• DeepSeek's success has put Silicon Valley on edge about Chinese competition.
• After DeepSeek released its latest model, AI investors panicked.
• Meta's chief AI scientist, Yann LeCun, said the market's reaction to DeepSeek was "unjustified."

Silicon Valley is melting down over DeepSeek, an emerging Chinese competitor in AI, but Meta's AI chief says the hysteria is unwarranted.

DeepSeek caused alarm among US AI companies when it released a model last week that, on third-party benchmarks, outperformed models from OpenAI, Meta, and other leading developers. It did so with subpar chips and, it said, vastly less money.

Bernstein Research found that DeepSeek priced its models significantly below equivalent models from OpenAI: DeepSeek's latest reasoning model, R1, cost $0.55 for every 1 million tokens inputted, while OpenAI's o1 reasoning model charged $15 for the same number of tokens. A token is the smallest unit of data an AI model processes.

The news hit the markets Monday, triggering a tech sell-off that wiped out $1 trillion in market cap. Nvidia — known for its premium chips, which can cost at least $30,000 — lost almost $600 billion.

Yann LeCun, the chief AI scientist for Facebook AI Research, however, says there's a "major misunderstanding" about how the hundreds of billions of dollars invested in AI will be used. In a Threads post, LeCun said the huge sums of money going into US AI companies were needed primarily for inference, not training AI.

Inference is the process in which AI models apply their training knowledge to new data. It's how popular generative AI chatbots like ChatGPT respond to user requests. More user requests means more inference is required, and processing costs increase.

LeCun said that as AI tools become more sophisticated, the cost of inference will rise. "Once you put video understanding, reasoning, large-scale memory, and other capabilities in AI systems, inference costs are going to increase," LeCun said, adding, "So, the market reactions to DeepSeek are woefully unjustified."

Thomas Sohmers, a founder of Positron, a hardware startup for transformer model inference, told Business Insider he agreed with LeCun that inference would account for a larger share of AI infrastructure costs.

"Inference demand and the infrastructure spend for it is going to rise rapidly," he said. "Everyone looking at DeepSeek's training cost improvements and not seeing that is going to insanely drive inference demand, cost, and spend is missing the forest for the trees."

This means that, as its popularity grows, DeepSeek is expected to handle more requests and spend a significant amount on inference.

A growing number of startups are entering the AI inference market, aiming to simplify output generation. With so many providers, some in the AI industry expect the cost of inference to drop eventually.

But this applies only to systems handling inference on a small scale. The Wharton professor Ethan Mollick has said that for models like DeepSeek V3 that provide free answers to a large user base, inference costs are likely to be much higher.

"Frontier model AI inference is only expensive at the scale of large-scale free B2C services (like customer service bots)," Mollick wrote on X in May. "For internal business use, like giving action items after a meeting or providing a first draft of an analysis, the cost of a query is often extremely cheap."

In the past two weeks, leading tech firms have stepped up their investments in AI infrastructure.

Meta CEO Mark Zuckerberg announced more than $60 billion in planned capital expenditures for 2025 as the company ramps up its own AI infrastructure. In a post on Threads, Zuckerberg said the company would be "growing our AI teams significantly" and had "the capital to continue investing in the years ahead." He did not say how much of that would be devoted to inference.

Last week, President Donald Trump announced Stargate, a joint venture between OpenAI, Oracle, and SoftBank set to funnel up to $500 billion into AI infrastructure across the US.

• DeepSeek, an open-source Chinese AI company, has riled Silicon Valley with its rapid rise.
• Meta's chief AI scientist said DeepSeek has benefited from the open-source community.
• Meta's AI program has remained open-source, while OpenAI has shifted to closed-source.

Silicon Valley was on edge this week after DeepSeek, a Chinese AI company, released its R1 model. In third-party benchmarks, it outperformed leading American AI companies like OpenAI, Meta, and Anthropic.

For Meta's chief AI scientist, Yann LeCun, the biggest takeaway from DeepSeek's success was not the heightened threat posed by Chinese competition but the value of keeping AI models open source so that anyone can benefit.

It's not that China's AI is "surpassing the US," but rather that "open source models are surpassing proprietary ones," LeCun said in a post on Threads.

