Hardcore gamers nurtured Nvidia and the AI boom. Now they feel betrayed by Big Tech.

By Christine Ji

'An ongoing culture war': AI is driving up the cost of key gaming components as Big Tech leaves the gaming industry, once an inseparable ally, in the dust

On most days after work, Jayden Weaver sheds his sales-manager role and logs in to his "Minecraft" account, where he and his YouTube audience of over 9,000 followers have built a sprawling modular kingdom. They've terraformed the desert terrain by planting trees, building houses, and even constructing a Walmart and an airport.

But not all is well outside of this idyllic "Minecraft" world on Weaver's gaming computer. The PC is equipped with a central processing unit, graphics card, random-access memory and high-capacity storage - the very same equipment that multitrillion-dollar tech companies are now hoovering up to power their artificial-intelligence ambitions.

The way 22-year-old Weaver sees it, the AI boom is threatening the hobby that has shaped his life since he was 8. Last fall, OpenAI inked deals with Samsung (KR:005930) and SK Hynix (KR:000660) to buy massive quantities of dynamic-random-access memory wafers for its Stargate data centers, contributing to a memory shortage. The average cost of the latest 32-gigabyte RAM kit, which Weaver considers the baseline standard for modern PC gaming, has increased to $560 from $150 in just one year, data from PCPartPicker shows. A 64-gigabyte RAM kit - a necessity for heavy game "modders," developers and diehard streamers - has surged to $1,000 from $250. Nvidia (NVDA) is slashing production of the company's consumer-facing GeForce graphics cards as it prioritizes delivering Blackwell and Rubin chips to its data-center customers, reports suggest.

Nvidia did not respond to MarketWatch requests for comment.

"They've taken the RAM that would have gone to consumers and flooded it into these data centers," said Weaver, who lives in Dawsonville, Ga. "That leaves people like me, who love building PCs, paying an arm and a leg for a crucial component." He hasn't been able to purchase any new components since the price hikes, he said. "It's held me back from buying and building PCs."

Alphabet (GOOGL) (GOOG), Amazon.com (AMZN), Meta Platforms (META) and Microsoft (MSFT) are planning to spend upwards of $700 billion on AI this year, a tidal wave of cash that's gobbling up endless amounts of chips, memory, compute hardware, networking equipment and power required for their data-center plans.

Big Tech's obsession with AI infrastructure is starting to feel like a betrayal to the legions of gamers who nurtured some of the key companies and technologies driving the boom. Today, they're deeply frustrated with the results as the gaming industry stagnates and consumers face higher prices on everything from memory to storage drives and graphics cards.

For decades, gaming subsidized artificial intelligence, forging the careers of today's AI visionaries and providing the scaffolding for major AI breakthroughs. Advancements in machine learning, in turn, spurred more complex game-play features. Jagged geometric shapes evolved into photorealistic video-game characters thanks to increased memory and more efficient architectures. The most important chip company in AI, Nvidia, and its graphics processing units, arguably would not exist if not for the gaming industry, which also spawned pioneering AI lab Google DeepMind.

A punishing hangover from pandemic-era overinvestment had already left the gaming industry reeling. Budgets initially ballooned as gaming studios raced to capture the attention of people stuck in their homes. As demand waned in the following years, companies conducted mass layoffs, leading private-equity firms to swoop in to pick over the remains of distressed assets.

Now, as Big Tech corners the hardware market and drives up component prices, gamers worry that the AI gold rush is coming at their direct expense. Microsoft alone is expected to account for $190 billion of AI outlays this year, fueling an infrastructure blitz anchored by a multibillion-dollar partnership with OpenAI that CEO Satya Nadella kicked off in 2019. Weaver is concerned that the company's gaming portfolio - home to massive properties such as "Minecraft" and the Xbox hardware line - will suffer as a result. "They're hurting their own console sales when they raise Xbox prices for a family that wants to play Xbox," he said.

Microsoft did not respond to a MarketWatch request for comment.

Weaver has aired his frustrations through YouTube videos, sparking a wave of agreement in the comments section. Steve Burke, editor-in-chief of the media channel Gamers Nexus, has garnered millions of views on YouTube videos bearing incendiary titles such as "Nvidia's AI Bubble" and "The Torture Will Continue Until Shareholder Value Improves."

