Singularity HUBModify

2025-12-09 07:55:01

‘Tis the season for overindulgence. But for people with allergies, holiday feasting can be strewn with landmines.

Over three million people worldwide tiptoe around a food allergy. Even more experience watery eyes, runny noses, and uncontrollable sneezing from dust, pollen, or cuddling with a fluffy pet. Over-the-counter medications can control symptoms. But in some people, allergic responses turn deadly.

In anaphylaxis, an overactive immune system releases a flood of inflammatory chemicals that closes up the throat. This chemical storm stresses out the heart and blood vessels and limits oxygen to the brain and other organs.

Early diagnosis, especially of shellfish or nut allergies, helps people avoid these foods. And in an emergency, EpiPens loaded with epinephrine can relax airways and save lives. But the pens must be carried at all times, and patients—especially young children—struggle with this.

An alternative is to train the immune system to neutralize its over-zealous response. This month, a team from the University of Toulouse in France presented a long-lasting treatment that fights off anaphylactic shock in mice. Using a vaccine, they rewired part of the immune system to battle Immunoglobulin E (IgE), a protein that’s involved in severe allergic reactions.

A single injection into mice launched a tsunami of antibodies against IgE, and levels of those antibodies remained high for at least 12 months—which is over half of a mouse’s life. Despite triggering an immune civil war, the mice’s defenses were still able to fight a parasitic infection. The vaccine is, in theory, a blanket therapy for most food allergies, from peanuts to shellfish.

Although it needs more testing before clinical trials, the treatment is a “very enticing therapeutic candidate that fills an important need,” wrote Danielle Libera at McMaster University and colleagues, who were not involved in the study.

Double Agent

An army of immune cells roams our bodies to surveil and fight off invaders. When the system detects danger—pathogens, cancer cells, or foreign organs—it springs into action.

Some cells locate the threat and act as a beacon to other immune troops. T cells activate and physically lock onto a target, releasing toxic chemicals that punch holes in the invader’s protective membrane. B cells send in tailored antibodies to further neutralize the enemy.

But sometimes the well-oiled immune machine goes awry. Allergies are caused by friendly fire from B cells as they churn out antibodies to suit the body’s needs. Immunoglobulin G (IgG) provides overall immune support. Immunoglobulin A (IgA) protects the lining of the gut and lungs. IgE fights off parasites—and also triggers severe allergic reactions.

In food allergies, for example, allergens in the gut trigger B cells to switch antibody production from IgG to allergen-specific IgE. In the bloodstream, IgE meets up with mast cells, sensitizes them to the allergen, and keeps them on high alert.

If the person eats food containing the same allergen again, the allergen grabs onto these sensitized cells and prompts them to release a deluge of chemicals, such as histamines.

Cue immediate symptoms: Blood vessels dilate and leak, causing flushing, swelling, and a sudden drop in blood pressure. Smooth muscles contract and restrict airways. Mast cells recruit more immune fighters, and mucus and inflammation in the lungs skyrocket.

EpiPens immediately counteract some of these responses and provide valuable time for more intensive treatment. But patients must have one nearby, and the pens aren’t preventative. In 2024, the US FDA approved an antibody therapy that lowers IgE in the body after accidental allergen exposure as a preventative measure. But the treatment requires an injection every two to four weeks, is costly, and ironically, can inadvertently trigger anaphylaxis in some people.

Self-Made Solution

Instead of injecting an antibody against IgE, why not coax the body to make its own?

The idea was first pitched in the early 1990s. But there were roadblocks, side effects being most notable. Earlier attempts at an IgE vaccine unexpectedly activated mast cells and triggered runaway immune reactions. The immune system also rapidly adapted. Newly formed IgE antibodies can be tagged as invaders, resulting in a counterattack that depletes levels of the antibodies levels over time.

However, the authors of the latest study had access to a wealth of new information. Atomic-level scans revealed that IgE toggles between two states. In an “open ” state, IgE grabs onto mast cells and allergens, forming a bridge that triggers allergic responses. But some antibodies can lock IgE into a “closed” state where it no longer connects with mast cells, severing the anaphylactic cascade.

The team engineered a vaccine using these antibodies to keep IgE in its closed state. The vaccine also stimulates the immune system to produce high levels of the antibodies.

