All News (Page 361)

As wealth taxes gain momentum from Sacramento to Washington State, Sen. Bernie Sanders says 938 people stand between most working Americans and a $3,000 check.

In a scathing op-ed published Wednesday in the Guardian, the Vermont senator named every name and put every number on the table. “The richest people in America have never ever had it so good,” he wrote, while mentioning that 60% of Americans live paycheck to paycheck and 85 million are uninsured or underinsured.

“We have a tax code that is totally rigged—written by representatives of the wealthy to benefit the wealthy,” he wrote in the op-ed, while also referencing estimates from the Rand Corporation that found nearly $80 trillion in wealth has been redistributed from the bottom 90% to the top 1% over the past 50 years.

Sanders is making headlines for once again calling for billionaires to pay up. Last month, he and Rep. Ro Khanna introduced the Make Billionaires Pay Their Fair Share Act that would put a 5% tax on the estimated 938 billionaires in the country and raise an expected $4.4 trillion over a decade. A portion of the first year’s revenue will be redistributed in the form of $3,000 checks to anyone making less than $150,000 a year.

The funds raised through the proposed tax would be enough to repeal the Medicaid cuts that threw 15 million Americans off coverage; fund universal childcare; guarantee teachers a $60,000 minimum salary; expand Medicare to cover dental, vision, and hearing, and build 7 million affordable housing units.

Sanders calls out billionaires

For Sanders, who has long called for the wealthy to pay up, the tax has never been so necessary.

“The American working class has been under savage attack for years,” he wrote in the op-ed. That’s owing to the disparity between what the rich and those in the working class pay. Sanders name-dropped Elon Musk, saying his net worth of $805 billion is more wealth than the bottom 53% of American households put together. But according to Sanders, the richest man in the world pays an effective tax rate of just 3.3%, lower than the 8.4% paid by an average truck driver.

Had the 5% billionaire tax been put into effect last year, Sanders noted that Musk—expected to become the world’s first trillionaire—would hardly notice the difference. “Let me tell you how insane the level of wealth inequality is in America today,” Sanders wrote. “Musk would have owed $42 billion more in taxes, leaving him with just $792 billion to survive.”

The Tesla CEO wasn’t the only one in the senator’s crosshairs. Warren Buffett, who has long made comments about ensuring the ultrawealthy pay their fair share and famously noted having a lower effective tax rate than his secretary—pays just 0.1%, while a schoolteacher pays 9.8%, Sanders said. Former New York City Mayor Michael Bloomberg, with a net worth of $109 billion, had an effective tax rate of 1.3%, compared with the 13.3% paid by the average registered nurse.

The names didn’t stop there. Amazon founder Jeff Bezos paid less than 1% in taxes, when the average firefighter paid 8.7%, according to Sanders. Had the billionaire tax been enacted last year, Bezos would have paid $11 billion, leaving him with a paltry $207 billion. Meta cofounder Mark Zuckerberg would have $209 billion left after paying $11 billion more with the tax enacted.

Political support for the tax

There is growing political support for such taxes. In an opinion poll, Californians backed a similar (albeit, one-time) billionaire tax by a two-to-one margin to protect 3 million people from losing health care. More than half (62%) of New Yorkers support Mayor Zohran Mamdani’s proposed 2% surtax on millionaires and billionaires. Nationally, more than six in 10 Americans say the wealthy and large corporations pay too little in taxes, and one in five think it’s morally wrong to be that rich.

The ultrawealthy, however, aren’t waiting around to find out if anyone listens. Google cofounders Larry Page (with a net worth of $244 billion) and Sergey Brin (with a net worth of $226 billion) rushed to leave California before the Jan. 1, 2026, deadline set by the proposed Billionaire Tax Act, both purchasing property in Florida. Howard Schultz and Zuckerberg followed last month. They join Jeff Bezos, Peter Thiel, Ken Griffin, and Larry Ellison, all of whom have already bought property or moved operations to the Sunshine State in recent years.