DeepSeek's R1 is itself open source, as is Meta's Llama. OpenAI, which was originally founded as an open-source AI company with a mission to create technology that benefits all of humanity, has on the other hand more recently shifted to closed-source.

LeCun said DeepSeek has "profited from open research and open source."

"They came up with new ideas and built them on top of other people's work. Because their work is published and open source, everyone can profit from it," LeCun said. "That is the power of open research and open source."

When DeepSeek unveiled R1 on January 20, which it said "demonstrates remarkable reasoning capabilities," the company said it was "pushing the boundaries" of open-source AI.

The announcement took Silicon Valley by surprise and was easily the most talked-about development in the tech industry during a week that included the World Economic Forum, TikTok uncertainty, and President Donald Trump's busy first few days in office.

Days after DeepSeek's announcement, Meta CEO Mark Zuckerberg said Meta planned to spend over $60 billion in 2025 as it doubles down on AI. Zuckerberg has been an outspoken advocate of open-source models.

"Part of my goal for the next 10-15 years, the next generation of platforms, is to build the next generation of open platforms and have the open platforms win," he said in September. "I think that's going to lead to a much more vibrant tech industry."

Those who support open source say it allows technology to develop rapidly and democratically since anyone can modify and redistribute the code. On the other hand, advocates for closed-source models argue that they are more secure because the code is kept private.

OpenAI CEO Sam Altman said the closed-source approach offers his company "an easier way to hit the safety threshold" in an AMA on Reddit last November. He added, however, that he "would like us to open source more stuff in the future."

• Tweens, teens, and college-aged kids showed off their Christmas hauls in TikTok videos.
• Casey Lewis, who writes about young consumers, watched about 1,000 haul videos, she told BI.
• Here are the top items that Gen Z kids bragged about getting for Christmas.

It was a very merry Christmas for some Gen Zers who took to social media to show off everything they unwrapped.

Casey Lewis, who writes the youth insights newsletter After School, analyzed Christmas haul TikTok videos from tweens, teens, and college-age consumers and compiled a list that she shared on her own TikTok.

"This is the third year I've done this sort of thing with the Christmas hauls, and I tried to refine my system just so that I'm able to actually crunch the data a little bit more scientifically," Lewis told Business Insider.

She said she watched hundreds of videos at double the speed to tally the standout gift items.

"I think conservatively, at least a thousand [videos]," Lewis said. "I was trying to discreetly binge Christmas haul TikToks while also spending time with my family."

From luxury clothing to throwback tech, these were the top gifts that the younger generation showed off in their Christmas hauls.

Digital cameras

"I think the thing that surprised me the most was how popular digital cameras were," Lewis said, noting that Gen Z has an affinity for Y2K nostalgia. "Everyone got digital cameras. It was also really interesting to see some of them got really expensive Sony ones, but then now Amazon makes those digital cameras that come in cute colors."

A basic Sony digital camera might run at around $750, but Lewis said she saw people showing off cheaper options from Amazon and Urban Outfitters for less than $100.

"They've sort of caught onto this trend, but then you kind of wonder how long is that going to last?" Lewis said.

UGG boots

UGG remained a hot item this year with the Ultra Mini boots and Classic Mini boots, which cost $150 to $160, reigning supreme.

"This year, it was the Minis, and last year, it was the Minis but also the Platforms," Lewis said. "Every year, they're just able to continue to be such a thing."

While she was going through the videos during the holidays at her childhood home, Lewis, who is 37, said she was surrounded by relics from her own childhood, like her own pair of UGG boots.

"Uggs and digital cameras — has anything changed? Am I still just a 16-year-old?" she said.

Rhode skin and beauty products

Lewis also said it was "staggering" how popular Rhode, Hailey Bieber's beauty brand, had become.

"Everyone got the lip peptide treatment," Lewis said. "It's such a popular skincare brand."

Rhode's peptide lip tint retails for $18.

"We know that celebrity brands are so fickle," Lewis said, "but it almost feels like this may have successfully reached the point where it's bigger than her and will thrive independently."

Sol de Janeiro products

Another popular beauty brand was Sol de Janeiro, which makes body and hair care as well as fragrances.