Called IgE-K, the vaccine protected mice from multiple allergic reactions, including to peanuts, and completely prevented anaphylaxis. Two vaccine doses produced persistent antibodies that lasted for a year at sufficiently high amounts to ward off additional allergic reactions.

The results indicate that IgE-K may overcome depletion and establish a long-term antibody reservoir, wrote Libera and colleagues. It’s an especially promising strategy for food allergies that are lifelong in more than 80 percent of affected people.

Although the vaccine dampened IgE activity, it didn’t interfere with the antibody’s ability to clear parasites. Vaccinated mice knocked out a worm infection similarly to their non-treated peers. However, the experimental model relied on mast cells to fight off the infection as opposed to IgE per se. The team is now exploring the vaccine’s impact on other parts of the immune system, especially the B-cells in charge of making antibodies.

The study is a first step. But if all goes well, kids with severe allergies could have their PB&J and eat it too.

The post Scientists Just Developed a Lasting Vaccine to Prevent Deadly Allergic Reactions appeared first on SingularityHub.

2025-12-06 23:00:00

The post This Week’s Awesome Tech Stories From Around the Web (Through December 6) appeared first on SingularityHub.

2025-12-05 23:00:00

OpenAI chief executive Sam Altman—perhaps the most prominent face of the artificial intelligence boom that accelerated with the launch of ChatGPT in 2022—loves scaling laws.

These widely admired rules of thumb linking the size of an AI model with its capabilities inform much of the headlong rush among the AI industry to buy up powerful computer chips, build unimaginably large data centers, and re-open shuttered nuclear plants.

As Altman argued in a blog post earlier this year, the thinking is that the “intelligence” of an AI model “roughly equals the log of the resources used to train and run it”—meaning you can steadily produce better performance by exponentially increasing the scale of data and computing power involved.

First observed in 2020 and further refined in 2022, the scaling laws for large language models (LLMs) come from drawing lines on charts of experimental data. For engineers, they give a simple formula that tells you how big to build the next model and what performance increase to expect.

Will the scaling laws keep on scaling as AI models get bigger and bigger? AI companies are betting hundreds of billions of dollars that they will—but history suggests it is not always so simple.

Scaling Laws Aren’t Just for AI

Scaling laws can be wonderful. Modern aerodynamics is built on them, for example.

Using an elegant piece of mathematics called the Buckingham π theorem, engineers discovered how to compare small models in wind tunnels or test basins with full-scale planes and ships by making sure some key numbers matched up.

Those scaling ideas inform the design of almost everything that flies or floats, as well as industrial fans and pumps.

Another famous scaling idea underpinned the boom decades of the silicon chip revolution. Moore’s law—the idea that the number of the tiny switches called transistors on a microchip would double every two years or so—helped designers create the small, powerful computing technology we have today.

But there’s a catch: not all “scaling laws” are laws of nature. Some are purely mathematical and can hold indefinitely. Others are just lines fitted to data that work beautifully until you stray too far from the circumstances where they were measured or designed.

When Scaling Laws Break Down

History is littered with painful reminders of scaling laws that broke. A classic example is the collapse of the Tacoma Narrows Bridge in 1940.

The bridge was designed by scaling up what had worked for smaller bridges to something longer and slimmer. Engineers assumed the same scaling arguments would hold: If a certain ratio of stiffness to bridge length worked before, it should work again.

Instead, moderate winds set off an unexpected instability called aeroelastic flutter. The bridge deck tore itself apart, collapsing just four months after opening.

Likewise, even the “laws” of microchip manufacturing had an expiry date. For decades, Moore’s law (transistor counts doubling every couple of years) and Dennard scaling (a larger number of smaller transistors running faster while using the same amount of power) were astonishingly reliable guides for chip design and industry roadmaps.

As transistors became small enough to be measured in nanometers, however, those neat scaling rules began to collide with hard physical limits.

When transistor gates shrank to just a few atoms thick, they started leaking current and behaving unpredictably. The operating voltages could also no longer be reduced without being lost in background noise.

Eventually, shrinking was no longer the way forward. Chips have still grown more powerful, but now through new designs rather than just scaling down.

Laws of Nature or Rules of Thumb?