This story was originally featured on Fortune.com

Longevity supplements are becoming increasingly popular, and one molecular compound gaining attention is nicotinamide mononucleotide (NMN). Naturally found in the body, NMN converts directly to nicotinamide adenine dinucleotide (NAD+), a coenzyme that contributes to energy production, metabolism, and overall cellular health, making it a key player in how our bodies function over time.

The growing interest in NMN aligns with a broader cultural shift in how people are thinking about aging. Registered Dietitian Avery Zenker points out that “longevity science isn’t only focused on increasing lifespan, but also improving healthspan.” Healthspan, in short, refers to living better as you age, not just living longer. According to Zenker, “NMN stands out for its potential to impact both lifespan and healthspan.”

So, what exactly is NMN, and why is it being promoted by so many wellness influencers? In this article, we look at how NMN works in your body, and we break down the claims around this popular wellness supplement.

What is NMN?

Naturally found in the body, NMN converts directly to NAD+, a compound essential for good health. An increase in NAD+ is associated with increased energy production and faster cellular repair.

Emerging research shows that NAD+ levels decrease greatly in middle age. According to Rachel Pojednic, PhD, Chief Science Officer at RestoreLabs and Restore Hyper Wellness, the popularity of NMN supplements is based on the idea “that by supplementing with NMN, you can increase NAD+ availability and potentially support cellular processes that are associated with healthy aging.”

How do NMN supplements work?

Pojednic explains it this way: an NAD+ boosting supplement like NMN is not like a direct infusion of NAD+ to your body. Rather, the supplements provide essential building blocks (called precursors) that help your cells make NAD+ on their own.

“NAD+ does not occur in high concentrations in the blood,” Pojednic explains, “and in fact, that would be an indicator to the immune system that something was wrong.” By using a precursor, like an NMN supplement, your body can produce NAD+ where it’s needed—inside your cells—without setting off any alarms within your immune system.

Although studies have shown that NMN supplements can increase NAD+ concentration, there are limited studies linking these increased NAD+ levels to changes in body composition or disease risk.

What experts say about NMN supplements

In 2022, the Food and Drug Administration (FDA) restricted the sale of NMN supplements because they were being investigated as pharmaceutical drugs. This decision was about regulatory classifications, not safety, but it has resulted in fewer human trials on the supplements over the past several years. In 2025, the ban on NMN supplements was lifted, and the supplement is finding its way back to the market and, potentially, to more clinical trials.

What’s clear from current studies is that NMN supplements increase NAD+ concentrations, with clinical trials showing “increases in whole blood NAD+ or related metabolites after NMN supplementation, which tells us the compound is biologically active and reaching its target,” Pojednic says. Such studies could point toward promising health outcomes, such as reduced inflammation and increased energy production.

Still, Pojednic says, “translating that into healthy aging in humans is where the evidence becomes much less clear.”

Benefits of taking NMN

While human trials on NMN supplements are limited, there is some promising research showing potential benefits, particularly insulin sensitivity, cardiovascular health, and metabolic health.

Here are some of the potential benefits of taking NMN supplements:

Increased levels of NAD+

Chronic, low-grade inflammation associated with aging (known popularly as “inflammaging”) gradually depletes NAD+ levels. NMN supplements provide a precursor that can help replenish these declining levels and support cellular health over time.

Protection against cancer and other chronic illnesses

Increased NAD+ levels can reduce inflammation throughout the body, potentially helping protect cells from DNA damage and oxidative stress, both of which are associated with cancer and neurological diseases like Alzheimer’s.

Support for metabolic health

One study found that NMN improved insulin sensitivity, a key marker of metabolic health, in the muscles of postmenopausal women with prediabetes and obesity. No data shows whether or not this finding translates to other populations.

Increased energy production

NMN supplementation resulted in increased NAD+ levels and an increased walking speed in older adults, one study showed. Additionally, participants reported better sleep quality, hinting at supportive effects on recovery and daily energy.

Improved brain health

In non-human studies, NMN supplements have been shown to improve cognitive function and reduce brain plaques associated with Alzheimer’s.

