Hey Nutters. ☕️

What happens when artificial intelligence gets dramatically cheaper?

Wall Street's first answer was: bad news for chipmakers.

But there's another possibility.

Cheaper AI could mean more users, more queries, more applications—and ultimately, a whole lot more computing power.

That's the debate after Moonshot AI's new Kimi K3 model arrived with frontier-level performance at a fraction of competitors' prices.

Investors initially hit the sell button on semiconductor stocks.

But the AI industry's oldest economic paradox may be about to show up again:

When technology gets cheaper, we tend to use a lot more of it.

And that could be very good news for the companies selling the picks and shovels.

Let's dive in. 👇

The AI trade is officially entering its discount era.

Moonshot AI released Kimi K3 last week, delivering performance that reportedly matched Anthropic's Fable 5 and OpenAI's GPT-5.6 Sol across several benchmarks.

The surprise?

It costs dramatically less.

💰 Kimi K3: $15 per million output tokens

🤖 OpenAI GPT-5.6 Sol: $30

🧠 Anthropic Claude Fable 5: $50

That's frontier-level AI at roughly half to one-third the price of major competitors.

Markets initially interpreted the development as bad news for the AI infrastructure trade.

If companies can achieve similar results with cheaper models, perhaps they won't need to spend quite as much on chips and data centers.

Chip stocks sold off.

But there might be one problem with that theory:

People really like cheap things.

Kimi K3's launch quickly demonstrated the other side of the equation.

Demand was so intense that Moonshot AI reportedly had to pause new user registrations over the weekend because of compute constraints.

Think about that for a second.

A cheaper AI model didn't reduce infrastructure demand.

It overwhelmed the available infrastructure.

That's why some investors are beginning to see falling AI prices as potentially bullish for semiconductor demand.

As inference becomes cheaper, companies can afford to deploy AI across more tasks:

💻 Software development

🤖 Autonomous AI agents

📊 Financial analysis

🛍️ Customer service

🔬 Scientific research

🏭 Industrial automation

One AI request getting cheaper matters less if businesses suddenly start making 100 times more requests.

The economics are similar to what happened with computing, internet bandwidth, and cloud storage.

Costs fell.

Usage exploded.

The biggest beneficiary could be one of the hottest corners of the semiconductor market:

High-bandwidth memory.

Kimi K3 reportedly contains roughly 2.8 trillion parameters.

Although only a fraction of those parameters may be actively used during each inference request, the entire model still needs to be stored and accessible.

That's a lot of memory.

And it makes running enormous frontier models on ordinary enterprise servers extremely difficult.

Instead, businesses increasingly need access to cloud-scale AI infrastructure packed with specialized chips and enormous amounts of HBM.

That could keep demand elevated for the three companies dominating advanced memory:

💾 Micron

🇰🇷 SK Hynix

🧠 Samsung

With HBM supply already tight, continued AI adoption could preserve the industry's pricing power longer than investors expect.

In other words:

The model might be cheaper.

The infrastructure running millions of copies of it isn't.

Memory isn't the only potential winner.

Larger AI models require massive clusters of accelerators working together.

That means continued demand for:

🟢 GPUs

🔗 Networking equipment

💾 High-bandwidth memory

⚡ Power infrastructure

🏢 Data centers

Wedbush analyst Matt Bryson has noted that increasingly large AI models could require more AI chips operating simultaneously, potentially benefiting semiconductor companies throughout the infrastructure stack.

That puts companies like Nvidia and Broadcom back in the conversation.

The first wave of the AI boom was about training increasingly powerful models.

The next wave could be about deploying those models everywhere.

And inference at enormous scale may ultimately require just as much infrastructure.

Kimi K3 also highlights another important shift.

China's AI startups aren't slowing down.

They're raising capital.

As models become cheaper to develop and operate, competition between U.S. and Chinese AI companies could intensify dramatically.

That means more companies building models.

More companies buying compute.

More companies competing for chips.

And more demand flowing through the global semiconductor supply chain.

The AI race may be getting cheaper at the software layer while simultaneously becoming more expensive at the infrastructure layer.

There's also a political wildcard.

OpenAI and Anthropic are reportedly pushing Washington to tighten restrictions around open-weight Chinese AI models, arguing that unrestricted access could create national security risks.

Critics see another motive:

Competition.

Low-cost Chinese models could put significant pricing pressure on American AI companies—particularly as some of the industry's largest players prepare for potential public listings.

The Trump administration has yet to make a major move.

But Kimi K3's rapid adoption shows why the debate matters.

If Chinese labs can deliver frontier-level AI at dramatically lower prices, the global AI market could become far more competitive.

Wall Street's initial reaction was simple:

Cheaper AI = less chip spending.

Reality might be more complicated.

Cheaper AI could unlock millions of new users and applications, pushing demand for compute, memory, networking, and data centers even higher.

The next AI winner might not be the company with the most expensive model.

It might be whoever sells the hardware required to run billions of cheap AI requests.

This 8.5% Fixed-Rate Investment Has a Perfect Track Record

Body w/o Links
When markets get unpredictable, investors start looking for two things: liquidity and yield. Groundfloor’s Signature Note offers both, with up to 8.5% fixed returns backed by a diversified pool of residential real estate loans.
Since launching Notes in 2018, every investor has been paid their full principal and interest on time.
Choose your term from 1 to 12 months, lock in your rate, and know exactly when your money comes back. You can let it roll over at maturity to keep earning or take your cash out with no fees or penalties. Get $100 on your first $1000 investment with NOTES100.

