• Some athletes who took podium spots at the Paris Olympics say their medals are deteriorating.
• Chaumet, a fine jewelry brand owned by LVMH, designed the medals.
• The International Olympic Committee said it will replace all "defective" medals.

All that glitters is not gold — and, as some athletes who competed in the Paris Olympics are finding out, even gold can lose its luster.

Since the 2024 Olympic Games last August, some Olympians who took home bronze, silver, and gold have taken to social media to complain that their medals are already showing signs of wear and tear.

They include French swimmers Clément Secchi and Yohann Ndoye Brouard, who posted photos on X of their gold medals in less-than-ideal shape in December.

"Paris 1924," Brouard wrote alongside crying face emojis in a post with images of his deteriorating gold medal.

The complaints mirror those of Team USA skateboarder Nyjah Huston. Shortly after the Games, he took to social media to show that his medal was already "looking rough."

"Olympic medals, we've got to step up the quality a little bit," Huston said in an Instagram story.

The medals were produced by the Monnaie de Paris, the French Mint, in partnership with the International Olympic Committee (IOC). Last week, the IOC said in a statement to France 24 that it was reviewing complaints and replacing "defective" medals.

In a statement to Business Insider, the Monnaie de Paris said it first received medal complaints in August, after which it "modified the varnish" used and "optimized its manufacturing process" to make them "more resistant to certain uses by athletes."

It also said it would replace and identically engrave "all damaged medals."

While the French Mint did not reveal the number of medals replaced, The New York Times reported on Tuesday that more than 100 athletes have issued complaints since the games.

Questions have also arisen for LVMH, the luxury conglomerate that partnered with the Olympics in 2024.

Ahead of the games, LVMH said that its fine jewelry brand Chaumet would design each medal — a task that the Maison embarked on with "creativity and passion," according to the LVMH website.

The Olympics marked one of the few highlights of 2024 for LVMH, a year in which its brands reported disappointing sales amid a widespread downturn in the luxury industry.

At the time, the Olympic partnership was a major marketing boost for LVMH, which — in light of the unfortunate medal situation — may no longer be the case.

This year is shaping up to be more promising for the French company. Its stock has risen sharply and and CEO Bernard Arnault's net worth is up almost $18 billion since January 1 to $194 billion, putting him in fifth place on the Bloomberg Billionaires Index.

LVMH and the IOC did not respond to requests for comment from Business Insider.

• An AI startup in China just showed how it's closing the gap with America's top AI labs.
• Chinese startup DeepSeek released a new AI model on Monday that appears to rival OpenAI's o1.
• Its reasoning capabilities have stunned top American AI researchers.

Donald Trump started his new presidency by declaring America must lead the world. He just got a warning shot from an AI crack team in China that is ready to show US technological supremacy is not a given.

Meet DeepSeek, a Chinese startup spun off from a decade-old hedge fund that calculates shrewd trades with AI and algorithms. Its latest release, which came on Trump's inauguration day, has left much of America's top industry researchers stunned.

In a paper released Monday, DeepSeek unveiled a new flagship AI model called R1 that shows off a new level of "reasoning." Why it has left such a huge impression on AI experts in the US matters.

Some of Silicon Valley's most well-resourced AI labs have increasingly turned to "reasoning" as a frontier of research that can evolve their technology from a student-like level of intelligence to something that eclipses human intelligence entirely.

To accomplish this, OpenAI, Google, Anthropic, and others have focused on ensuring models spend more time thinking before responding to a user query. It's an expensive, intensive process that demands a lot from the computing power buzzing underneath.

As a reminder, OpenAI fully released o1 — "models designed to spend more time thinking before they respond" — to a glowing reception in December after an initial release in September. DeepSeek's R1 shows just how quickly it can close the gap.

DeepSeek narrows the gap

What exactly does R1 do? For one, DeepSeek says R1 achieves "performance comparable to OpenAI o1 across math, code, and reasoning tasks."

Its research paper says that this is possible thanks to "pure reinforcement learning," a technique that Jim Fan, senior research manager at Nvidia, said was reminiscent of the secret behind making Google DeepMind's AlphaZero a master at games like Go and Chess from scratch, "without imitating human grandmaster moves first." "This is the most significant takeaway from the paper," he wrote on X.

DeepSeek, which launched in 2023, said in its paper that it did this because its goal was to explore the potential of AI to "develop reasoning capabilities without any supervised data." This is a common technique used by AI researchers. The company also said that an earlier version of R1, called R1-Zero, gave them an "aha moment" in which the AI "learns to allocate more thinking time to a problem to reevaluating its initial approach."