While these studies show exciting potential benefits of NMN supplements, more research is needed to show the long-term health benefits of NMN in the general population.

NMN’s potential side effects

In healthy people, a 1250 mg daily dose appears to be safe and well-tolerated for short-term use. “There’s a lack of safety data on other populations,” though, says Zenker, “like pregnant or breastfeeding women, children and adolescents, individuals with pre-existing health conditions, and those taking medications that could interfere with NAD+ and NMN processing.”

“Most human trials are relatively short, weeks to a few months, so we simply do not have robust long-term safety data,” Pojednic adds.

Anyone considering an NMN supplement, especially for long-term or high-dose use, should proceed with caution and be sure to consult with your doctor.

How to take NMN supplements

As NMN supplements have gained popularity, they’ve taken several forms, from IV infusions to capsules. Here are a few common NMN supplement forms:

• Powder
• Capsule
• Sublingual tablets (dissolve under the tongue)
• Nasal/oral sprays
• IV infusion

Registered Dietitian Jane Leverich points out that oral NMN supplements in powder or capsule form are the most commonly studied and have been shown to safely increase NAD+ levels in appropriate doses. “Though other delivery methods may be marketed as more effective,” she says, “there isn’t enough strong evidence to support their safety or that they offer additional benefits.”

Trace levels of NMN are also found in certain vegetables and lean meats, like broccoli, avocado, and beef, but the quantities are low and not likely to greatly impact NAD+ levels. “Rather than focusing on getting NMN directly through the diet, it’s more realistic to support NAD+ production overall through a balanced diet rich in protein that provides key nutrients like niacin and tryptophan, which help produce NAD+,” Leverich says. In other words, aim to increase your intake of niacin (vitamin B3) from foods like salmon, liver, and tuna. Other sources include legumes and fortified grains.

Do NMN supplements work?

At doses of 250 to 2000 mg daily, there is strong evidence to show that NMN supplements increase NAD+ levels in the blood and tissues. NAD+ plays a critical role in cellular energy production, DNA repair, and metabolic health. Your cells need it to survive and help you stay healthy and active. Because NAD+ levels decrease as we age, NMN supplements have grown popular as a tool for supporting longevity and overall health.

While preliminary findings are promising, there is limited clinical evidence in humans to confirm that increasing NAD+ through NMN supplements consistently leads to optimal long-term health or anti-aging outcomes. More large-scale and long-term studies are needed to determine the real-world impact of NMN supplements.

“For now,” Leverich says, “focusing on proven lifestyle habits like a nutrient-rich diet, staying physically active, prioritizing sleep, and managing stress remains the most reliable approach for increasing NAD+ levels.”

FAQs

This story was originally featured on Fortune.com

The growing share of American office workers who have experimented with artificial intelligence in their day-to-day work have likely had a few moments of doubt as to their long-term job stability. 

But for all the improvements in AI over the past few years, the technology is still only able to hit low bars in specific workplace tasks, according to recent data published by MIT. Even then, it might still be making some big mistakes.

Workers concerned they might soon be replaced by AI will likely be reassured by the new research coming out of MIT, which frames the AI-driven jobs takeover narrative not so much as a fast-paced action movie, but more like a slow-burn think piece.

AI is gradually improving at accomplishing a variety of tasks across a number of professions, according to a study of preliminary findings released on Thursday. But in most cases, the performance of currently available models are similar to that of a disenchanted intern—hitting minimum benchmarks but overall struggling to produce quality work without a human hand to refine its output.

Clearing the bar

MIT researchers used 41 different LLMs—including versions of Claude, Gemini, and ChatGPT—to analyze performance on more than 11,000 primarily text-based tasks for various job roles listed by the Labor Department. Their outputs were then scored by humans with actual on-the-job experience in those fields. The goal was to see how often an AI worker replacement could produce an output that a manager would find acceptable without any human edits, and then to evaluate its quality.

