Google‘s corporate parent on Monday announced an agreement to buy data center energy specialist Intersect for $4.75 billion as part of its effort to secure the vast amounts of electricity needed to power artificial intelligence technology.
Alphabet, which depends on Google’s search engine and other online services for most of its revenue, is buying out Intersect in its entirety after purchasing a stake in the San Francisco-based startup a year ago. Intersect had previously raised $2.1 billion from Google and other early investors.
After the acquisition is completed during the first half of next year, Alphabet intends to allow Intersect to operate independently while continuing its efforts to secure the energy sources needed for the sprawling data centers being built to power the AI services rolling out from Google and rivals such as OpenAI.
The power-hungry data centers, which are being likened to AI factories, are fueling a backlash in communities where residents have been interpreting their steadily rising electricity bills as evidence that they are helping foot the boom for a technology boom that could still go bust.
Intersect focuses on finding ways to increase and diversify the supply of energy in an effort to keep hold down power costs. It’s already working on a data center for Mountain View, California-based Alphabet in Haskell County, Texas.
Google’s inroads with AI have helped increase Alphabet’s stock price by more than 60% so far this year, generating an additional $1.4 trillion in shareholder wealth.
Barnes & Noble is on a roll, and its not slowing down in 2026.
The bookstore has confirmed plans to open 60 new locations across the country in 2026. For Barnes & Noble, the move is part of a major comeback thats been years in the works.
After the company saw a sales peak in 2008, it spent a decade watching sales dwindle as customers moved to other retailers (especially digital booksellers like Amazon), ultimately leading the retailer to shutter 150 stores.
In 2019, the company was acquired by the UK’s Elliott Advisors and taken private.
But since 2023, Barnes & Noble has annually added dozens of new locations across the U.S. A post-pandemic boom in the bookstore economy and an influx of new readers from online communities like TikToks #BookTok have helped the store crawl its way back to bookselling dominanceand this latest announcement shows that the growth is more than a passing fad.
The retailer’s latest expansion plans were reported earlier by USA Today.
Where will the new Barnes & Noble locations be?
In a statement to Fast Company, a Barnes & Noble spokesperson confirmed that the company is looking to add 60 new stores before June 2026 across 10 locations, though the representative declined to share specific addresses or opening dates.
The locations include bookstores in the following states:
Ohio
Texas
Florida
Illinois
Colorado
Washington state
California
Virgina
Georgia
D.C.
How #BookTok made reading cool
This new expansion plan is on par with Barnes & Nobles annual growth levels over the past few years.
In 2023, it opened around 30 new stores, followed by 61 in 2024. According to the spokesperson, 2025 was the company’s most impressive year yet, notching an additional 67 stores.
Barnes & Nobles newfound success is part of a larger return to the stacksand return to reading as a hobby, more generallypowered, in large part, by readers on TikTok.
As younger communities discover popular reads through their #BookTok recommendations, everyone from retailers to libraries are seeing a rise in readership.
Back in February, a Barnes & Noble spokesperson told Fast Company, Since the rise of BookTok during the pandemic, bookstores have seen a significant surge in popularity, especially among young people. Our stores have become popular social spots, offering an experience that online shopping simply cant match.
BookToks popularity is having ripple effects in other areas, too. Some public institutions, like the Milwaukee Public Library and Columbus Metropolitan Library, are riding this wave by building their own presences directly on TikTok.
And companies like the book tracking app Goodreads and publisher Harlequin have both recently adopted new branding thats designed to appeal to the TikTok crowd.
While BookToks effect on booksellers and libraries mightve initially seemed like a brief phase, Barnes & Noble’s consistent expansion and strong sales appear to signal that reading is officially cooland profitableagain.
If youve sold equity in your company at some point, youre probably used to keeping track of your cap table and knowing which shareholders have how big of a stake.
But how would you feel about the U.S. government taking a sizable chunk of your business too?
Thats the increasingly common reality for a growing number of U.S. companies, a new analysis by The New York Times findsand for all the chatter over the summer about the feds moving to own almost 10 percent of Intel, computer chips arent the only sector seeing increased governmental involvement.
Instead, its the minerals, metals, and mining industry that comes up again and again in the Times breakdown of where the Trump administration is making money moves. Included on that list are investments made by the Department of Defense ($400 million in the mining company MP Materials; $35.6 million in Trilogy Metals; $80 million in mineral refiner ReElement Technologies; $620 million in rare earth magnet manufacturer Vulcan Elements) as well as the departments of Commerce (another $50 million in Vulcan Elements) and Energy ($182 million in deferred debt service with Lithium Americas).
Thats not to mention the governments golden share in U.S. Steel, its option to take a future 8 percent stake in the nuclear power company Westinghouse, or of course, the 9.9 percent stake in Intel, priced at $8.9 billion, that came out of the deal this summer.
The companies in which Trump is pushing for federal investment tend to be those deemed vital to national security, which is no surprise when you look at the sectors being prioritized: metals and mining, nuclear power, semiconductors.
If business-as-usual policies worked, America would not be reliant on foreign countries for critical minerals, semiconductors, and other products that are key for our national and economic security, an administration spokesperson said.
Still, its an unusual move by the typically free market-focused Republicans, with the Times estimating that over $10 billion in taxpayer money has already been dedicated to these investments (with more expected to be on the way).
Some officials are hopeful the equity stakes will generate a windfall for taxpayers, but the likelihood of that is unclear, the report reads. Many of the companies are facing financial headwinds, and some could take years to become profitable.
Critics have also raised concerns that these investments could create room for corruption and backroom dealings. The Times analysis did not look at stakes taken by the governments Development Finance Corporation, which more typically invests in private firms.
Brian Contreras
This article originally appeared on Fast Companys sister publication, Inc.
Inc. is the voice of the American entrepreneur. We inspire, inform, and document the most fascinating people in business: the risk-takers, the innovators, and the ultra-driven go-getters that represent the most dynamic force in the American economy.
Gold and silver prices rose to record highs in early trading on Monday, spurred on by a confluence of a few different political events and economic factors, including tensions over the U.S. seizing of possibly another oil tanker from Venezuela, speculation of future Federal Reserve rate cuts, overall economy insecurity, and bets on U.S. monetary policy in 2026.