The language-model scaling curves that Altman celebrates are real, and so far they’ve been extraordinarily useful.

They told researchers that models would keep getting better if you fed them enough data and computing power. They also showed earlier systems were not fundamentally limited—they just hadn’t had enough resources thrown at them.

But these are undoubtedly curves that have been fit to data. They are less like the derived mathematical scaling laws used in aerodynamics and more like the useful rules of thumb used in microchip design—and that means they likely won’t work forever.

The language model scaling rules don’t necessarily encode real-world problems such as limits to the availability of high-quality data for training or the difficulty of getting AI to deal with novel tasks—let alone safety constraints or the economic difficulties of building data centers and power grids. There is no law of nature or theorem guaranteeing that “intelligence scales” forever.

Investing in the Curves

So far, the scaling curves for AI look pretty smooth—but the financial curves are a different story.

Deutsche Bank recently warned of an AI “funding gap” based on Bain Capital estimates of a $800 billion mismatch between projected AI revenues and the investment in chips, data centers, and power that would be needed to keep current growth going.

JP Morgan, for their part, has estimated that the broader AI sector might need around $650 billion in annual revenue just to earn a modest 10 percent return on the planned build-out of AI infrastructure.

We’re still finding out which kind of law governs frontier LLMs. The realities may keep playing along with the current scaling rules; or new bottlenecks—data, energy, users’ willingness to pay—may bend the curve.

Altman’s bet is that the LLM scaling laws will continue. If that’s so, it may be worth building enormous amounts of computing power because the gains are predictable. On the other hand, the banks’ growing unease is a reminder that some scaling stories can turn out to be Tacoma Narrows: beautiful curves in one context, hiding a nasty surprise in the next.

This article is republished from The Conversation under a Creative Commons license. Read the original article.

The post AI Companies Are Betting Billions on AI Scaling Laws. Will Their Wager Pay Off? appeared first on SingularityHub.

2025-12-05 02:06:06

On the surface, cystic fibrosis and Tay-Sachs disease have nothing in common. Although both are inherited genetic disorders, one causes thick mucus buildup in the lungs, making it progressively harder to breathe; the other gradually leads to a buildup of fatty molecules that kills brain cells.

But under the hood, the diseases share a common villain: Nonsense mutations.

Like a molecular “stop sign,” these mutations instruct cells to abandon making certain proteins, resulting in truncated versions that don’t work and lead to disease. Gene-editing tools can correct mutated genes by targeting them one by one for each disease. While this approach can save lives, it takes time and a lot of resources to develop.

Why not aim for the common villain?

This month, a team led by David Liu at Harvard University developed a gene-editing tool to correct nonsense mutations. Called PERT, the tool tackles the mutations head on, allowing cells to ignore the mutations and continue producing proteins to their full length and potential.

In cultured human cells and mice with nonsense mutations, a single dose of PERT rescued protein production and eased disease symptoms. Because PERT is inserted into the genome, the treatment is, in theory, one-and-done.

The gene therapy could be a boon for rare diseases. More than 7,000 inherited diseases affect hundreds of millions of people around the world. Nonsense mutations are involved in roughly 30 million. Few have any treatment or cure.

“The most effective strategy for addressing this unmet need would be to develop therapies that are effective against multiple rare disorders,” wrote Kim Keeling at the University of Alabama, who was not involved in the study.

A more universal gene editor fits the bill. If proven safe and effective in humans, PERT sets the stage for more affordable gene therapies, faster development times, and more importantly—hope for people with rare diseases that have been largely sidelined in the past.

Hacking the Protein Machine

Proteins are the workhorses of our bodies. They’re the foundational building blocks of tissues and organs and are in charge of intricate biological functions, from regulating immune responses to digesting food.

The blueprints for proteins are embedded in our DNA as three-letter codons. Each codon represents a single amino acid, the basic molecules that make up proteins. These codons are transcribed into molecular transports called mRNA, which shuttle the information to the cell’s protein-making factory for assembly.

The factory reads the codons one by one and instructs a team of molecular chauffeurs called tRNA to grab the correct amino acid and bring it to the assembly line. In this way, the factory translates the body’s genetic code into a ribbon-like protein chain.

