
Shares in Upexi fell 7.5% on Tuesday after it filed a $1 billion shelf registration, suggesting its Solana holdings could again grow after over five months of no purchases.


Shares in Upexi fell 7.5% on Tuesday after it filed a $1 billion shelf registration, suggesting its Solana holdings could again grow after over five months of no purchases.

The Ontario Securities Commission has cleared Matador to raise $58 million, which it intends to use to expand its Bitcoin treasury.

Bitcoin has been a “monster in financial markets” even though it hasn’t hit the most optimistic 2025 price targets, says Anthony Pompliano.
At first glance, it looks like the start of a human pregnancy: A ball-shaped embryo presses gently into the receptive lining of the uterus and then grips tight, burrowing in as the first tendrils of a future placenta appear.
This is implantation—the moment that pregnancy officially begins.
Only none of it is happening inside a body. These images were captured in a Beijing laboratory, inside a microfluidic chip, as scientists watched the scene unfold.

In three papers published this week by Cell Press, scientists are reporting what they call the most accurate efforts yet to mimic the first moments of pregnancy in the lab. They’ve taken human embryos from IVF centers and let these merge with “organoids” made of endometrial cells, which form the lining of the uterus.
The reports—two from China and a third involving a collaboration among researchers in the United Kingdom, Spain, and the US—show how scientists are using engineered tissues to better understand early pregnancy and potentially improve IVF outcomes.
“You have an embryo and the endometrial organoid together,” says Jun Wu, a biologist at the University of Texas Southwestern Medical Center, in Dallas, who contributed to both Chinese reports. “That’s the overarching message of all three papers.”
According to the papers, these 3D combinations are the most complete re-creations yet of the first days of pregnancy and should be useful for studying why IVF treatments often fail.
In each case, the experiments were stopped when the embryos were two weeks old, if not sooner. That is due to legal and ethical rules that typically restrict scientists from going any further than 14 days.
In your basic IVF procedure, an egg is fertilized in the lab and allowed to develop into a spherical embryo called a blastocyst—a process that takes a few days. That blastocyst then gets put into a patient’s uterus in the hope it will establish itself there and ultimately become a baby.