In addition, there is the issue of increasing U.S. debt and increasing demands of AI. “The continued strength in gold is part of the so-called ‘debasement trade,’ where investors are seeking a hedge against the unprecedented global G7 debt, Ben McMillan, CEO of IDX Advisors, told Fast Company. For silver, we have the added element that roughly 70% of silver is used for industrial purposes, which recently has been increasing due to increased demand for data centers, solar panels and electronic vehicles (all of which use silver).”
In 2025, gold has surged by nearly 70%, according to Bloomberg.
Here’s what to know.
What happened today?
At the time of this writing, in midday trading on December 22, gold bullion (GC=F) was up over 1.9% to $4,472.20 and silver (SI=F) was up about 3.4%, heading toward $70 an ounce, making both headed for highest levels since 1979, per Bloomberg.
How did the Fed rate cut affect gold and silver prices?
Earlier this month, the Federal Reserve cut interest rates by 25 basis points, the third cut in 2025, driving federal rates to their lowest in three years and pointing to possible further cuts in the year ahead.
Historically, gold prices have risen following U.S. Federal Reserve interest rate cuts, with notable gains of 31% in 2000, 39% in 2007, and and 26% in 2019 (all within 24 months).
That’s because those rate cuts are often an attempt to jump-start slowing economic growth, and gold is considered a safe haven in times of economic uncertainty.
In this case, gold became more appealing as bonds and other fixed-income investments yielded lower investments, according to Investopedia.
Will gold hit $5,000 an ounce in 2026?
Now the question is, will gold break the $5,000-an-ounce threshold next year?
Nobody can say for sure, but as the bullion inches upward, Goldman Sachs is predicting gold will hit $4,900 an ounce by December 2026, and oil prices will decline to a low mid-year.
Meanwhile, McMillan notes, “we’re seeing a structural shift in the demand curves for both [gold and silver] that could result in sustained price strength for longer than many expect.”
The white dome of Boudhanath rises like a silent guardian over the chaotic sprawl of Nepals capital, Kathmandu, crowned by a golden spire that pierces the sky. Painted on each of the spires four sides are the benevolent eyes of the Buddha wide, calm, and unblinking said to see all that unfolds below.
Those eyes have served as a symbol of sanctuary for generations of Tibetans fleeing the Chinese crackdown in their homeland. But today, Tibetan refugees are also watched by far more malevolent eyes: Thousands of CCTV cameras from China, perched on street corners and rooftops to monitor every movement below. This intense surveillance has stifled the once-vibrant Free Tibet movement that had resonated around the world.
Nepal is just one of at least 150 countries to which Chinese companies are supplying surveillance technology, from cameras in Vietnam to censorship firewalls in Pakistan to citywide monitoring systems in Kenya. This technology is now a key part of Chinas push for global influence, as it provides cash-strapped governments cost-effective, if invasive, forms of policing turning algorithms and data into a force multiplier for control.
The irony at the heart of this digital authoritarianism is that the surveillance tools China exports are based on technology developed in its greatest rival, the United States, despite warnings that Chinese firms would buy, copy, or outright steal American designs, an investigation by The Associated Press has found.
For decades, Silicon Valley firms often yielded to Beijings demands: Give us your technology and we will give you access to our market. Although tensions fester between Washington and Beijing, the links between American tech and Chinese surveillance continue today.
For example, Amazon Web Services offers cloud services to Chinese tech giants like Hikvision and Dahua, assisting them in their overseas push. Both are on the U.S. Commerce Departments Entity List for national security and human-rights concerns, which means transactions with them are not illegal but subject to strict restrictions.
AWS told AP it adheres to ethical codes of conduct, complies with U.S. law, and does not itself offer surveillance infrastructure. Dahua said they conduct due diligence to prevent abuse of their products. Hikvision said the same, and that they categorically reject any suggestion that the company is involved in or complicit in repression.
Chinese technology firms now offer a complete suite of telecommunications, surveillance, and digital infrastructure, with few restrictions on who they sell to or how theyre used.
China pitches itself as a global security model with low crime rates, contrasting its record with the United States, said Sheena Greitens, a political scientist at the University of Texas at Austin.
Its got a set of solutions that its happy to share with the world that nobody else can offer, she said. (But) theyre certainly exporting the tools and techniques that are very important to authoritarian rule.
The AP investigation was based on thousands of Nepali government procurement documents, corporate marketing material, leaked government and corporate documents, and interviews with more than 40 people, including Tibetan refugees and Nepali, American, and Chinese engineers, executives, experts and officials.
While thousands of Tibetans once fled to Nepal every year, the number is now down to the single digits, according to Tibetan officials in Nepal. In a statement to AP, the Tibetan government in exile cited tight border controls, Nepals warming ties with China and unprecedented surveillance as reasons for the drastic plunge.
A 2021 internal Nepali government report, obtained by AP, revealed that China has even built surveillance systems within Nepal and in some areas of the border buffer zone where construction is banned by bilateral agreements. In a statement to AP, the Chinese Ministry of Foreign Affairs denied coercing Western companies to hand over technology or working with Nepal to surveil Tibetans, calling it a sheer fabrication driven by ulterior motives.”
Attempts to use Tibet-related issues to interfere in Chinas internal affairs, smear Chinas image, and poison the atmosphere of China-Nepal cooperation will never succeed, the statement said.
The Nepali government and the Chinese-controlled Tibetan authorities did not respond to requests for comment.
Under pressure, many Tibetans are responding the only way they can: Leaving. The Tibetan population in Nepal has plunged from over 20,000 to half that or less today.
Former activist Sonam Tashi gave up protesting years ago. Now 49, today he’s just a father trying to get his 10-year-old son out before the net pulls tighter. The boy was born in Nepal but has no document proving he is either a refugee or a citizen, a result of Chinese pressure.
Tashi described how those considered likely to protest are picked up in advance around key dates like March 10, which marks the 1959 Tibetan uprising, or July 6, the Dalai Lamas birthday. In 2018, Nepal’s police magazine confirmed that it was building predictive policing, which allows officers to watch peoples movements, identify in advance who they think will protest, and arrest them preemptively.