The data-center boom is so lucrative that companies no longer need to cater to consumers as they did in the past, Burke told MarketWatch. It's posed an "existential risk" to gamers owning their own computing devices.

But worse than the price hikes is the unsettling feeling that consumers are being gaslit by tech giants paying lip service to the gaming community while delaying product releases or peddling AI-generated "slop."

"Really where the pain comes in for people is feeling like they're being lied to or manipulated," Burke said.

The gaming industry, once a partner, finds itself in the path of AI's advance.

Origin stories

NYU computer-science professor Julian Togelius studies the intersection of AI and gaming.

When Julian Togelius was 13, he used the proceeds from his gardening job to buy a used computer. The clunky machine had around the same amount of computing power as a modern-day toaster. Fascinated by computer games, Togelius taught himself to program so he could create his own.

"This is not a super uncommon kind of origin story," Togelius said of his childhood gaming experiences. "Many software people and developers and sometimes researchers of my age have similar experiences."

Today, the 47-year-old Togelius is a computer-science professor at New York University. As an artificial-intelligence researcher with a specialty in videogames, Togelius is caught in the middle of what he calls "an ongoing culture war" between his passions.

The gaming community has developed a "very, very fraught relationship with AI, and it's not getting better," according to Togelius. In fact, it's a "disaster."

Togelius is clinging to hope, grounded in his decades of industry experience, that the potential of AI will ultimately create more immersive virtual worlds.

"I've long had a vision of a big open-world game where you drive for hours in a random direction, and it creates new cities with new quests and architecture styles," Togelius said. Such a game would seamlessly mold itself to individual players' skills and tastes, constantly anticipating the narrative twists and visual landscapes that they would enjoy next.

From board games to first-person shooters, games have historically provided a structured, rules-based playground for digital algorithms to learn. Long before the modern GPU, computer-science legends like Alan Turing and Claude Shannon - the namesake of Anthropic's large language model - used chess to model human logic and cognition.

"You can have different scenarios, you can run them over and over again, [and] you can simulate them very fast," Togelius said of games. "When you fail, nothing bad happens."

After IBM's Deep Blue supercomputer scored a landmark victory over chess champion Garry Kasparov in 1997, researchers sought new simulations to push the frontiers of machine intelligence. Videogames, with their virtual worlds and continuous decision-making, were one promising avenue.

Against this backdrop rose one of the most influential figures of the modern AI field, Google DeepMind co-founder Demis Hassabis. DeepMind - established in 2010 by Hassabis and fellow technologists Shane Legg and Mustafa Suleyman - became the first modern artificial-general-intelligence lab, and its 2014 acquisition by Google famously galvanized the creation of OpenAI. Suleyman would go on to become Microsoft's CEO of AI.

Hassabis's life is a microcosm of the AI industry's broader evolution. At the age of 6 he was a chess prodigy. At 12 he was programming his own games. By 17 he had landed a full-time job at the game studio Bullfrog Games. It was there that Hassabis co-designed "Theme Park," a landmark game with bleeding-edge algorithmic capabilities that earned both critical acclaim and millions of sales. At 22, Hassabis graduated from the University of Cambridge and founded the gaming company Elixir Studios. Over a decade would pass before Hassabis could convince investors to back a pure AI research lab.

Around the same time as Hassabis was working on "Theme Park," Jensen Huang and two other men sat down in a Denny's diner and drafted a plan for a 3-D graphics company. Nvidia's founders, like Hassabis, were drawn to the economic prospects of videogames - an industry that offered a rare combination of high sales volumes and voracious appetite for computing power.

Early on, Nvidia struggled to find its footing. The company famously teetered on the edge of bankruptcy in 1996, surviving only through a $5 million cash injection from the Japanese videogame developer Sega. In the following years, Nvidia's fortunes turned around as the company pioneered the modern graphics processing unit. GPUs, with their parallel processing capabilities, could render vast swaths of pixels far better than central processing units, which can only carry out tasks sequentially.

A wave of consumer adoption powered Nvidia to an initial public offering in 1999. Nvidia scored a $200 million contract in 2000 from Microsoft to develop the hardware for Xbox consoles. Gamers, eagerly snapping up graphics cards, were the reason Nvidia endured.

The AI pivot

(MORE TO FOLLOW) Dow Jones Newswires

05-23-26 0830ET