While AI dominates the headlines, biotech has quietly emerged as one of 2026's strongest market themes.

Large pharmaceutical companies including Merck and AbbVie are aggressively hunting for acquisition targets as some of their biggest drugs approach patent expiration.

The industry has already seen 37 transactions worth more than $1 billion each, creating a powerful cycle of capital recycling.

Big Pharma buys a biotech company.

Investors get paid.

That money flows back into new biotech companies.

And suddenly the IPO window opens again.

Biotech IPOs have reportedly gained roughly 55%, attracting investors back into a sector that spent years struggling with high interest rates and limited funding.

But drug development remains brutally unpredictable.

One failed clinical trial can erase billions of dollars overnight, which is why some investors are choosing diversified vehicles such as the SPDR S&P Biotech ETF ($XBI) rather than betting on individual companies.

The biotech trade is officially back—but the volatility never left.

Commercial aviation has a strange problem:

There's plenty of demand.

There just aren't enough airplanes.

Airlines around the world are waiting years for new aircraft as Boeing and Airbus struggle with enormous order backlogs, supplier shortages, regulatory challenges, and production bottlenecks.

Boeing's immediate priority is getting its 737 MAX 7 and MAX 10 models certified, a crucial step that could unlock additional deliveries and improve cash flow.

Airbus, meanwhile, is trying to stabilize production of its popular A320 family as engine availability continues to create headaches.

For investors, the opportunity is obvious.

Global airlines need newer, more fuel-efficient aircraft.

The risk is execution.

The winner won't necessarily be the manufacturer with the largest order book.

It'll be the one that can actually deliver the planes.

General Motors just delivered a strong quarter, reporting roughly $48 billion in revenue and beating profit expectations.

Management also raised its full-year outlook for the second time this year, supported by strong demand for high-margin trucks.

But underneath those numbers sits a very expensive EV hangover.

GM has reportedly accumulated roughly $10.9 billion in EV-related charges since 2025, forcing the automaker to rethink the speed of its electric transition.

One of the clearest signals?

The company has reversed plans to turn Cadillac into an EV-only brand.

GM isn't abandoning electric vehicles.

But it is acknowledging something increasingly obvious across the auto industry:

Consumers—and profits—aren't moving away from gasoline as quickly as automakers once expected.

For now, GM's old-school trucks are helping finance its next-generation ambitions.

The U.S. administration is preparing to impose 50% tariffs on certain Canadian products, including autos, dairy, and alcohol.

The measures could affect roughly $20 billion in trade and would mark the first use of the Section 338 Tariff Act against a major U.S. trading partner.

The potential consequences stretch beyond Canada.

Higher tariffs can increase costs throughout tightly integrated North American supply chains, particularly in industries such as automobiles, where components frequently cross the U.S.-Canada border multiple times before a finished vehicle reaches consumers.

If the tariffs remain in place, companies may have to choose between absorbing the costs, restructuring supply chains, or passing higher prices to customers.

None of those options are particularly fun.

The happiest place on Earth is apparently getting more ketchup.

Kraft Heinz signed a multiyear partnership with Disney that will make the food giant an exclusive condiment and food supplier across Disney's theme parks and cruise lines.

For Disney, the partnership adds another major consumer brand to its enormous hospitality ecosystem.

For Kraft Heinz, the strategy goes deeper.

The company has been looking for ways to modernize some of its aging household brands and reconnect with younger consumers.

Disney gives it access to millions of families every year—and plenty of opportunities for co-branded products.

It's part licensing deal, part distribution agreement, and part marketing campaign.

Mickey Mouse, meet mac and cheese.

The world's most important chip manufacturer may be getting more expensive.

TSMC is reportedly planning price increases of roughly 5% to 10% beginning in 2027, as rising costs for materials, semiconductor equipment, and global manufacturing expansion squeeze margins.

Mature-node chips could see some of the largest adjustments.

The company has spent aggressively expanding manufacturing beyond Taiwan, including major investments in overseas facilities.

That geographic diversification may reduce supply-chain risk—but building semiconductor factories across multiple countries isn't cheap.

For customers, higher TSMC prices could eventually ripple throughout the technology industry.

For TSMC, the message is simple:

Everyone wants more chips.

And making them isn't getting any cheaper.

Save up to $130 on print orders with Staples.

Getting your business noticed on a budget is hard — especially without the right help.
 
With Staples, Print & Marketing Services, it’s easy to boost your business’s presence. Running a promotion? Hosting an event? Launching something new? Staples makes it easy to create, customize and print exactly what you need, when you need it.
 
Think flyers that fly in the foot traffic, direct mailers that gets right to the source, outdoor banners that’ll outlast any weather and so much more. Choose from ready-to-go templates or upload your own designs, then pick up in store or have it delivered straight to your business. Some prints are even available the same day when you order by 12 pm.

Thanks for reading Cash Nut! 🥜

Today's big idea: Don't automatically assume cheaper AI means less hardware.

If falling prices bring AI to billions more users, the semiconductor industry's biggest demand wave could still be ahead.

We'll be back with more business, markets, tech, and the stories moving your money.

See you tomorrow. 🚀