There are cameras everywhere, Tashi said, sitting on a bus winding toward the Indian border. There is no future.
They gave us all the hardware
After China crushed a Tibetan uprising in 1959, thousands fled across the Himalayas to Nepal, carrying only what they could: Religious paintings, prayer wheels, and the weight of families left behind.
Their exodus, led by the charismatic Dalai Lama, captured the American imagination, with Hollywood films and actor Richard Geres congressional appeals putting Tibet in the spotlight. Washington trod a careful line, defending the rights and religious freedom of Tibetans without recognizing independence.
Today, the future of the Free Tibet movement is in question. Without refugee cards that grant basic rights, Tibetans in Nepal can no longer open bank accounts, work legally, or leave the country.
Cameras are now everywhere in Kathmandu, perched on traffic lights and swiveling from temple eaves. Most link back to a four-story brick building just a few blocks down from the Chinese embassy, where officers watch the country in real time.
The building hums with the low breath of cooling fans. Inside, a wall of monitors blinks with feeds from border towns, busy mrkets and clogged traffic crossings.
Officers in crisp blue uniforms and red caps sit in the glow, scanning scenes. Beneath the screens, a photo published in a Nepali daily shows, a sign in English and Chinese reads: With the compliments of the Ministry of Public Security of China.
Their reach is vast.
Operators can track a motorbike weaving through the capital, follow a protest as it forms, or patch an alert directly to patrol radios. Many cameras are equipped with night vision facial recognition and AI tracking able to pick a single face out of a festival crowd or lock onto a figure until it disappears indoors. The system not only sees but is learning to remember, storing patterns of movement, building a record of lives lived under its gaze.
A 34-year-old Tibetan cafe owner in the city watched the city change in quiet horror. Now you can only be Tibetan in private, he said. He and other Tibetans in Nepal spoke to AP anonymously, fearing retaliation.
The first cameras in Boudhanath were installed in 2012, officially to deter crime. But after a Tibetan monk doused himself in petrol and set himself ablaze in front of the stupa in 2013, police added 35 night vision cameras around it.
The Chinese embassy in Kathmandu worked closely with the police, said Rupak Shrestha, a professor at Simon Fraser University in Canada who studied surveillance in Nepal. He said the police received special training to use the new cameras, identify potential symbols associated with the Free Tibet movement, and anticipate dissent.
In 2013, a team of Nepal Police officers crossed the northern border into Tibet for a seemingly straightforward mission: Collect police radios from Chinese authorities in Zhangmu, a remote border town, about 120 kilometers (75 miles) from Kathmandu. A truck was loaded with equipment and a few handshakes later, they were driving back to Kathmandu.
The radios made by the partly state-owned Chinese firm Hytera looked like walkie-talkies but ran on a digital trunking system, a scaled-down mobile network for police use. Officers could talk privately, coordinate across districts, even patch into public phone lines. The entire system radios, relay towers, software was a $5.5 million gift from China.
They didnt give us the money, recalled a retired Nepali officer who made the trip. They gave all the hardware. All Chinese.
He remembered not the border guards but the tech sleek, reliable, and far ahead of anything theyd used before. He spoke on condition of anonymity to describe sensitive internal discussions.
He said Nepal had initially considered buying the technology from the U.S. and only wanted to deploy the system in its two biggest cities. Hytera was a fraction of the cost and performed comparably, but China also wanted coverage near the border with Tibet. Nepal acquiesced.
They installed the technology in Sindhupalchowk, a border district with a key road to China used by Tibetan refugees. We understood their mindset, the retired officer said. A secure border.
A police envoy from the Chinese embassy began making regular visits to the Nepal Police headquarters. Hed chat over coffee, flip through brochures from Chinese companies. Hed say, You want anything? the retired officer recalled.
China began donating tens of millions in police aid and surveillance equipment, including a new school for Nepals Armed Police Force. Hundreds of Nepali police traveled to China for training on policing and border control, according to Chinese government posts.
Ahead of a summit of South Asian leaders in 2014, among the goods on offer were ones from Uniview, Chinas pitch for an all-seeing eye.
The company was the Chinese surveillance business of what was then Hewlett Packard, or HP, before it was spun off in a 2011 deal. Since 2012, Uniview has been selling mass surveillance solutions to the Tibetan police, such as a command center, and developed cameras that track ethnicities such as Uyghurs and Tibetans.
Uniview installed cameras in Kathmandu for Nepals first safe city project in 2016. It started with the citys roads, then went up across the capital in tourist areas, religious sites, high-security zones like Parliament and the prime ministers home.
The cameras didnt just record. Some could follow people automatically as they moved. Others were designed to use less data, making it easier to store and review footage.
Hewlett Packard Enterprise, or HPE, a successor company to HP that sells security solutions, has no ownership in Uniview and declined to comment. Hytera and Uniview did not respond to requests for comment.
Nearly all the cameras installed in Nepal are now made by Chinese companies like Hikvision, Dahua, and Uniview, and many come bundled with facial recognition and AI tracking software.
Hikvision’s website and marketing materials advertise camera systems in Nepal linked via Hik-Connect and HikCentral Connect, cloud products that rely on Amazon Web Services. Hikvision sells to the Nepali police and government, and a template for Nepali tenders indicates CCTV cameras procured for the government are required to support Hik-Connect.
In return for Beijings support, top Nepali officials have thanked China repeatedly over the years, promising never to allow anti-China activities on Nepali territory.
The Nepali police head offices arent far from the now-forlorn Tibetan reception center, which used to shelter tired, hungry Tibetans fleeing across the border.
The building is nearly empty. The gates are locked. Those who do escape, like Namkyi, arrested at 15 for protesting Chinese rule, often have to wait for weeks confined indoors until theyre smuggled out again to the Tibetan capital in exile in India.
Silence has become survival.
They know they are being watched, she said. Even though we are free, the surveillance cameras mean were actually living in a big prison.
From clients to competitors
From the start, U.S. companies eager for Chinas vast markets exchanged technology for enry.
