditional infrastructure cannot meet the needs of machine commerce, agents are more likely to use stablecoins, blockchains, or other programmable financial instruments than to create a monetary system detached from the human economy,” Xie added.

Programmable payment instruments and digital settlement assets could give machines something closer to what they need: value that can move continuously through automated networks without relying on traditional banking processes for every transaction.

Rather than consciously designing a new monetary system, their collective behavior could produce one instead.

“They’ll gravitate to whatever settles fastest and cheapest with the fewest permissions, and the sum of billions of those cold, unsentimental choices will look, in retrospect, like a monetary order nobody voted for,” Zalan explained.

Machines may not need to hold a conference or vote on a new monetary standard. If autonomous agents select the cheapest and most efficient way to exchange value, their choices could concentrate economic activity around certain networks and assets. The system would emerge from usage rather than deliberate design.

Agents cannot hold bank accounts. Public payment rails are currently the only place software holds value directly.Mariya/Getty Images

Why programmable payment rails are entering the AI conversation

Programmable payment networks operate around the clock, can be accessed directly by software, and can use automated contracts to enforce transaction conditions. Those characteristics make them a natural candidate for machine-to-machine transactions.

The major card networks are already acting on that logic. Mastercard launched Agent Pay for Machines on June 10, 2026, a platform built for machine-speed transactions across cards, accounts, and digital settlement assets, as TheStreet reported.

Visa, Stripe, and Mastercard have all built out tools and protocols in anticipation of agent-driven commerce, Fortune reported. If traditional banking infrastructure could carry this commerce alone, there would be no reason for the card networks to build on public networks. They are building on them anyway.

Lee’s position reflects a bet on that structural shift. BitMine, which Lee chairs, has built one of the largest corporate Ethereum treasury positions, holding approximately 5.85 million ETH, roughly 4.8% of the circulating supply.

Lee believes programmable settlement networks are best positioned to become the financial foundation for the machine economy.

What the AI agent economy means for investors

The idea that machine-to-machine commerce could require new transaction infrastructure does not automatically mean any particular network, token, or company will benefit. The competition will come down to familiar factors, including transaction costs, speed, liquidity, security, and developer adoption.

Agents cannot hold bank accounts. Public payment rails are currently the only place software holds value directly. Which network’s machines actually select for settlement will be decided by fees, finality, and neutrality, not by whose treasury holds the most of any given asset.

The agent economy could create opportunities across identity systems, digital wallets, cybersecurity, payment infrastructure, and financial rails. The crucial issue running through all of it is accountability.

An AI agent may execute a transaction, but someone ultimately needs to be responsible for what it does. Identity, permissions, and governance could become just as important as transaction speed.

The most realistic near-term outcome is probably not AI agents declaring independence from the human financial system. It is the gradual development of financial infrastructure designed around the needs of software.

Those changes could become significant enough to reshape how value moves through the economy, and the companies that solve the infrastructure problem early may prove to be among the most consequential investments of the AI era.

Related: Google DeepMind prepares for risk of AI agents going rogue

Viking Holdings (VIK) had a great run this year, and then it gave a big chunk of it back.

The luxury cruise operator hit an all-time high of about $108 on Aug. 5. By Sept. 1, the stock had slipped to roughly $86, a drop of nearly 20% in under a month.

Jim Cramer says that drop is a buying opportunity, not a warning sign.

On CNBC’s “Mad Money,” Cramer told viewers to buy Viking into its weakness. 

“I think it’s crazy that people have been selling this thing,” he said. “I’m telling you to buy the stock into its recent weakness.”

Cramer has backed Viking since shortly after its May 2024 initial public offering, so his call carries some weight. 

Why Viking Holdings stock dropped nearly 20% from its August high

The sell-off came from two directions at once.

The first is bigger than Viking. Cruise stocks in general came under pressure from rising oil prices and geopolitical uncertainty, both of which raise costs and make investors nervous about leisure spending.

More Jim Cramer Coverage:

Jim Cramer’s 5 investing themes for the rest of 2026

Jim Cramer says wait for a pullback before buying Target stock

Jim Cramer recommends buying Reddit stock

The second problem is specific to Viking. According to CNBC, historically low water levels on the Danube and Rhine rivers forced the company to change some itineraries.

Viking is best known for its river cruises, so disruptions there hit a core part of the business. The company is now issuing vouchers to affected passengers, and those costs will carry into 2027 and 2028 as travelers redeem them.

For a stock trading at a premium, any threat to the core business tends to trigger a fast repricing. That’s a big part of what happened here.

Viking targets older, affluent travelers, a customer base Cramer says holds up better when budgets tighten.oksanaphoto / Getty Images

Cramer’s case for buying Viking Holdings stock now

Cramer’s argument is simple: The business is doing better than the chart suggests.

His first point is the customer base. Viking caters to older, affluent, English-speaking travelers, a group that keeps spending even when inflation or higher energy prices squeeze everyone else.

His second point is demand, and Viking’s second-quarter earnings release backs it up.

As of Aug. 9, Viking had already sold 96% of its core capacity for the 2026 season and 53% for 2027. 

Advance bookings for 2027 reached $4.71 billion, up 21% from the same point a year earlier.

That gives Viking rare visibility into future revenue. When most of next year’s cabins are already booked, a rough month for the stock says little about the underlying business.

The Q2 numbers were strong, too. 

Viking posted adjusted earnings of $1.31 per share on revenue of $2.19 billion, up 16.5% year over year, Investing.com reported. That beat Wall Street expectations of about $0.95 per share. 

“What can I say? Buy the dip,” Cramer told viewers, according to CNBC.