Gen Z kids showed off their "Cheirosa '62" perfume mist, which Lewis said was a big hit this year.

A full-size, 240 ml bottle retails for $38.

Jellycats

Tweens and teens went crazy over Jellycats, small plush toys that retail for $30 to $50.

"Jellycats were mentioned on every wish list, and they were very popular in hauls," Lewis said.

Lewis saw many Jellycats in haul videos but not as many as she expected, prompting her to question whether the kidult purchasing trend is declining.

"Are parents tired of buying their kids, their almost grown kids, stuffed animals? I don't know," Lewis said. "It feels very similar to Beanie Babies where it was a craze, but it wasn't able to sustain because no craze is."

White Fox apparel

Luxury loungewear remained popular this holiday season.

The $50 sweatpants from White Fox, which is headquartered in Australia, were "very popular" in haul videos, Lewis said.

"Athleisure had such a moment coming out of COVID, but young people are still very much prioritizing comfort clothes," Lewis said, noting that brands like Lululemon were also popular. "Teen and college-age girls, so many of them just wear sweat sets."

Roller Rabbit pajamas

Roller Rabbit pajamas were a popular gift pick as well, according to Lewis' analysis.

Available in dozens of different brightly colored patterns as well as in short- and long-sleeve options, the pajamas retail from $138 to $158.

Lewis noted the pajamas convey a sense of status.

Shark hair tools

Whereas last year saw a craze for Dyson hair tools, this year was all about Shark tools.

"I don't think I could have been trusted when I was a teen with an expensive hair tool," Lewis said. "I just don't think I could have taken care of it and not accidentally broken it."

While the classic set of Dyson hair tools retails for $600, the Shark set is comparatively more affordable at $300.

Vanity desk and mirror

While not a name-brand item, many tweens, teens, and college-age girls said in their Christmas haul TikToks that they got a vanity desk or a vanity mirror to put on their desk.

Vanity mirrors often come with lighting that is optimal to use while applying makeup. Depending on the brand, a desk with a vanity mirror might cost about $1,000.

Dae hair styling cream

A styling product from the brand Dae was a popular stocking stuffer, Lewis said.

The styling cream comes with a small wand that's helpful for doing a slick back hairstyle.

A 0.6 oz tube retails for $18.

Adidas Campus shoes

Adidas also continued its reign of popularity.

The Campus 00s, which retail for $110, were the go-to pick, Lewis said.

In previous years, Adidas Gazelles and Sambas were the choice picks.

Alani Nu energy drink

Alani Nu energy drinks were a popular, small-dollar item. Lewis referred to it as the "cool girl energy drink" in her TikTok analysis.

A 12-pack retails for $30.

"What's fascinating about that is it is a very accessible energy drink, but it's also very aesthetic," Lewis told BI. "The energy drink that appeared in so many Christmas hauls this year was nowhere to be found in Christmas hauls last year. So that's a little bit about how quickly some of this stuff changes."

Touchland hand sanitizer

Touchland hand sanitizers were another popular stocking stuffer, Lewis said in her analysis.

"$10 for a tiny hand sanitizer is kind of crazy," she told BI.

But for a 30ml hand sanitizer, it still carries some clout, she said.

"These more affordable, or at least accessible, items that have a little bit of status associated, a little bit of clout," Lewis said. "You don't need to have the Louis Vuitton, or you don't need to even have the Sony camera."

LoveShackFancy Perfume

LoveShackFancy's perfume in the scent "Forever In Love" was a hot gift, Lewis said.

A 2.5 oz bottle retails $125.

Other popular perfume runner-ups were Billie Eilish's "Eilish Eau de Parfum," which retails for $72 for a 3.4 oz bottle, and Glossier's "You," which costs $112 for a 100 ml bottle.

ONE/SIZE setting spray

Wrapping up the list was waterproof setting spray from ONE/SIZE by Patrick Starrr.

A 3.4 oz can of the mattifying spray retails for $32, adding to the subset of more affordable items that Lewis noted.

"There were not a lot of, I don't know, designer sunglasses. I did see a couple of designer purses," Lewis said. "It's not like there's one emerging or one dominant luxury item that everyone is feeling like they need to have."