Many were required to start joint ventures and research operations in China as a precondition for being allowed in. Dozens, if not hundreds, complied, transferring valuable know-how and expertise even in sensitive areas like encryption or policing.
Little by little, Chinese companies chipped away at the lead of American tech companies by luring talent, obtaining research, and sometimes plain copying their hardware and software. The flow of technology continued, even as U.S. officials openly accused China of economic espionage and pressuring American companies for their technology.
China is by far the most egregious actor when it comes to forced technology transfer, Robert D. Atkinson, then-president of a think tank focused on innovation, warned Congress in a 2012 hearing.
American tech resistance came to a final, definitive end later that year with Edward Snowdens revelations that U.S. intelligence was exploiting American technology to spy on Beijing. Spooked, the Chinese government told Western firms they risked being kicked out unless they handed over their technology and provided security guarantees.
After companies like HP and IBM agreed, their former partners became their fiercest global competitors and unlike American firms, they faced few questions about the way their technology was being used. Companies like Huawei, Hikvision, and Dahua have now become global behemoths that sell surveillance systems and gear all over the world.
American technology was key to this:
– Uniview, the Chinese AI-powered CCTV camera supplier, supplied the first phase of Nepals safe city project in 2016, installing cameras in Kathmandu. Uniview was carved out of California-based HPs China surveillance video business.
– Hytera provided data infrastructure for the Nepali police, such as walkie-talkies and digital trunking technology, which enables real-time communication. Earlier this year, Hytera acknowledged stealing technology from U.S. company Motorola in a plea agreement, and had acquired German, British, Spanish, and American tech businesses in their growth phase.
– Hikvision and Dahua, Chinas two largest surveillance camera suppliers, sell many of the cameras now in Nepal. They partnered with Intel and Nvidia to add AI capabilities to surveillance cameras. Those ties ended after U.S. sanctions in 2019, but AWS continues to sell cloud services to both companies, which remains legal under what some lawmakers call a loophole. AWS has advertised to Chinese companies expanding overseas, including at a policing expo in 2023.
– Chinese tech giant Huawei has become one of the worlds leading sellers of surveillance systems, wiring more than 200 cities with sensors. In Nepal, they supplied telecom gear and high-capacity servers at an international airport. Over the years, the company benefited from partnerships with American companies like IBM, and has been dogged by allegations of theft including copying code from Cisco routers wholesale, a case which Huawei settled out of court in 2004.
Huawei said it provides general-purpose products based on recognized industry standards. Intel has said it adheres to all laws and regulations where it operates, and cannot control end use of its products. Nvidia has said it does not make surveillance systems or work with police in China at present.
IBM and Cisco declined to comment. Policing gear maker Motorola Solutions, a successor company to Motorola after it split, did not respond to requests for comment.
U.S. technology transfer to Chinese firms has mostly stopped after growing controversy and a slew of sanctions in the past decade. But industry insiders say its too late: China, once a tech backwater, is now among the biggest exporters of surveillance technologies on earth.
Few realized the U.S. shouldnt be selling the software to China because they might copy it, they might use it for these types of surveillance and bad stuff, said Charles Mok, a Hong Kong IT entrepreneur and former lawmaker now living in exile as a research scholar at Stanford. Nobody was quick enough to realize this could happen.
The great big eye in the sky
Inside a 15th-century monastery in Lo Manthang in Nepals Mustang district, light slants through wooden slats, catching motes of dust and the faded faces of bodhisattvas.
Crumpled notes of Chinese currency lie at the feet of deities in the walled city along the Tibetan border. Here, shops stock Chinese instant noodles and cars with Chinese plates rumble down mountain roads.
A gleaming white observation dome just inside Chinese territory looms over the city. Visible from 15 kilometers (9 miles) away, its trained on the district that has long been a refuge for Tibetans, including a guerrilla base in the 1960s.
The dome is just one node in Chinas vast 1,389-kilometer (863-mile) border network with Nepal a Great Wall of Steel of fences, sensors, and AI-powered drones.
Chinese forces have barred ethnic Tibetans from accessing traditional pastures and performing sacred rites. They have pressured residents of Lo Manthang to remove photos of the Dalai Lama from shops. And a China-Nepal joint command mechanism meets several times a month on border patrols and repatriations, according to a post by the Chinese-run Tibetan government.
The result is that the once-porous frontier is now effectively sealed, and Chinas digital dragnet reaches deep into the lives of those who live near it.
In April 2024, Rapke Lama was chatting with a friend across the border on WeChat when he received an invitation to meet. He set out from his village and crossed into Tibet only to be arrested almost immediately.
Lama believes his WeChat exchange was monitored; Chinese police appeared with unsettling precision, as if they knew where to look. After accusing him wrongly, he maintains of helping Tibetans flee into Nepal, the police seized his phone, which had photos of the Dalai Lama and Tibetan music. Then came months in a Lhasa prison, where isolation and inadequate medical care hollowed him out.
Lama did not return to Npal until May 2025, gaunt and shaken. He later said he entered Tibet to harvest caterpillar fungus, valued in traditional Chinese medicine. Another friend who crossed the border remains in custody.
Even now, Im scared, Lama says. He wears masks when wandering the streets, he says, because of that lingering fear.
The Chinese observation dome is a giant symbol of the same fear, towering over the border.
Its the great big eye in the sky, said a 73-year-old Tibetan hotel owner in Nepal, who spotted the installation during a trip near the border last year. For Tibetan refugees, Nepal has become a second China.
__
Associated Press journalists Niranjan Shrestha and Binaj Gurubacharya in Kathmandu, Manish Swarup and Rishi Lekhi in New Delhi, Ashwini Bhatia in Dharamshala, India, and David Goldman in Washington contributed to this report.
Aniruddha Ghosal and Dake Kang, Associated Press
Just when you think youve wrapped your mind around computers that can put your dog in front of the Eiffel Tower or chatbots that act like your best friend (or lover), the AI behemoths surprise you with a fully AI-powered TikTok or the ability to virtually bring back your dead relatives.
Ive worked in the AI space for 15 years. I served as an early beta tester for OpenAI in 2020, when I predicted that a little model called GPT-3 had world-changing potential.