What Cramer’s valuation call gets right, and where it’s still steep

The stock trades at about 28.6 times trailing earnings, which is what you’ll see on most quote pages. But Cramer’s case rests on forward earnings.

The sell-off dropped Viking’s stock price to 22 times its expected earnings for next year. Jim Cramer argued this higher price tag is worth it, given Viking’s strong growth and profits. 

That’s still pricier than traditional operators such as Carnival (CCL) or Norwegian (NCLH). 

Investors are paying up for Viking’s faster growth, and that premium can shrink quickly if the river problems drag on.

The risks Viking Holdings investors shouldn’t ignore

The balance sheet is the first thing to understand. 

Viking is funding an aggressive fleet expansion, and heavy spending means the company is sensitive to interest rates and any slowdown in bookings.

Related: Jim Cramer reveals 6 AI stocks to watch in 2026

The good news is that leverage looks manageable today. Viking reported net leverage of 1.2 times and about $4 billion in cash as of June 30.

The valuation is the second risk. 

At a premium multiple, Viking has less room for error than cheaper cruise names, so any extended river disruption could pull the stock lower before the story improves.

Consider these three points before buying.

3 things to weigh before following Cramer into VIK

Time horizon: Cramer treats Viking as a long-term hold, not a trade.

River risk: Low water levels and voucher costs will pressure margins into 2027 and 2028.

Valuation: A forward multiple near 22 times leaves little cushion if growth slows.

How Viking Holdings stock stacks up against the S&P 500

Even after the pullback, Viking is up about 19% year to date, well ahead of the S&P 500, which set a record close of 7,737 on Aug. 4.

The past five days tell the other side of the story. 

Viking fell about 7% over that stretch, a reminder that the stock swings harder than the broad market in both directions.

So the same stock is both a strong yearly performer and a shaky recent one. 

With a stock this volatile, picking the right time to buy is important. 

What Viking Holdings investors should do next

If you find Cramer’s case convincing, the practical question is how to buy, not just whether to.

One approach is to scale in gradually instead of buying all at once. Spreading your purchases over time lowers the risk of buying right before another drop while volatility stays high.

The stock is trading above its 200-day moving average, which sits at about $74. 

Long-term investors watch this line to spot the big market trend, and Viking is staying safely above it. 

Cramer’s Viking call fits a pattern he repeats often: buy quality businesses when fear pushes the price down, then hold. 

Analysts broadly agree the company is solid, with a Moderate Buy consensus.

Viking’s bookings and wealthy customer base support the bull case, but the premium valuation and river disruptions are real. Size the position to your own risk tolerance and treat it as a long-term hold rather than a quick gain.

Related: Jim Cramer reveals his 20% rule for winning stocks

With multiple pressures on the airline industry this year, many smaller airlines have been forced to shut down.

In other cases of funds running out, an airline that was in the middle of launch preparations can fail to get off the ground entirely. British carrier Ecojet Airlines entered voluntary liquidation, or the British equivalent of a voluntary Chapter 7 bankruptcy, in January 2026 after its plan of converting their kerosene engines to hydrogen-electric ones bled too much money to be worth continuing.

In July 2026, shareholders of what was supposed to be a new low-cost regional Italian airline called Aerolinee Siciliane voted to dissolve and liquidate before operations could begin due to high expenses and challenges securing the necessary approvals.

Creditor petition asks to put Global Airlines into involuntary liquidation

While not currently in liquidation or any other form of formal process that can lead to it shutting down, virtual airline Global Airlines is the latest name to face mounting pressure to be put into involuntary liquidation by a creditor.

The British airline startup was first proposed in 2019 by online travel platform Holiday Swap founder James Aqsuith with a plan to purchase Airbus A380 planes and launch as a transatlantic budget airline similar to Norse or French Bee.

One flight from Glasgow to JFK in New York ran through a damp lease, or leased aircraft and pilots, from Hi Fly Malta but the one A380 plane purchased by Global Airlines has since been put in long-term storage in southern France and much uncertainty has swirled about the carrier’s future.

Leadership changes, depreciating funds and delays with completing the necessary maintenance have all delayed progress to the point that some are speculating that the airline is a failed project.

Global Airlines is a British virtual airline that has been grounded since 2025.Global Airlines

What is happening with British airline startup Global Airlines

As was first reported by DJ’s Aviation, a woman claiming to be a creditor filed a petition asking for the High Court of Justice, Business and Property Courts of England and Wales to place Global Airlines in involuntary liquidation back in May.

A similar petition was filed by another creditor in August 2025 but has since been withdrawn. As it can theoretically be rejected by the court for being meritless, this type of petition alone is not an indication of an airline’s state in one way or another.

A further hearing (the second one occurred behind closed doors in July) has been set for Sept. 15 while Global Airlines could not be immediately reached for comment on the situation.

More Travel News:

Airline to launch unusual new flight to Cayman Islands from the U.S.

There is a very cool Irish version of swimming pigs in The Bahamas

Unexpected country is most luxurious travel destination for 2026

Low-cost airline launches easier way to get to Sri Lanka

Some airlines that shut down or filed for bankruptcy in 2026:

Spirit Airlines: The largest airline shutdown of the year occurred when Spirit Airlines canceled all remaining flights on May 2. Although the airline had filed for Chapter 11 protection twice before, the skyrocketing price of jet fuel dealt the final blow to its operations.

Magnicharters: The Mexican low-cost airline canceled all of its flights and eventually shut down in a collapse that left thousands stranded.

Starflite Aviation: Houston-based Starflite Aviation had its AOC license revoked in March 2026, amid FAA claims that owners falsified pilot training records to bypass safety audits.

AlpAvia: Slovenian charter airline AlpAvia also shut down in March 2026 over financial problems.

Related: Another airline shuts down, cancels all flights due to low demand