Nonsense mutations bring the process to a screeching halt. The factory needs instructions on when a protein chain is complete so it can be released for further processing. These instructions are called “stop codons” and are made of several unique genetic letter combinations.

Some genetic diseases have a single genetic mutation that turns a protein-making codon into a stop codon—basically, pulling an emergency switch to shut down production. Rather than making the entire functional protein, the cell destroys mRNA shuttles and truncates the resulting proteins. These proteins are less stable or struggle to perform their roles.

Previous studies found a workaround: Nonsense suppression.

Engineered suppressor tRNA molecules can skip over nonsense mutations. Like molecular smugglers, these synthetic RNA molecules sneak amino acids to spots where the protein should have been terminated. This trick rewires the code and lets the protein-making factory skip the stop command and keep making the rest of the protein.

The strategy has already had successes. In one study, synthetic suppressor tRNA molecules delivered by a virus were shown safe and effective in mice with a nonsense mutation, and beneficial effects from a single treatment lasted for more than half a year. Another suppressor tRNA molecule wrapped up in a fatty bubble for delivery—a commonly used system in gene therapies—restored production of a protein in mice with cystic fibrosis, allowing them to better breathe.

Both methods have downsides though. Viral carriers, even when stripped of their disease-causing traits, can still stir up immune responses. And although using fatty bubbles to deliver therapies is relatively safer, they require multiple doses in chronic genetic diseases.

Precision Editor

Liu and colleagues brainstormed a one-and-done therapy that directly inserts instructions for suppressor tRNA molecules into cells or an animal’s genetic code.

After screening thousands of tRNA variants, they found a highly active candidate as a starting point. Using prime editing, a type of small and precise gene editor, they altered natural versions of tRNA into suppressor versions that recognized a specific mutated stop codon.

Rather than terminating the building project, the engineered tRNA shuttled an amino acid into place to override the mutation and finish constructing the full-length protein.

The team tested the new tool, called PERT, in several human cell types in petri dishes. The cells harbored nonsense mutations for different genetic diseases, including cystic fibrosis and Tay-Sachs disease. A single dose increased working proteins by 20 to 70 percent regardless of the disease.

The therapy also worked in mice with a nonsense mutation causing a severe disease called Hurler syndrome in humans. Here, the body struggles to make a protein that degrades a type of sugar molecule, which builds up and causes cellular mayhem. Seven weeks after a single treatment, the mice had 8 percent more of the protein—enough to decrease harmful sugars and alleviate symptoms.

Making Sense Out of Nonsense

PERT’s strength is in its versatility. In a screen of over 14,000 mutated stop codons, the gene editor bypassed mutations roughly 70 percent of the time.

But while the results are promising, altering nonsense mutations can be fickle.

Inserting an amino acid into a growing protein chain can impact its function and stability. Proteins largely depend on their 3D structures to interact with other biological molecules, and a single change in amino acid makeup could alter the overall architecture.

It’s therefore unlikely “that the engineered tRNA will recover optimal function for all proteins” that have nonsense mutations, wrote Keeling. The current study focused on one stop codon: UGA. Several others exist and are now targets for other suppressor tRNA molecules.

Using prime editing, the molecules can linger in the body to continually produce the engineered versions without need for repeated jabs. From here, scientists must conduct long-term animal studies to test the edited tRNA’s stability and side effects.

There’s also the dosage problem. An optimal amount for liver tissue may be too large or ineffective for the heart or lungs. Eventually the team envisions a library of PERT tools, tailored to each organ and frozen in a fridge for use on command. With their work, the team has brought the therapeutic use of suppressor tRNA molecules “a step closer,” wrote Keeling.

The post One Dose of This Gene Editor Could Defeat a Host of Genetic Diseases Suffered by Millions appeared first on SingularityHub.

2025-12-03 05:29:25

Ollie Chu was three years old when he received an infusion that would change his life.

Born with a rare inherited condition called Hunter syndrome, Ollie’s body couldn’t produce an enzyme that breaks down complex sugars.

Just a few months after his birth, the sugars had built up everywhere, wreaking havoc on lungs, liver, skin, and brain. In Hunter syndrome, joints stiffen and airways narrow, making it hard to breathe. The brain also struggles to grow, resulting in developmental delays and cognitive problems. Most kids diagnosed with the condition don’t live past 20.