It was later released as something called ChatGPTperhaps youve heard of it?
Ive also called several big AI trends correctly, including the rise of video generators and the AI Wars between Google and OpenAI.
Based on my experience, here are my six AI predictions for 2026 and beyond.
1. OpenAI goes garlic
In late 2025, Googles Gemini model started to gain ground on OpenAI and its GPT-5.1 system.
That apparently really irked Sam Altman and the OpenAI team. Altman reportedly called a code red, directing staff to focus all their efforts on besting Google.
Rumor had it they were developing a new, fully re-trained thinking model, codenamed Garlic. When OpenAI did a surprise drop of a new GPT-5.2 model in mid December, lots of people thought it might be Garlic coming to market.
Based on my own testing, its not. Or at least, its not the complete model.
GPT-5.2 is indeed better than the previous model. Its faster and more efficient, and makes fewer errors. Its also notably better at scientific tasks, and practical ones like coding.
But now it appears there will be a second new OpenAI release, which I expect to come out in January. Thats most likely the full Garlic model.
This new model, I predict, will have a new knowledge cutoff sometime in 2025, a broader context window, and much better image generation capabilities. It will also be faster and more efficient to run, especially on thinking tasks.
2. Googles Gemini continues its march toward domination
Whenever it finally arrives, Garlic will enter the world with plenty of competition.
Google was very slow to roll up to the generative AI table. For a company thats been working in deep learning for decades and has some of the most intelligent people in the world working for it, that felt like a big miss.
Google had reportedly developed its own ChatGPT years before OpenAI, but chickened out on releasing it. In the beginning of the AI race, that allowed OpenAI to very loudly and publicly eat Googles lunch.
The history of science, though, is littered with examples where early innovators werent the ones who successfully commercialized a new technology.
Just ask Joseph Swan, the true inventor of the lightbulb. Youve never heard of him. But you do know Thomas Edison, who made the lightbulb a widely available technologyand did a great job promoting his invention (and himself) in the process.
Historically, first-mover advantage has proven surprisingly inconsequential in the tech space. And now that Google has woken up to the importance of AI, theyre aggressively building out their Gemini model and integrating it into almost all their products, including their core search experience.
Google has more data, more resources (including its own custom AI chips), more people, and a much broader reach than OpenAI.
In 2026, Google will continue to throw its weight around in generative AI. Gemini will go from being an also-ran to one of the most powerful models on the market. Because it will be broadly integrated into products that normal people use on a day-to-day basis, it will immediately have an audience in the billions.
The struggle now isnt for newer companies like OpenAI to create the best product. Its to create a product better than Googles. That will be very hard in 2026 and beyond.
3. Chatbots become therapy (and a bit more)
Users have already realized that ChatGPT can take the place of a human therapist. In a recent poll by the Economist, 25% of people reported turning to chatbots for mental health support.
As cases of AI psychosis and alleged suicides demonstrate, this can go very badly. But for people who cant afford any kind of psychological supportor simply dont have access to it in their language or countryusing chatbots as cheap therapists is incredibly appealing.
Without directly saying so, OpenAI has implied that theyre moving ChatGPT further into this space, with improvements to how the bot handles sensitive medical and mental health conversations.
This could be a huge boon for mental health. Many people appear more comfortable discussing their problems with an unthinking bot than with a human. The fact that a session with ChatGPT doesnt cost $300 per hour is also a big plus!
ChatGPTs capabilities will expand in other ways, too. A rumored adult mode will arrive in 2026, allowing ChatGPT to write risqué material. Prepare yourself for a wave of frenzied op-eds about how people are turning to these newly salacious for relationships instead of fellow humans.
4. AI-generated videos take overand not just on Sora
OpenAIs AI-powered Sora video generator is incredibly powerful, and their Sora-based social network is incredibly fun to us.
In 2026 and beyond, expect to see the reach and importance of AI-generated vertical video accelerate.
Vertical video is the perfect format for AI. The clips tend to be short, which caters to AIs ability to generate about 10 seconds of reliable video before things go off the rails.
They also tend to be grabby and compelling. Again, AI excels at making videos of things like people falling into wedding cakes or having heated arguments with their roommates.
In 2026, expect the reach of AI-powered social networks like Sora to grow dramatically.
The biggest growth, though, will come from these videos migrating off the Sora platform and onto other social media. Already, my Facebook Reels feed is dominated by clearly AI-generated videos of things like a cat saving her kittens from a flood or grandmothers fighting grizzly bears.
In 2026, AI videos wont stay put. Theyll travel into every vertical video space on the webfrom TikTok to Nextdoorfurther blurring the lines between whats real and whats imagined.
5. Electricity becomes the limiting factor
I have friends who build data centers for a living. They tell me the only thing stopping them from building more data centers is finding enough electrical power to keep up with AIs demands.
Some companies are reportedly even going nuclear, building or recommissioning fully functional reactors to power their electricity-hungry AI chips.
The need for more electrical power for AI will start to limit the techs growth in 2026. It will also rub up against societys other needs.
In 2026, I expect a populist backlash against the fact that data centers voracious energy demands are raising electricity rates for everyday people.
Ultimately, the deficiencies of todays gridstrained as it is by the rise of AIwill drive innovative, new models that are good for everybody. Cheap solar power at midday may be redirected toward data centers, for example, or stored in giant batteries to keep servers running overnight.
This demand will create a huge market for green technologies, ultimately benefiting the planet and everyone on it.
6. AI invades the real world
No, the robot uprising isnt here just yet. But AI is increasingly invading the real world.
Self-driving cars were once a novelty. In 2026, usage will explode, with Zoox, Waymo and their competitorsincluding entrants from Chinaserving more cities.
The rise of self-driving vehicles and other physical manifestations of AI technology will surprise people. Youll blink, and one day it will feel like nearly every car on the road is self-drivingas it currently does in my home city of San Francisco.
I expect to see other experiments with physical AI in 2026, from robot baristas to caregiving machines, and plenty of military AI tech, too.
Again, though, the self-driving car blitz coming in 2026 will be the most profound and surprising (to everyday people, anyway) implementation of the technology. It will arrive far sooner than you think.