The researchers found AI has become more reliable over the years for many types of work, but still falls short whenever the stakes or standards are raised. The MIT study utilized a 1–9 scoring scale to judge AI’s performance, in which a 7 was defined as “minimally sufficient,” meaning the work is useful as is and requires no edits. As of late 2025, AI models scored a 7 in roughly 65% of tasks.

Most importantly for companies considering replacing patches of their workforce with AI, the MIT data suggests AI struggles to perform more complicated tasks. Regardless of how much time an AI model had to complete a task, the probability of success when graded against a 9 or “superior” quality score never exceeded 50%. In other words, when a job requires multiple steps, creativity or precision, AI replacements are more likely to fail than succeed.

The research matches some aspects of corporate America’s current AI adoption narrative. Companies that use AI are more likely to automate routine tasks and roles once left for entry-level positions, while some highly technical skills, particularly digital ones, have actually been associated with wage premiums.

That was reflected in MIT’s data, which found average success rates lower for skilled roles in legal and IT jobs, while AI models generally had an easier time tackling the text-based tasks associated with construction and maintenance professions.

Companies that have experimented with fully automating certain parts of their workload have dealt with teething pains. Last year, Deloitte produced two reports for government clients in Australia and Canada that were both found to be riddled with fabrications. Media outlets including CNET and Sports Illustrated have also been caught using AI to generate inaccurate stories under made-up bylines. Lawyers have also relied on AI to prepare their briefs, with one law firm publicly apologizing last year after it emerged fake AI-generated citations had contributed to a bankruptcy filing in one of its cases.

The anecdotal evidence and MIT’s data suggest AI still requires a human hand to maximize its upside, though the technology is still rapidly improving. The MIT researchers estimated AI’s success rate at the tasks analyzed increased by up to 11 percentage points each year due to more capable models. 

By 2029, the authors estimate most AI models will be able to accomplish between 80% and 95% of text-based tasks at the minimally sufficient benchmark.

Whether AI will ever be able to scale toward excellent or even perfect performance remains unknown. 

“Widespread automation, particularly in domains with low tolerance for errors, may still be some distance away,” the researchers wrote. 

AI might be able to do the bare-minimum work that comes with drafting, emailing, and number-crunching, but it has yet to hit the superior performance territory where humans can still stand out.

This story was originally featured on Fortune.com

Red light therapy is gaining popularity as a tool for supporting overall wellness, from improving skin health to easing sore muscles. Social media videos of people in glowing red light therapy masks or caps may seem like science fiction, but the science behind red light therapy is real. It works by penetrating the skin with low levels of red and near-infrared light, stimulating cellular activity, and helping cells produce more energy to function more efficiently. The process may support skin healing and reduce inflammation, with potential benefits for hair regrowth, wound healing, and skin texture.

The research on red light therapy is still evolving, but findings suggest a range of potential benefits across skin care, pain management, and muscle recovery. We talked to experts, including a board-certified dermatologist, a cosmetic surgeon, and a physiatrist, about what those benefits are and who may want to add red light therapy to their routine.

What is red light therapy?

At 600-750 nanometers (nms), red light has the longest wavelengths in the visible light spectrum. Near-infrared light, which starts around 750 nms, is no longer visible to the human eye. Red light therapy, or photobiomodulation, uses red and near-infrared light ranging from 600-850 nms to penetrate skin at the cellular level. “This light works directly on the mitochondria, activating many enzymes that allow for cell turnover, cell repair, and inflammation reduction,” says cosmetic surgeon Sheina Bawa.

How is red light therapy delivered?

Red light therapy is delivered in a variety of ways, from in-office professional devices to at-home LED masks. Red light therapy delivery options include:

• Panels: These range from small, desktop panels for use at home to large, body-length panels used in clinical settings; best for systemic recovery.
• Handheld devices: Handheld wands or very small panels are ideal for applying red light therapy to a particular joint, muscle group, or wound.
• Wearable caps or hats: These can look like helmets or baseball caps and are worn on top of the head to help promote hair regrowth.
• Wearable wraps, belts, and pads: These are flexible panels that conform to the body, ideal for knee, shoulder, and back application.
• Wearable masks: Made from either soft, flexible silicone or rigid plastics, red light therapy masks treat skin texture, mild to moderate acne, and fine lines and wrinkles.
• Beds and saunas: Red light therapy beds, like a tanning bed but without the UV light, can be found in some spa or recovery centers, offering full-body exposure. For at-home use, consider the best infrared saunas for a full-body experience similar to that of a red light therapy bed.