The end result offers what Wharton professor Ethan Mollick described as responses from R1 that read "like a human thinking out loud."

Notably, this level of transparency into the development of AI has been hard to come by in the notes published by companies like OpenAI when releasing models of a similar aptitude.

Nathan Lambert, a research scientist at the Allen Institute for AI, noted on Substack that R1's paper "is a major transition point in the uncertainty in reasoning model research" as "until now, reasoning models have been a major area of industrial research without a clear seminal paper."

Staying true to the open spirit, DeepSeek's R1 model, critically, has been fully open-sourced, having obtained an MIT license — the industry standard for software licensing.

Together, these elements of R1 provide complications to US players caught up in an AI arms race with China — Trump's main geopolitical rival — for a few reasons.

First, it shows that China can rival some of the top AI models in the industry and keep pace with cutting-edge developments coming out of Silicon Valley.

Second, open-sourcing highly advanced AI could also challenge companies that are seeking to make huge profits by selling their technology.

OpenAI, for instance, introduced a ChatGPT Pro plan in December that costs $200 per month. Its selling point was that it included "unlimited access" to its smartest model at the time, o1. If an open-source model offers similar capabilities for free, the incentive to buy a costly paid subscription could, in theory, diminish.

Nvidia's Fan described the situation like this on X: "We are living in a timeline where a non-US company is keeping the original mission of OpenAI alive — truly open, frontier research that empowers all."

DeepSeek has shown off reasoning know-how before. In November, the company released an "R1-lite-preview" that showed its "transparent thought process in real time." In December, it released a model called V3 to serve as a new, bigger foundation for future reasoning in models.

It's a big reason American researchers see a meaningful improvement in the latest model, R1.

Theo Browne, a software developer behind a popular YouTube channel for the tech community, said, "The new DeepSeek R1 model is incredible." Tanay Jaipuria, a partner investing in AI at Silicon Valley's Wing VC, also described it as "incredible."

Awni Hannun, a machine learning researcher at Apple, said that a key advantage of R1 was that it was less intensive, showing that the industry was "getting close to open-source o1, at home, on consumer hardware," referring to OpenAI's reasoning model introduced last year.

The model can be "distilled," meaning smaller but also powerful versions can run on hardware that is far less intensive than the computing power loaded into servers in data centers many tech companies depend on to run their AI models.

Hannun demonstrated this by sharing a clip on X of a 671 billion parameter version of R1 running on two Apple M2 Ultra chips, responding with reason to a prompt asking if a straight or a flush is better in a game of Texas Hold 'em. Hannun said its response came "faster than reading speed."

AI censorship

R1 does appear to have one key problem. Former OpenAI board member Helen Toner pointed out on X that there are demos of R1 "shutting itself down when asked about topics the CCP doesn't like."

Toner did suggest, however, that "the censorship is obviously being done by a layer on top, not the model itself." DeepSeek did not immediately respond to Business Insider's request for comment.

It is worth noting, of course, that OpenAI has introduced a new model called o3 that is meant to be a successor to the o1 model DeepSeek is currently rivaling. Lambert said it was "likely technically ahead" in his blog, with the key caveat that the model is "not generally available," nor will basic information like its "weights" be available anytime soon.

Given DeepSeek's track record so far, don't be surprised if its next model shows parity to o3. America's tech leaders may have met their match in China.

The spread of influenza A, COVID and RSV is "high" or "very high" across much of the U.S. at the same time norovirus cases are well above normal levels, Centers for Disease Control and Prevention and wastewater surveillance data shows.

Why it matters: The result is "quad-demic" of illness hitting simultaneously in what's shaping up to be a more active virus season than last year.

The big picture: The simultaneous threats are straining some hospitals to capacity and leading administrators to recommend masking among staff, ABC News reports.

• The surge follows what was a slow start to the respiratory virus season.
• "Predictions for this respiratory virus season were that we would see peak January 1 and that it would likely mirror previous respiratory virus seasons. We're obviously seeing it peak a little bit later," Saskia Popescu, a member of APIC's Emerging Infectious Diseases Task Force, told Axios.

The details: Flu activity is high or very high across 33 states and Washington, D.C., according to the CDC tracking of outpatient visits to health care providers for influenza-like illness.

• Mississippi, Oklahoma and Texas had the highest levels as of the week of Jan. 18, according to the Walgreens Flu Index, compiled using retail prescription data for antiviral medications used to treat the flu across Walgreens locations.
• The CDC data shows flu-like illness is "very high" in New York City while the Walgreens index identifies Oklahoma City; Lafayette, Louisiana; and Montgomery-Selma, Alabama, among the areas with the most activity.