A pat on the back
So thats what I anticipate for the year ahead. As someone whos been in the AI space for a long time, Ive missed some things.
But I also wrote, two years before ChatGPT, that OpenAI and its founders could easily make billions (and likely challenge the advertising and content recommendation engines of rivals like Google) by throwing caution to the wind and throwing open the doors to GPT-3 to all comers.
I still pat myself on the back for that one.
How will my 2026 predictions shape up? Ask me in a year!
The author of The Art of the Deal always likes to claim hes a big winner when it comes to any business arrangement he makes. And in some ways, Donald Trump appears to have won big by finalizing a deal that will see Oracle, Silver Lake, and MGX take part-ownership of a new joint venture designed to oversee operations in the United States of TikTok, the wildly popular social video appl. But dig into the details and youll see that what Trumps White House is keen to present as a big win for national security looks more like a standard business dealor more cynically, a shakedown.
Concerns around TikTok first bubbled up at the end of Trumps first term, when the 45th president, running unsuccessfully for re-election in 2020, presented the app, owned by the Chinese tech champion ByteDance, as a national-security concern. On both sides of the aisle, China hawks worry that TikToks algorithm could be used to catalyze opposition to the American way of life, and indoctrinate U.S. teens into Chinese ways of thinkingor nudge public opinion to be more favorable to the Chinese regime.
Trump tried unsuccessfully to ban the app outright from the United States, a gambit that didnt stand up in court. He has continued to try and alter its ownership even as he appears to have changed tack about whether the platform ought to be banned outright.
Trump has claimed that TikToks purported Chinese links still gave him pause, but that he was willing to allow the app to continue existing in the U.S. provided that it was brought under American control. With this finalized deal, even thats not guaranteedwhich could suggest this is little more than a shakedown and carve-out to ensure the U.S. capitalizes on the only non-American social network that has managed to gain a mass global foothold in the last 20 years.
The leaked details of this deal seem to imply that the public debate and concerns were a red herring, says Catalina Goanta, associate professor in private law and technology at Utrecht University. The U.S. just wanted in on a profitable business model that has been growing faster and with more potential than any of its competitors, she argues.
The terms of the agreement suggest that the joint venture will own between 45%and 50% of the new U.S.-Tiktok entity (reports differ on the precise percentages involved). Around one-third of the entity will be owned by ByteDances current investors, with the remainderan estimated 20% or sostill under ByteDances control. The deal is due to close by January 22.
Others are equally uncertain that the deal matches up to what Trump claimed was the core concern. Will the sale enrich the new investors or protect American interests? asks Hussein Kanji, founder of Hoxton Ventures. Lets see if the algorithm changes in the new leadership to support a particular political viewpoint.
So far, theres no suggestion that the apps algorithm will change in any way, beyond being fed U.S. user data to ensure the content is free from outside manipulation, said an internal memo sent by TikTok CEO Shou Chew to staff last week. That isnt significantly different from what already happens, except it draws slightly stronger fences around U.S-only users.
The terms of the deal are believed to adopt the current TikTok algorithm, while the storage of user data within the United States will remain within the country and overseen by a local partner. In this case, that partner will be Oracle, under terms similar to those TikTok has already instigated elsewhere voluntarily, including in Europe (U.K. cybersecurity firm NCC Group oversees data access) where TikTok has built dedicated data centers for local users. ByteDance will reportedly still have control of the apps ecommerce, advertising, and marketing arms, all of which are core components of the business.
In short, basically nothing has changed, except several U.S. firms get a part of the new companyand presumably, a share of its income.
Its no surprise, then, that China has nodded the deal through: Little changes for them, except for homegrown champion ByteDance losing a proportion of its income from the app. Chinese state media sees this deal as a win for China, and it emphasized retaining global connectivity, which can also affect what kind of content is seen from outside of the U.S., says Goanta.
Of course, the app could still change. It certainly would be easier to do so when U.S. companies control the data, the algorithm, and any decisions are overseen by a U.S. board. But itd be highly unusualsome might say self-defeatingfor TikTok in the United States to try and diverge from what made its global product successful.
Instead, it looks like a classic Trump deal: Plenty of sound and fury, and a whole lot of hyperbole to justify very few changes that actually address the underlying issue that caused the brouhaha in the first place. The deal allows the president to portray action on a politically potent issue while avoiding a total ban that could alienate younger voters or provoke corporate backlash. For China, the arrangement shows it can be flexible without surrendering, allowing ByteDance to preserve that global reach for its flagship app.
But as often happens under Trumps America First policy, American entities get a cut of the dealwhether theyre deserving of it or not.
As many families are preparing to gather for the holidays, influenza (flu) cases are spiking across the country.
According to the latest data from the Centers for Disease Control and Prevention (CDC), positive test results have reached the highest levels seen so far this season.
The most frequently reported influenza virus this season is the influenza A H3N2 virus. Last week, Fast Company reported on a new mutated strain of influenza A H3N2, known as the subclade K flu variant, which emerged after multiple mutations. Heres what you need to know.
Recent data shows positive cases are spiking
According to CDC data for the week ending December 13, 14.8% of samples tested positive for influenza.
This is the highest level of positive cases so far this season. A total of 927 influenza viruses were reported by public health laboratories.
Of those, 911 were influenza A and 16 were influenza B. Of the 706 influenza A viruses that were subtyped, 89.9% were influenza A H3N2.
For the week ending December 13, states with the highest flu activity include:
Colorado
Louisiana
New Jersey
New York
Rhode Island
Tracking data from the New York State Department of Health shows cases at their highest for the season, with over 5,300 hospitalizations so far.
In New York City, flu cases have spiked significantly.
[Screenshot: via CDC]
Holiday season means data will lag this week
Unfortunately, the CDC wont provide further updates on the spread of the virus until the very end of the year. Data reporting will be delayed due to the Christmas holiday.
According to the CDC, data for the week ending December 20 will be posted on December 30.
What are symptoms of the flu?
Typical symptoms of the flu vary from mild to severe. The CDC says these are the main symptoms to watch out for:
fever or feeling feverish/chills
cough
sore throat
runny or stuffy nose
muscle or body aches
headaches
fatigue (tiredness)
possible vomiting and diarrhea (more common in children than adults)
The agency notes that not everyone with the flu gets a fever.