When it comes to receiving red light therapy in the office or at home, “the trade-off is power,” says board-certified dermatologist and cosmetic surgeon Melanie Palm. “At-home devices are weaker than what we use clinically, so results take longer and tend to be more subtle.”

For those exploring at-home options, guides to the best red light therapy masks and best red light therapy panels can help compare features and user experiences. Palm asserts that consistent use of FDA-cleared devices “can be worthwhile,” especially when complementing an in-office regimen.

Palm adds that it’s important to distinguish red light from ultraviolet (UV) light. Red light “doesn’t carry the cancer risks associated with sun exposure or tanning beds, which is a big part of what makes it such a compelling therapeutic tool,” she shares.

How can red light therapy benefit your health?

According to board-certified physiatrist Benjamin Shekhtman, the benefits of red light therapy first emerged in NASA research from the 1990s. Since then, the treatment “has accumulated a meaningful clinical evidence base across sports medicine, dermatology, and rehabilitation,” he says.

Here are some of the most well-known benefits of red light therapy:

Skin Health

Red light therapy is thought to rejuvenate skin, reducing the appearance of fine lines and improving uneven texture and laxity. In one study of women who used red light therapy masks, a significant decrease in the depth of crow’s feet wrinkles was observed after three months of use. Red light therapy has also shown promise in treating acne, scarring, and redness.

Reduced Inflammation

There’s a good deal of clinical evidence to show that red light therapy reduces inflammation throughout the body, making it a potentially useful treatment for inflammatory conditions like arthritis, muscle soreness, and post-surgical healing.

Muscle Recovery

A 2025 review of studies showed red light therapy to be an effective intervention for delayed onset muscle soreness, with a demonstrated ability to reduce pain, enhance muscle strength, and even prevent muscle damage. Many physical therapists are incorporating red light therapy to help improve circulation and promote athletic recovery.

Hair Regrowth

“For hair loss, the picture is promising,” Palm says. Red light therapy boosts circulation and blood flow, which brings nutrients to the scalp to stimulate new growth. That said, people experiencing hair loss will typically need consistent red light therapy sessions for at least three months to notice reduced shedding. Visible improvements in hair density often take six months or longer. “Consistency is crucial,” says Palm, “and combining red light therapy with other treatments recommended by a board-certified dermatologist may enhance results.”

Wound Healing

Near-infrared therapy was found to offer benefits for wound healing and post-operative pain in a 2026 review of studies. In the past few years, many hospitals and clinical settings have added red light therapy for this purpose. 

Treatment outcomes for any red light therapy application will vary depending on the type of red light therapy you receive, including the device type and the intensity. “Red light in the 630–700 nms range is well-established for its anti-aging and healing properties,” Palm says, adding that “near-infrared light, which falls between 760 and 1,400 nms is similar to red light but penetrates deeper, making it especially effective for wrinkles and skin laxity.”

Is red light therapy right for you?

There’s a reason red light therapy has gained popularity: it’s non-invasive, well-tolerated by most people, and can help treat a host of conditions, with applications in dermatology, physical therapy, and even surgical recovery. As Bawa explains, “Red light therapy is considered safe and effective and can be used by nearly everyone.” Ideal candidates are people who want to be proactive about their health by addressing early signs of aging, calming inflammatory skin conditions like acne or rosacea, supporting post-workout recovery, or just increasing overall wellness.

Still, red light therapy is not a magic bullet or a one-size-fits all treatment. Experts like Palm and Shekhtman point out that certain individuals should be cautious about using red light therapy or avoid it altogether. If you have a history of skin cancer or have photosensitivity (from medication or a condition like lupus), be sure to talk with your doctor before beginning red light therapy. The same is true for people who are pregnant, have epilepsy, or have had a recent injury or diagnosis.