What we're watching: Human metapneumovirus, or HMPV — which comes with symptoms of a cough, fever, nasal congestion and shortness of breath — made headlines in China and has also been spreading in the U.S.

• Activity in the U.S. remains low compared with other viruses, per CDC data, and experts have said it shouldn't be cause for panic. "It is a seasonal bug that we know how to manage," said Popescu, an assistant professor at the University of Maryland School of Medicine.
• For HMPV, as well as these other viruses that are circulating, the tried-and-true advice is particularly worth heeding right now, she said.
• "All of those mitigation efforts that you can do — washing your hands, covering your cough, cleaning, disinfection, being mindful of ventilation in shared spaces — that's all going to help," she said.

Over just eight hours on Inauguration Day, Presidents Trump and Biden forever stretched the immense public and private power of the presidency to once-unimaginable dimensions:

• Presidents can preemptively pardon family and friends in case of any accusation of grift or crimes.
• Presidents can pardon violent criminals convicted of sedition and violence in defense of their politics.

• Presidents and their families can start businesses — or even currencies — and profit without restriction or outcry.
• Oh, and they can do this with the presumption of presidential immunity.
• America doesn't have a king. But we're dancing close to king-like power.

Why it matters: Presidents could always pardon, profit or protect friends, family and allies. It just never has been done this broadly, this brazenly, this quickly. And with this much of a public shrug.

The big picture: So much of modern political and presidential power flows from precedent and imagination: doing unto others what the predecessors did — or did to them. And then stretching the hell out of it.

Biden, under the guise of protecting his family from unfair political and legal persecution, preemptively pardoned his brothers James and Frank Biden, his sister Valerie Biden Owens, and John Owens and Sara Biden, the spouses of Valerie and James. This is unprecedented.

• "It's disgusting," Bill Daley — a longtime Biden friend who was White House chief of staff under President Obama — told us. It "confirms that there are serious concerns about culpability." Daley said the Bidens will never wipe this "stain" from the former president's legacy.
• Trump blasted the pardons, moments before offering his own to approximately 1,500 people convicted or charged in the Jan. 6  attack on the Capitol — including violent criminals who attacked police officers.

Trump also pardoned Enrique Tarrio, the fascist Proud Boy leader convicted of seditious conspiracy — and serving a 22-year sentence in federal prison — for coordinating the attack on the Capitol from outside Washington.

• Trump also commuted the sentences of 14 extremist members of the Proud Boys and Oath Keepers convicted for plotting to violently overthrow the U.S. government and keep Trump in office.
• The sweeping acts of clemency stunned Washington and contradicted prior statements from Republicans — including Vice President JD Vance — that violent offenders should not be pardoned.

Between the lines: Biden, who earlier pardoned his son, Hunter, basically offered blanket immunity to family members who might be accused of profiting from this presidency.

Trump tested new limits by launching a surprise meme coin, $TRUMP, that vaulted him to crypto billionaire status two days before being inaugurated.

• Crypto insiders fear that $TRUMP — as well as the hastily launched $MELANIA meme coin — could destroy credibility that the scam-plagued industry has spent years trying to build.
• Remember, Trump once was a crypto skeptic and converted only during the 2024 campaign. He then became a beneficiary, to the tune of tens of billions of dollars, of the industry's open wallet. What an ROI!

Most Americans don't realize there are basically no limitations on presidents profiting off their reins of power through new businesses or business deals.

• Thanks to the Supreme Court, presidents also enjoy the presumption of immunity for "official acts" if they're ever accused of crossing any legal lines.

So Trump and his family conceivably could make billions through deals worldwide, new businesses and new currencies, funding the family — or even a political movement — for a generation. Their only limitation is imagination.

• America has drifted into uncharted waters in the rule of law. Trump and future presidents can test the limits with a presumption of success. And Biden's final act of pardons show Democrats have lost a lot of ability to cry foul.
• Senate Majority Leader John Thune (R-S.D.) told reporters on Capitol Hill Tuesday when asked about Trump's blanket Jan. 6 pardons: "We said all along that Biden opened the door on this."

We'll leave you with this: Now that presidential power is so broad, so deep, so uncontainable, why forfeit it? Well, here's an apparent loophole in the constitutional limit on two presidential terms:

• Trump or future presidents could simply run for a de facto third term — as the vice presidential nominee, with the understanding they will take power back once elected. That's but one of the once-unthinkable scenarios that seem more thinkable than ever.

Axios' Zachary Basu contributed reporting.