Down arrow button icon

I’m on the mend today, having just returned from several days in northeast Louisiana visiting the site of Meta’s massive AI data center, known as Hyperion, for a feature story I’m reporting on.

I’ve been looking through the thesaurus, trying to come up with the right word to describe how big, noisy, and chaotic this construction site is. enormous? Mammoth? Sprawling? Let’s put it this way: It only takes a while to drive the length of the site – it stretches about five miles from top to bottom.

During that journey, I discovered that Hyperion had become a bigger company. Although the expansion had long been suspected and an open secret among some locals, I confirmed it Meta quietly purchased approximately 1,400 additional acresan area roughly twice the size of Manhattan’s Central Park, is adjacent to its massive 2,250-acre campus. I also noticed active construction going on on the newly acquired land – when I wasn’t worrying about being bombarded by the endless parade of 18-wheelers transporting materials in and around the site.

Possible Amazon-OpenAI deal to power Alexa

Now that I’m back in One Piece, I want to move on to a different piece of news that caught my attention this week. like Amazon OpenAI is weighing a tens of billions of dollars in equity investment in OpenAI, and is reportedly in talks to use OpenAI models to power some of its internal AI products, including the Alexa voice assistant. Transaction first I mentioned by Informationwill include OpenAI employees who help customize the models to Amazon’s needs.

This news comes just one day after Amazon finally made its Alexa+ AI assistant available to everyone in the US, nearly a year after its initial launch. I attended a colorful unveiling in New York City last February, when the company pitched the service as an improved version of the original, 11-year-old Alexa — a version that can handle multiple inquiries at once and act as an “agent,” taking actions on your behalf like booking a repairman or ordering an Uber car.

But even recently, it was beta testing Express a lot of frustration. “When I ask Alexa to turn off the light, it should turn off the light — not everything on the bar,” one software engineer wrote last October in an internal channel seeking feedback on unreleased Alexa+ features. “This turned off the power strip of my aquarium filter and killed my fish.”

Other testers complained that the assistant was talking nonstop, ignoring repeated commands to be quiet, or playing music at the highest level when no one was home.

Alexa’s AI has been a long, slow journey

I’ve been following Amazon’s Alexa’s journey closely since the dawn of post-ChatGPT generative AI in 2023. That September, Amazon held another splashy event — this time at its second headquarters in Washington, D.C. — where David Limb, the company’s then head of devices and services, previewed the new AI-powered generative Alexa by saying, “Alexa, let’s talk.”

But after about a year, I mentioned— based on interviews with more than a dozen former employees who worked on Alexa’s AI — found that the organization was plagued by structural dysfunction and technological challenges that repeatedly delayed shipping of new-generation Alexa AI. These former employees painted a picture of a company lagging behind its big tech rivals like Google, Microsoft and Meta in the race to deploy chatbots and AI agents, and struggling to catch up.

They confirmed that the demo in September 2023 was just a demo. The big language model at the heart of the new Alexa, which Amazon positioned as a competitor to OpenAI’s ChatGPT, was, according to former employees, far from state-of-the-art. Research scientists said Amazon lacks the data and specialized computing infrastructure needed to train and employ master’s degree holders at scale.

Since that report, Amazon He has transformed To a more hybrid approach. Alexa+ is now powered by a mix of Amazon’s own Nova models and models from Anthropic, the AI ​​startup in which Amazon has invested $8 billion.

However, problems remain. Can OpenAI really solve Alexa’s long-standing problems? Or will the Amazon-OpenAI deal distract OpenAI at a time when it is locked in fierce competition with Google and Anthropic? Looming on the horizon is Apple, whose deal with Google to power Siri has complicated an already crowded AI landscape.

Most important of all, the reported conversations reveal how desperate even the biggest players have become to stay ahead in the brakeless AI race.

With that said, here’s more AI news.

Sharon Goldman
sharon.goldman@fortune.com
@sharonggoldman

Fortune on artificial intelligence

OpenAI announces Frontier, an enterprise AI agent platform to run applications like Salesforce and Workday, but could it eventually replace them? – Written by Sharon Goldman

EXCLUSIVE: Lawhive, a startup using AI to reimagine the general law firm, raises $60M in new venture capital funding – Written by Jeremy Khan

Meta’s Hyperion AI data center will span four times the size of Central Park in Manhattan – Written by Sharon Goldman

Palmer Luckey says AI will make hardware so cheap that you’ll be able to buy a “Ford F-150 for $1,000.” – Written by Jake Angelo