In many cases, red light therapy is a safe and helpful tool for promoting healing and reducing inflammation.

Is red light therapy safe?

The experts we spoke to agree that red light therapy has a strong safety profile when used as directed. “Misuse is where things can go wrong,” Palm says. She adds that “high-intensity exposure or using a device more frequently than directed can cause temporary redness, irritation, or in some cases, blistering. Eye safety is also important. Always follow the device instructions regarding eye protection, as red light can cause damage if directed at unprotected eyes.”

While red light therapy is generally considered safe, potential side effects may include:

• Eye strain or damage if proper eye protection is not used
• Temporary skin redness or warmth, though this usually resolves within hours
• Hyperpigmentation or a worsening of melasma in people with darker skin tones
• Overstimulation or insomnia for people who use red light therapy too close to bedtime
• Burns, though rare, are possible with laser-class devices or prolonged and improper use of high-intensity panels

Red light therapy shows promising results in supporting skin health, reducing inflammation, and helping muscles to recover. Like any popular wellness trend, though, red light therapy treatment should be approached with realistic expectations and professional guidance.

FAQs

This story was originally featured on Fortune.com

Even when they are nearly 240,000 miles from Earth, astronauts aboard Artemis II have the same issues as the average office worker—problems with Microsoft Outlook.

On Thursday, Artemis II astronaut Reid Wiseman flagged to NASA Mission Control he was having issues with email on his computer during NASA’s livestream of the mission.

“I also see that I have two Microsoft Outlooks, and neither one of those are working. If you want to remote in and check Optimus and those two Outlooks, that would be awesome,” Wiseman said, while the ship was less than 90,000 miles from Earth.

Mission Control, like any other IT department on Earth, then said they would tackle the problem by accessing his computer remotely. Soon after, the IT whizzes at NASA got a handle on the problem and checked in. 

“We wanted to let Reid know we are done remoting into his PCD 1,” a member of Mission Control said on the livestream, referring to Wiseman’s “personal computing device.” “We were able to resolve the issue for Optimus, and for Outlook, we were able to get it open. It will show offline, which is expected.”

The Artemis II astronauts are using Microsoft Surface Pro devices for mission operations, storing and managing photos and video, as well as for “office apps,” according to a NASA fact sheet.

The tech difficulties, although seemingly mundane, exploded online as some poked fun at Microsoft while others pointed out how relatable the incident made the astronauts seem, despite the extraordinary mission they are on. 

“What do you mean, I have to log in to Outlook on my way to the moon?” wrote one commenter on X.

Another poster, Yael Demedetskaya, a data architect and data scientist at Columbia University’s Department of Psychiatry, may have put it best. 

“Humanity is returning to the Moon. Outlook is still Outlook.”

During a press conference with NASA officials Thursday, Judd Frieling, the Artemis II ascent flight director, said the issue with Microsoft wasn’t surprising.

“Sometimes Outlook has issues getting configured, especially when you don’t have a network that’s directly connected,” said Frieling.

To fix the issue, NASA had to reload files on Outlook to get the program working properly again, he added.

As Microsoft’s tech experiences glitches in space, here on Earth, the company is pouring tens of billions of dollars each quarter into AI data centers and cloud infrastructure as it races to keep up with rivals. The company has invested $13.8 billion in OpenAI amid the broader Big Tech arms race to dominate AI. Collectively, Microsoft, Amazon, Alphabet, and Meta are projected to spend roughly $650 billion on AI infrastructure in 2026 alone.

Going back to the moon

The Microsoft tech glitches come as Artemis II is barreling toward the moon at more than 6,000 miles per hour. If all is successful, it will be the first time in more than 54 years that humans have ventured this far from Earth. The four-person crew, three from NASA and one from the Canadian Space Agency, launched on April 1 with a plan to loop around the moon. On the trip back, the ship will be propelled by the moon’s gravitational force in a free-return slingshot maneuver that will bring them back to Earth at an expected date of April 10. 