But that’s a common failure point. Many patients will learn that their IVF procedure didn’t work because an embryo never attached.
In the new reports, it’s that initial bond between mother and embryo that is being reproduced in the lab. “IVF means in vitro fertilization, but now this is the stage of in vitro implantation,” says Matteo Molè, a biologist at Stanford University whose results with collaborators in Europe are among those published today. “Considering that implantation is a barrier [to pregnancy], we have the potential to increase the success rate if we can model it in the laboratory.”
Normally implantation is entirely hidden from view because it occurs in someone’s uterus, says Hongmei Wang, a developmental biologist at the Beijing Institute for Stem Cell and Regenerative Medicine, who co-led the effort there. Wang often studies monkeys because she can interrupt their pregnancies to collect the tissues she needs to see. “We’ve always hoped to understand human embryo implantation, but we have lacked a way to do so,” she says. “It’s all happening in the uterus.”
In the Beijing study, researchers tested about 50 donated IVF embryos, but they also ran a thousand more experiments using so-called blastoids. The latter are mimics of early-stage human embryos manufactured from stem cells. Blastoids are easy to make in large numbers and, since they aren’t true embryos, don’t have as many ethical rules on their use.
“The question was, if we have these blastoids, what can we use them for?” says Leqian Yu, the senior author of the report from the Beijing Institute. “The obvious next step was implantation. So how do you do that?”
For the Beijing team, the answer was to build a soft silicone chamber with tiny channels to add nutrients and a space to grow the uterine organoid. After that, blastoids—or real embryos—could be introduced through a window in the device, so the “pregnancy” could start.
“The key question we want to try to answer is what is the first cross-talk between embryo and mother,” says Yu. “I think this is maybe the first time we can see the entire process.”
This isn’t the first time researchers have tried using organoids for this kind of research. At least two startup companies have raised funds to commercialize similar systems—in some cases presenting the organoids as a tool to predict IVF success. In addition to Dawn Bio, a startup based in Vienna, there is Simbryo Technologies, in Houston, which last month said it would begin offering “personalized” predictions for IVF patients using blastoids and endometrial organoids.
To do that test, doctors will take a biopsy of a patient’s uterine lining and grow organoids from it. After that, blastoids will be added to the organoids to gauge whether a woman is likely to be able to support a pregnancy or not. If the blastoids don’t start to implant, it could mean the patient’s uterus isn’t receptive and is the reason IVF isn’t working.
The Beijing team thinks the pregnancy organoids could also be used to identify drugs that might help those patients. In their paper, they describe how they made organoids out of tissue taken from women who’ve had repeated IVF failures. Then they tested 1,119 approved drugs on those samples to see if anything improved.
Several seemed to have helpful effects. One chemical, avobenzone, an ingredient in some types of sunblock, increased the chance that a blastoid would start implanting from just 5% of the time to around 25% of the time. Yu says his center hopes to eventually start a clinical trial if they can find the right drug to try.
The Beijing group is working on ways to improve the organoid system so that it’s even more realistic. Right now, it lacks important cell types, including immune cells and a blood supply. Yu says a next step he’s working on is to add blood vessels and tiny pumps to his chip device, so that he can give the organoids a kind of rudimentary circulation.
This means that in the near future, blastoids or embryos could likely be grown longer, raising questions about how far scientists will be able to take pregnancy in the lab. “I think this technology does raise the possibility of growing things longer,” says Wu, who says some view the research as an initial step toward creating babies entirely outside the body.
However, Wu says incubating a human to term in the laboratory remains impossible, for the time being. “This technology is certainly related to ectogenesis, or development outside the body,” he says. “But I don’t think it’s anywhere near an artificial womb. That’s still science fiction.”
Demis Hassabis, CEO of Google DeepMind, summed it up in three words: “This is embarrassing.”
Hassabis was replying on X to an overexcited post by Sébastien Bubeck, a research scientist at the rival firm OpenAI, announcing that two mathematicians had used OpenAI’s latest large language model, GPT-5, to find solutions to 10 unsolved problems in mathematics. “Science acceleration via AI has officially begun,” Bubeck crowed.
Put your math hats on for a minute, and let’s take a look at what this beef from mid-October was about. It’s a perfect example of what’s wrong with AI right now.
Bubeck was excited that GPT-5 seemed to have somehow solved a number of puzzles known as Erdős problems.
Paul Erdős, one of the most prolific mathematicians of the 20th century, left behind hundreds of puzzles when he died. To help keep track of which ones have been solved, Thomas Bloom, a mathematician at the University of Manchester, UK, set up erdosproblems.com, which lists more than 1,100 problems and notes that around 430 of them come with solutions.
When Bubeck celebrated GPT-5’s breakthrough, Bloom was quick to call him out. “This is a dramatic misrepresentation,” he wrote on X. Bloom explained that a problem isn’t necessarily unsolved if this website does not list a solution. That simply means Bloom wasn’t aware of one. There are millions of mathematics papers out there, and nobody has read all of them. But GPT-5 probably has.
It turned out that instead of coming up with new solutions to 10 unsolved problems, GPT-5 had scoured the internet for 10 existing solutions that Bloom hadn’t seen before. Oops!
There are two takeaways here. One is that breathless claims about big breakthroughs shouldn’t be made via social media: Less knee jerk and more gut check.
The second is that GPT-5’s ability to find references to previous work that Bloom wasn’t aware of is also amazing. The hype overshadowed something that should have been pretty cool in itself.
Mathematicians are very interested in using LLMs to trawl through vast numbers of existing results, François Charton, a research scientist who studies the application of LLMs to mathematics at the AI startup Axiom Math, told me when I talked to him about this Erdős gotcha.
But literature search is dull compared with genuine discovery, especially to AI’s fervent boosters on social media. Bubeck’s blunder isn’t the only example.
In August, a pair of mathematicians showed that no LLM at the time was able to solve a math puzzle known as Yu Tsumura’s 554th Problem. Two months later, social media erupted with evidence that GPT-5 now could. “Lee Sedol moment is coming for many,” one observer commented, referring to the Go master who lost to DeepMind’s AI AlphaGo in 2016.
But Charton pointed out that solving Yu Tsumura’s 554th Problem isn’t a big deal to mathematicians. “It’s a question you would give an undergrad,” he said. “There is this tendency to overdo everything.”
Meanwhile, more sober assessments of what LLMs may or may not be good at are coming in. At the same time that mathematicians were fighting on the internet about GPT-5, two new studies came out that looked in depth at the use of LLMs in medicine and law (two fields that model makers have claimed their tech excels at).
Researchers found that LLMs could make certain medical diagnoses, but they were flawed at recommending treatments. When it comes to law, researchers found that LLMs often give inconsistent and incorrect advice. “Evidence thus far spectacularly fails to meet the burden of proof,” the authors concluded.
But that’s not the kind of message that goes down well on X. “You’ve got that excitement because everybody is communicating like crazy—nobody wants to be left behind,” Charton said. X is where a lot of AI news drops first, it’s where new results are trumpeted, and it’s where key players like Sam Altman, Yann LeCun, and Gary Marcus slug it out in public. It’s hard to keep up—and harder to look away.
Bubeck’s post was only embarrassing because his mistake was caught. Not all errors are. Unless something changes researchers, investors, and non-specific boosters will keep teeing each other up. “Some of them are scientists, many are not, but they are all nerds,” Charton told me. “Huge claims work very well on these networks.”
*****
There’s a coda! I wrote everything you’ve just read above for the Algorithm column in the January/February 2026 issue of MIT Technology Review magazine (out very soon). Two days after that went to press, Axiom told me its own math model, AxiomProver, had solved two open Erdős problems (#124 and #481, for the math fans in the room). That’s impressive stuff for a small startup founded just a few months ago. Yup—AI moves fast!
But that’s not all. Five days later the company announced that AxiomProver had solved nine out of 12 problems in this year’s Putnam competition, a college-level math challenge that some people consider harder than the better-known International Math Olympiad (which LLMs from both Google DeepMind and OpenAI aced a few months back).
The Putnam result was lauded on X by big names in the field, including Jeff Dean, chief scientist at Google DeepMind, and Thomas Wolf, cofounder at the AI firm Hugging Face. Once again familiar debates played out in the replies. A few researchers pointed out that while the International Math Olympiad demands more creative problem-solving, the Putnam competition tests math knowledge—which makes it notoriously hard for undergrads, but easier, in theory, for LLMs that have ingested the internet.
How should we judge Axiom’s achievements? Not on social media, at least. And the eye-catching competition wins are just a starting point. Determining just how good LLMs are at math will require a deeper dive into exactly what these models are doing when they solve hard (read: hard for humans) math problems.
This story originally appeared in The Algorithm, our weekly newsletter on AI. To get stories like this in your inbox first, sign up here.