EXCLUSIVE: Former Cohere CEO Sarah Hooker raises $50M for AI startup Adaption Labs, a bet on smaller, smarter models – Written by Jeremy Khan

Artificial intelligence in the news

The defeat of artificial intelligence that has finally stunned the market. this Bloomberg This article has highlighted the fact that after years of AI-fueled volatility, this week’s market rout stood out both for its speed and the fear behind it: not that AI is a bubble, but that it is already starting to replace entire business models. Hundreds of billions of dollars were wiped off global stocks and credit in just two days, with software companies being the hardest hit, as investors reacted to signs that AI tools were moving from promise to practice – sparked in part by Anthropic’s new legal product, and reinforced by wider concern across the sector. Even potential AI winners showed nervousness, with Alphabet warning of rising AI spending and Arm Holdings failing to meet revenue expectations. While companies like Salesforce and ServiceNow have yet to report losing customers to AI-native competitors like Anthropic and OpenAI, the disappointing adoption of some internal tools — like Microsoft’s Copilot — has deepened concerns that AI leaders may overtake SaaS incumbents faster than expected.

Google is on its way to success thanks to artificial intelligence. A look at Google’s results for the fourth quarter of 2025 shows that its AI strategy is paying off. the New York Times I mentioned Gemini’s models have outperformed competitors, helping boost engagement on YouTube through recommendations and increasing search traffic by delivering concise, AI-generated answers, momentum that lifted Alphabet’s quarterly revenue to $113.8 billion, up 18% year-over-year, and pushed annual sales above $400 billion for the first time. Profits jumped 30% in the December quarter, but the gains come with a trade-off: Google plans to nearly double capital spending this year to as much as $185 billion as it races to build AI data centers to support Gemini at scale, underscoring how the clear winners in AI are trading squeezed profit margins for dominance in an increasingly capital-intensive race.

Elon Musk is betting his business empire on artificial intelligence. I enjoyed this article on Economistwhich focused on how Elon Musk doubled down on his AI efforts through a sweeping and risky restructuring of his business empire. His plan to merge SpaceX with xAI, which values ​​the combined entity at about $1.25 trillion, aims to give xAI an advantage by pairing it with SpaceX’s launch capabilities and even pursuing the far-fetched idea of ​​orbiting data centers. But the deal also links a highly profitable space company to a money-burning AI lab that lags behind competitors like OpenAI, while sitting on massive debt and regulatory risks inherited from Musk’s social platform The bet is stark: If AI proves as transformative as Musk believes, the gamble could cement its dominance; If not, he may be stretching his “Elon support” – the idea that he will personally intervene if one of his companies gets into trouble – to its limits.

Hollywood is losing audiences due to AI fatigue. I’m sure of this Wired The story strikes a chord, arguing that Hollywood’s obsession with malicious and rogue AI — from Metropolis to HAL 9000 and Skynet — has distorted how the public understands AI today. Rather than autonomous superintelligences plotting the downfall of humanity, the real dangers AI faces are far more mundane and immediate: sloppy automation, degraded creative work, misinformation, and systems deployed widely without accountability. The author suggests that these science fiction tropes distract from how AI is currently reshaping culture and business—not through dramatic rebellion, but through scale, speed, and economic pressure—inflaming a public backlash against what many see as “low-quality AI” and hollow innovation. In other words, the risk is not that AI will turn evil; Rather, it is actually reshaping media and creativity in ways that people neither like nor would choose.

Eye on the AI ​​numbers

7

That’s at least the number of Super Bowl ads bought this year either by AI companies, companies selling AI products, or companies touting that the ads themselves were created with AI.

In addition to OpenAI running an ad, its competitor Anthropic is AIIt launched its first commercial at the Super Bowl about its Claude system and took a swipe at OpenAI over its addition of ads to ChatGPT. Rule 44, a The AI ​​app building platform is also powering its first Super Bowl location.

Meanwhile, Meta is pitching its AI-powered glasses in an ad, and Amazon Ring is debuting an AI-powered feature to find lost dogs. Wix is ​​a website builder too Run an ad for Weeks Harmonyits own AI design platform.

Finally, vodka brand Svedka is promoting its Super Bowl ad using AI-generated creative.

You have a calendar

February 10-11: Artificial Intelligence Action Summit, New Delhi, India.

February 24-26: International Association for Safe and Ethical Artificial Intelligence (IASEAI), UNESCO, Paris, France.

March 2-5: Mobile World Congress, Barcelona, ​​Spain.

March 16-19: Nvidia GTC, San Jose, California.

April 6-9: HumanX, San Francisco.