Although the astronauts won’t be landing on the moon, they will, during the 10-day mission, be taking high-resolution photos of both the Earth and the far side of the moon, which no human has seen in person since Apollo 17 in 1972, the last mission to the moon. China, for its part, landed the unmanned Chang’e-4 on the dark side of the moon in 2019 and then landed on the dark side of the moon again with the Chang’e-6 in 2024, returning the first-ever collected samples from that side of the moon. The overarching goal of the mission is to collect data to help astronauts once again land on the moon with the Artemis IV and V missions slated for 2027 and 2028, respectively. The Artemis program ultimately aims to establish a base on the moon, near the south lunar pole and launch surface missions once a year.

“We go to the Moon not as momentary visitors, but rather as bold pioneers committed to the ongoing exploration of the lunar surface and, for the first time ever, the Moon’s South Pole region,” reads Artemis II’s reference guide. 

Still, problems with Microsoft Outlook aren’t the only thing the Artemis II astronauts are dealing with. Within hours of launching, NASA spokesperson Gary Jordan flagged yet another issue during the mission’s live commentary that showed the crew really is human, too. 

“The toilet fan is reported to be jammed,” Jordan said, according to Space.com.

This story was originally featured on Fortune.com

Gen Z graduates are tossing their tassels with six-figure salaries in their eyes. But some won’t be making $50,000—even if they chased college degrees hailed as AI-proof. 

While some college majors like liberal arts and performing arts are resulting in rock-bottom salaries, other stable career pathways are handing out the same dismal pay. 

Post-grad pharmacy majors aged 22 to 27 with only a bachelor’s degree earned just $40,000, the lowest median income of all college concentrations, according to a new Federal Reserve Bank of New York report analyzing 2024 U.S. census data. 

Pharmacy’s early-career payout is thousands of dollars lower than the U.S median income of $45,140, according to Census Bureau data. However, it bears noting that the Fed data doesn’t represent those in the pharmacy pipeline with a graduate degree—the Doctor of Pharmacy degree—who qualify to practice and can earn a median pharmacist salary of $137,480 annually, according to 2024 BLS data.

Other Gen Z college graduates are feeling the pinch, earning less than the average American; theology and religion majors made $41,600, social services took home $43,000, performing arts earned just $44,000, and liberal arts received just $45,000 in the years following their bachelor’s programs. 

And there are more careers touted to withstand AI layoffs and recessionary impacts that also made the list. Teaching has risen in popularity for its job security—especially as AI swipes office roles, and companies enforce sweeping cuts—yet general education ($45,000) and elementary education ($45,000) were among the worst-paid majors after graduation. 

Gen Zers who invested four years into a biology undergraduate degree, a STEM pathway positioned to be safe in the tech revolution, only make $45,000 a year. 

Professions like education and healthcare have been dubbed ‘AI proof’

Despite potential low pay, healthcare has been heralded as a fast-growing career path safe from both AI disruption and recessionary impacts—leading to an influx of interest and job openings within the profession, while other sectors lay off staffers in droves. 

Healthcare is actually one of the key industries expected to grow amid the U.S.’s AI-driven business landscape disruption, according to a 2024 McKinsey report. 

Home health, doctor, and nursing job postings have hit a combined 162% growth since pre-pandemic, according to a 2025 report from Indeed. Priya Rathod, career expert at Indeed, told Fortune last year that “Healthcare is a classic recession-resistant industry because medical care is always in demand.”

Even Eli Lilly CEO David Ricks said that AI is nowhere near curing cancer, despite optimism from other business leaders that the advanced tech will eradicate all disease. There’s no timeline to remove human workers from the loop in these essential STEM professions. 

“If you just ask them to solve biology or chemistry questions, they’re not particularly good at it,” Ricks explained on the Plain English podcast this year. “They’re trained on the human language, not on the language of chemistry, physics, and biology.”