Total flu illnesses reach 4.6 million nationwide
So far this season, the CDC estimates that at least 4.6 million flu illnesses have occurred. The agency estimates that there have been 49,000 hospitalizations and 1,900 deaths from the flu.
If youre feeling sick this holiday season, you should limit contact with others.
The CDC recommends that everyone 6 months of age and older who has not yet been vaccinated this season receive an annual influenza vaccine. Approximately 130 million doses of the influenza vaccine have been distributed in the U.S. this season.
SoftBank Group is racing to close a $22.5 billion funding commitment to OpenAI by year-end through an array of cash-raising schemes, including a sale of some investments, and could tap its undrawn margin loans borrowed against its valuable ownership in chip firm Arm Holdings, sources said.
The “all-in” bet on OpenAI is among the biggest yet by SoftBank CEO Masayoshi Son, as the Japanese billionaire seeks to improve his firm’s position in the race for artificial intelligence. To come up with the money, Son has already sold SoftBank’s entire $5.8 billion stake in AI chip leader Nvidia, offloaded $4.8 billion of its T-Mobile US stake, and slashed staff.
Son has slowed most other dealmaking at SoftBank’s Vision Fund to a crawl, and any deal above $50 million now requires his explicit approval, two of the sources told Reuters.
Son’s firm is working to take public its payments app operator, PayPay. The initial public offering, originally expected this month, was pushed back due to the 43-day-long U.S. government shutdown, which ended in November. PayPay’s market debut, likely to raise more than $20 billion, is now expected in the first quarter of next year, according to one direct source and another person familiar with the efforts.
The Japanese conglomerate is also looking to cash out some of its holdings in Didi Global, the operator of Chinas dominant ride-hailing platform, which is looking to list its shares in Hong Kong after a regulatory crackdown forced it to delist in the U.S. in 2021, a source with direct knowledge said. Investment managers at SoftBank’s Vision Fund are being directed toward the OpenAI deal, two of the above sources said.
SoftBank’s scramble to marshal funds offers a window into the strain faced even by the world’s biggest dealmakers as they scramble to finance ambitious AI data center projects worth hundreds of billions of dollars.
SoftBank declined to comment.
SOFTBANK HAS OPTIONS
OpenAI has not yet received the remaining funding, but expects the money to come in by the end of 2025, as stipulated in the contract, sources said.
SoftBank has multiple sources of capital it could tap, including margin loans, cash on its balance sheet, stakes in listed companies, and corporate bonds or bridge loans, sources said.
Son has strong reasons to draw on a range of funding mechanisms to fulfill those obligations. SoftBank secured a deal to invest in OpenAI at a $300 billion valuation in April. Since then, the valuation of OpenAI has risen dramatically and the company is in talks to raise additional funding from investors, including Amazon, tripling its valuation to close to $900 billion, one of the sources added, which would give SoftBank a significant paper gain once the transaction is completed.
A major pool of capital for SoftBank is its undrawn capacity of margin loans borrowed against its ownership of British semiconductor and software design company Arm Holdings. SoftBank recently expanded its margin loan capacity by $6.5 billion, bringing the total undrawn capacity to $11.5 billion. Arms stock has since tripled from its IPO price, providing SoftBank with additional collateral headroom to expand its borrowing capacity.
SoftBank reported parent-level cash of 4.2 trillion yen ($27.16 billion) as of September 30. The group still owns about 4% of T-Mobile US, remaining the wireless carriers second-largest shareholder, a stake worth roughly $11 billion at the end of September, according to LSEG data.
Despite investing at a less active pace, it has continued to back AI startups such as Sierra and Skild AI.
OPENAI NEEDS THE MONEY
Both OpenAI and SoftBank are investors in Stargate, a $500 billion initiative to build AI data centers for training and inference that executives say is crucial to the U.S. government’s ambitions to keep ahead of China in AI.
The rush to build data centers has also prompted tech giants including Meta Platforms to commit unprecedented sums to these buildouts – which need chips, power, cooling, and servers – and they have brought in deep-pocketed partners to spread the risk. Their hefty capital outlays have sparked concerns about what happens if the investments fail to bring commensurate returns, raising the specter of an “AI bubble” bursting.
SoftBank promised in April to invest up to $30 billion in OpenAI – $10 billion of which the startup would receive the same month. The rest of the payment was contingent on the AI startup transitioning to a for-profit corporation by the end of the year, an ambitious feat that OpenAI achieved in October.
The new funding is crucial for covering OpenAIs rising costs to train and run its AI models as competition from Alphabet’s Google ratchets up. OpenAI CEO Sam Altman told employees recently that the company is now entering a code red phase to improve ChatGPT delaying other product rollouts to fend off the momentum behind Googles Gemini.
In October, Altman said OpenAI aimed to build 30 gigawatts of computing capacity for $1.4 trillion. He said he ultimately wants OpenAI to add 1 gigawatt of compute every week – an enormous target given that each gigawatt currently comes with a capital cost of more than $40 billion.
(Reporting by Echo Wang in New York, Krystal Hu and Deepa Seetharaman in San Francisco and Miho Uranaka in Tokyo; Editing by Sayantani Ghosh and Matthew Lewis)
Echo Wang, Miho Uranaka and Krystal Hu, Reuters
This year delivered whiplash: geopolitics, tariffs, and technology all shifting at once. And heading into 2026, the disruption isnt easing up. Bob Safian distills hard-won lessons from his Rapid Response podcast this year on how to lead when the ground wont stop movingfeaturing standout moments from Airbnbs Brian Chesky, Runways Cristóbal Valenzuela, Metas Clara Shih, LinkedIns Aneesh Raman, Planned Parenthoods Alexis McGill Johnson, and the NWSLs Jessica Berman, with practical takeaways for turning uncertainty into advantage.
This is an abridged transcript of an interview from Rapid Response, hosted by the former editor-in-chief of Fast Company Bob Safian. From the team behind the Masters of Scalepodcast, Rapid Response features candid conversations with todays top business leaders navigating real-time challenges. Subscribe to Rapid Response wherever you get your podcasts to ensure you never miss an episode.