There are a few treatments. One drug on the market counteracts some bodily symptoms but at a hefty price. It must be taken weekly for life and can’t rescue the brain. Another option is a full bone marrow replacement. While this offers a long-term solution, the procedure is risky for toddlers and depends on the availability of matching donors, who are few and far between.

Ollie’s treatment is new. Roughly a year ago, researchers at the University of Manchester removed stem cells from his body, genetically inserted a functional copy of the gene encoding the missing enzyme, and infused the edited cells back into his body through a catheter.

Now, he no longer depends on weekly drug infusions. “[He] is doing great since having the gene therapy. We have seen dramatic improvements, and he continues to grow physically and cognitively,” said his dad, Ricky, in a press release.

Ollie is one of five very young children in an ongoing clinical trial of gene therapy for Hunter syndrome. Led by the Royal Manchester Children’s Hospital and collaborators, researchers hope the one-and-done therapy will slash treatment time and offer a lasting solution.

“Gene therapy is not only safer and more effective [than bone marrow transplant], but it enables us to use the child’s own cells which cuts out the need to find a donor,” said joint clinical lead Rob Wynn. If successful, the principles could be adapted for other genetic diseases.

Broken Waste Plant

Cells are constantly building, destroying, and recycling proteins. They monitor the levels of different molecules—sugars, fats, and proteins—and shuttle excess to the lysosome.

Think of the lysosome as a cell’s “stomach.” Each bubble-like structure contains acidic fluids and a menagerie of enzymes to break down different types of molecules.

One of these enzymes, called iduronate-2-sulfatase (IDS), is missing in Hunter syndrome. The enzyme exists in all cells, but it’s most active in the liver, skin, immune system, and brain. Rather than staying put, IDS loves to roam about and explore neighboring cells. In other words, if only a fraction of cells can make the enzyme, its effects would still spread beyond just the treated ones.

The enzyme replacement therapy Ollie and other kids with Hunter syndrome begin early in life relies on IDS. Here, the enzyme is infused into the bloodstream where it’s absorbed into multiple tissues to help clean out toxic sugars. The treatment improves lung and liver function and helps with joint mobility. But due to its large size, it can’t enter the brain. Hence, the disease continues to attack neural function.

At the root of Hunter syndrome is the gene that produces IDS. Using viruses and gene editing, studies have shown that delivering a healthy version of the gene to mice boosts production of the enzyme. Some genetic diseases have only a single DNA letter change. But the IDS gene mutates in hundreds of ways, making it difficult to engineer a universal gene therapy.

A bone marrow transplant from a matching healthy donor is one workaround. Donor stem cells gradually develop into a range of healthy blood and immune cells. Because they have a normal version of the IDS gene, these cells pump the missing enzyme throughout the body.

A transplant is a one-and-done treatment, but the recipient must take immunosuppressant drugs for the rest of their life, increasing the chance of infections. And the wait for a matching donor can be very long.

Complete Replacement

In Ollie’s treatment, researchers harvested his own stem cells for gene therapy. Because the cells come from his body, they’re more likely to evade immune rejection.

The approach is based on a mouse study by Brian Bigger and colleagues, who is also co-leading the clinical trial. It uses a viral carrier, stripped of disease-causing genes, to shuttle a healthy IDS gene into blood stem cells outside the body. The edited cells are then infused back into the patient. The virus inserts the gene directly into the cell’s genome, ensuring the replacement isn’t lost when the cells divide.

Rather than using a natural version of IDS, the team added a snippet to the gene that helps the enzyme better tunnel into the brain. Once infused, the edited stem cells multiply into a variety of blood and immune cells that roam the body and release the working enzyme.

In mice modeling Hunter syndrome, a single treatment completely reversed brain symptoms for up to 16 months—or almost their entire lifespan. Other organs also benefited without notable side effects.

In late 2024, Ollie, at just three years of age, underwent a similar procedure. His doctors collected and isolated his blood stem cells and genetically tweaked them to churn out the missing enzyme. As he watched cartoons, the team infused two doses of the edited cells through a catheter. He quickly recovered and was discharged from the hospital a few days later.