Lesson number one: As tech moves fast, we need to move even faster
Among the most challenging aspects for leaders is the speed of change and how it requires us to reset our expectations and practices. Here’s the CEO of AI video company Runway, Cris Valenzuela, talking with me about planning in the eye of the AI storm.
How far out do you think of your product roadmap? Or is that something you’re reassessing all the time?
Cris Valenzuela: Yeah, it’s a weekly thing, to be honest. If you’re planning on a quarterly basis, you’re not going to make it. You’re done. In four weeks, you’re going to get leapfrogged and things will change.
We’ve historically taken this open-ended research approach. Instead of defining very specific goals you want to accomplish, you define the boundaries on which you want the team to play and experiment. And then setting the boundaries and the limits is kind of the hard thing because if it’s too open, then there’s nothing really directionally happening. If it’s too broad, then it’s just an objective that’s very clear. If it’s broad enough and has enough of the right incentives, then people are going to stumble on things that are new, that you’ve never thought of before, that have a great value. And those are the things that we care the most.
Valenzuela’s approach is so different from traditional leadership, leaning into experimentation rather than specific goals, and reframing plans on a weekly basis. It’s an approach that could make a lot of people uneasy. I talked about this with Clara Shih, who’s led AI business at Salesforce and at Meta. She offered practical insights about navigating what’s hype and what’s imperative. Here’s me and Shih.
How do leaders strike the balance between I got to be in this, versus it’s not really showing any measurable impact now yet?
Clara Shih: I see this all the time from various leaders that I meet with. I think it’s first being hands-on and really getting in there and understanding the capabilities because I think with that judgment, with that firsthand experience, only then can leaders really know, “Okay, I want to apply it here, but not here.” Another really great success formula is splitting up the team, right? Having people focus on immediate use cases, what can I unlock today that will show me ROI this quarter, next quarter, versus what are the bigger bets where just I see the secular trend and we have to skate to where the puck is going. But just know that it’s going to take longer and more experiments to asymptotically hopefully get to the right answer. And then just having space and time to live in both time horizons simultaneously. What’s the day-to-day? What’s the quarter? How could I be completely screwed in six months or 12 months if I don’t have this tiger team that’s incubating experiments at startup speed?
Lesson two: In an AI world, human connection is a competitive advantage
AI technology is so powerful. But there’s an equally strong thread about emphasizing the human factor within enterprises, that that will truly differentiate the winners. Here’s Brian Chesky, CEO of Airbnb.
Brian Chesky: The term AI, the important term is artificial. We’re going to live in a world where it’s not clear that what you’re seeing is real. And the opposite of artificial is real. The opposite of screen is the real world. People want real connections in the real world. Why are people feeling so lonely right now? Because they were connecting with people they don’t know, arguing with people on the internet and your Instagram followers aren’t coming to your funeral. No one changed someone else’s mind in YouTube comment section. And now pretty soon we’re going to have a situation where your friends are going to be AIs. So there has to be this movement to real.
Chesky’s business at Airbnb, of course, relies on in-person interaction through home stays and experiences, but that doesn’t diminish the leadership implications of what he’s saying. The challenges and opportunities of this age come down to human choices. The choices we make about how we interact with each other defines leaders and organizations, especially because AI is changing how we’re interacting. Here’s LinkedIn’s Chief Economic Opportunity Officer, Aneesh Raman, talking about what he calls the five Cs, the core human skills for this era.
Aneesh Raman: What you’ve got is sort of what I call the five Cs, this list I’ve developed with neuroscientists, courage, compassion, creativity, curiosity, and communication. Those are kind of the core skills I think that make us humans.
Remember, our species until about 40,000 years ago wasn’t the only sapiens around. And we were never the biggest, we were never the fastest. What allowed us to emerge as the apex species on this planet is that we were able to adapt in really important ways by those five Cs in how we both told really complex stories through language and then how we organized to increase scale around things like nation states. So that’s going to come to the center of it all for us and we’ve got to shore those skills up.
Lesson three: The most important decisions are simple and brave
When uncertainty is high, clarity of mission matters more than ever. For some, new pressure served as a valuable reminder of what was most important. Here’s an exchange I had with Alexis McGill Johnson, president of Planned Parenthood, who’s been in the crosshairs of the Trump administration all year.
Alexis McGill Johnson: I feel concern that a number of really critical institutions in our society are feeling a financial pressure to, I think in many ways, go against thir core values. Your values are … That’s your integrity. That’s who you are. So I cannot actually stand here and say, “We’re going to walk away from the very communities that we have committed ourselves to providing care for.”
It sounds like you wish maybe that there was a little more bravery from some other leaders whose, I don’t know, whose missions may not be as clearly values-based all the time.
Two things come to mind here. One is watching people obey in advance, comply in advance before the actual directives come, which I think sends a signal that people are willing to kind of stand down. But I think we’re also missing the collective action here, that there is a logic of collective action that means that when we actually stay kind of arms linked and say, “You know what? We are going to stand with the rule of law and what we believe the Constitution says here.” I think it really is about linking arms and understanding that that is really the kind of strongest attack back in some ways to the kinds of things that we are facing.
When it comes to decision making, I often think of a framework Brian Chesky has talked about, focusing on what he calls principle decisions versus business decisions, choices that you’ll be proud of even if things don’t go your way. It seems pretty simple, but then the most important decisions often are if boiled down to their essence. Jessica Berman, commissioner of the National Women’s Soccer League, keeps a pile of children’s books on the coffee table in her office to remind her team to ground themselves in the basics. Here’s Berman.
Jessica Berman: Every single leadership lesson you need in life, you learned when you were five It’s such a great analog for people to humanize and boil down sometimes hard to talk about or complex concepts that are really interpersonal.
Is there a book that right now, a children’s book that you find yourself going to more?
I have one. We’re Going on a Bear Hunt. Guess what? You can’t go around it. You can’t go over it. You can’t go under it. You just have to go through it. And that is the story of challenges in life, and so we cite that in our office almost every single day.