Within three months of the infusion, Ollie was able to come off the weekly drug infusions that had dominated his life. His speech and motor abilities improved, allowing him to ride a tricycle, hang out with friends, and enjoy a normal childhood.

“I want to pinch myself every time I tell people that Oliver is making his own enzymes,” his mother Jingru told the BBC. “Every time we talk about it I want to cry because it’s just so amazing.”

The team is recruiting other children with Hunter syndrome in the ongoing clinical trial to further test safety and efficacy. Because symptoms progress so rapidly before causing brain damage, the trial only accepts patients between three and 12 months of age. (At first, doctors thought Ollie was too old, but testing showed his condition had progressed only a little.) Once treated, the children will be followed for two years to gauge the therapy’s effects against common symptoms, such as delayed learning, hearing issues, and heart and lung problems.

If successful, the same gene-editing approach could be used to treat other inherited diseases involving stem cells. Ollie’s parents are hopeful the therapy might be extended to older children, including his five-year-old brother Skyler, who also has Hunter syndrome but is currently too old for the trial.

Still, to his father Ricky, the experimental treatment has been a success.

“We’re excited for Ollie’s future. Seeing the difference for Ollie pre-and post-transplant has made us believers,” he said. “We hope that one day, a treatment becomes available for all children at all stages of Hunter syndrome.”

The post Groundbreaking Gene Therapy Transforms Life of Boy With Devastating Disorder appeared first on SingularityHub.

2025-12-02 04:52:51

Space is getting busier as national space agencies and private companies increase the tempo of launches. But today’s approach to space exploration is unsustainable, say researchers, and we need to do more to make sure the orbital economy is a circular one.

While companies like SpaceX have made progress with reusable rockets, most launch vehicles are used only once, and their remains are left to either burn up in the atmosphere or clutter low-Earth orbit. They also dump huge quantities of greenhouse gases and ozone-depleting chemicals into the upper atmosphere.

Satellites are similarly unsustainable. After completing missions, they’re often either moved to a “graveyard orbit,” or worse, they add to the growing pile of space junk making low-Earth orbit increasingly hard to navigate.

As the pace of launches grows, these approaches are no longer viable, say researchers. In a paper published in Chem Circularity, scientists argue we need to shift to a “circular space economy” designed around the principles of reducing, repairing, and recycling.

“As space activity accelerates, from mega-constellations of satellites to future lunar and Mars missions, we must make sure exploration doesn’t repeat the mistakes made on Earth,” the University of Surrey’s Jin Xuan says in a press release. “A truly sustainable space future starts with technologies, materials, and systems working together.”

Progress already made in shifting industries like electronics and automotive manufacturing to more circular practices could provide a template for the space economy, say the researchers.

To reduce waste in the industry, they say spacecraft need to be more durable to increase their lifespans. This could slash material waste from the vehicles themselves and reduce the number of launches required.

Making spacecraft more repairable could also play an important role, they note. To make this possible, space stations would need to become hubs that carry out maintenance and build spacecraft components. They could also refuel satellites already in orbit to extend their lifespans.

Recycling spacecraft is more challenging due to the enormous amount of wear and tear they undergo in the harsh conditions of space and the punishing process of re-entering the atmosphere. The researchers say companies need to further develop soft-landing systems like parachutes and airbags to ensure vehicles can be brought back safely.

The study also calls for systematic efforts to clear existing orbital debris. These would reduce the risk of collisions but could also recover valuable materials. The work would require new tools like robotic arms and nets that can safely capture spacecraft moving at thousands of miles per hour.

The biggest challenge, the researchers say, is this would represent a fundamental shift in the way the space industry operates. That means piecemeal progress on individual components or processes won’t cut it: What’s needed is a system-wide commitment to radically different ways of operating.

“We need innovation at every level, from materials that can be reused or recycled in orbit and modular spacecraft that can be upgraded instead of discarded, to data systems that track how hardware ages in space,” says Xuan. “But just as importantly, we need international collaboration and policy frameworks to encourage reuse and recovery beyond Earth.”

That may prove challenging in an arena that has been characterized by intense geopolitical competition. But we have a golden opportunity to avoid the same mistakes we have made here on Earth.

The post Scientists Say We Need a Circular Space Economy to Avoid Trashing Orbit appeared first on SingularityHub.