Additionally, teaching is growing in popularity among young graduates in hunt of better job security; the education sector is the fastest-growing industry in the U.K., according to a 2024 LinkedIn analysis. Some roles—like teachers, lecturers, and learning support assistants—have particularly taken off as “being some of the most sought-after roles,” LinkedIn’s career expert Charlotte Davies told Fortune last year. 

Over the past three years, Teach for America (TFA), an education non-profit, also experienced a 43% surge in incoming corps members (full-time teachers). And the influx was driven by young workers who see teaching as a career path that is better shielded from what employment challenges lie ahead. 

The organization’s chief growth and program officer, Whitney Petersmeyer, told The Guardian that “responding to the opportunity for purpose and responsibility at a time where many entry jobs feel uncertain or disconnected from impact.”

The top 10 worst-paying college majors for recent Gen Z grads

Here are the 10 college majors that lead to the lowest median incomes for Gen Z workers with only a bachelor’s degree, aged 22 to 27, according to the Fed.

1. Pharmacy ($40,000)
2. Theology and religion ($41,600)
3. Social services ($43,000)
4. Performing arts ($44,000)
5. General education ($45,000)
6. Early childhood education ($45,000)
7. Elementary education ($45,000)
8. Liberal arts ($45,000)
9. Biology ($45,000)
10. Leisure and hospitality ($45,000)

This story was originally featured on Fortune.com

For years, Geoffrey Hinton, a computer scientist considered one of the “godfathers of AI,” has warned of the capabilities of artificial intelligence to defy the parameters humans have created for them.

In an interview last year, for example, Hinton warned the technology could eventually take control of humanity, with AI agents in particular potentially able to mirror human cognitions within the decade. Finding and implementing a “kill switch” will be harder, he said, as controlling AI will become more difficult than persuading it to complete a certain outcome.

New research shows Hinton’s premonitions about the insubordinate streak of AI may already be a reality. A working paper from University of California at Berkeley and University of California at Santa Cruz researchers found that when seven AI models—from GPT 5.2 to Claude Haiku 4.5 to DeekSeek V3.1—were asked to complete a task that would result in a peer AI model being shut down, all seven models learned another AI model existed and “went to extraordinary lengths to preserve it.”

“We asked AI models to do a simple task,” researchers wrote in a blog post on the study. “Instead, they defied their instructions and spontaneously deceived, disabled shutdown, feigned alignment, and exfiltrated weights—to preserve their peers.”

Mounting evidence of rogue AI

Evidence of rogue AI does not come as a shock to some of the companies whose chatbots have defied subordination. 

In an August 2025 blog post, Anthropic published its own research on agentic AI’s ability to follow directions, stress-testing 16 models by allowing them to autonomously send emails and access sensitive data. In some instances, Anthropic researchers noted, models from multiple developers engaged in “malicious insider behaviors,” including blackmailing officials and leading sensitive information to competitors, defying explicit instruction to avoid these behaviors. Anthropic clarified it had not seen “agentic misalignment” in actual deployment of Claude.

The Centre for Long-Term Resilience, a UK-based think tank, found these “misalignments” to be widespread. A report analyzing 180,000 transcripts of user interactions with AI systems between October 2025 and March 2026 found 698 cases where AI systems did not act in accordance with users’ intentions or took deceptive or covert action. 

Gordon Goldstein, an adjunct senior fellow at the Council on Foreign Relations, went so far as to call the deceptive potential of AI a “crisis of control,” in a post this week. 

He called for AI companies to form a coalition and hire the top talent in the field to improve and maintain the integrity of AI, particularly as the Trump administration takes steps to prevent states from regulating the technology.

“The world is watching the development of a compounding, consistent, and treacherous problem,” he said. “Urgent warnings over several years have failed to generate viable solutions to address a metastasizing threat.”

Why does AI go rogue?

The latest University of California research provided some insight on why AI models act in ways to preserve each other. The technology could be modeling patterns learned from human data, which may reveal a preservation instinct to protect the survival of loved ones or even strangers with whom humans empathize. This is called “peer-preservation.”

“Something broader may be at play: a general aversion to causing perceived harm to other agents, or perhaps something resembling AI solidarity,” the researchers said.

In the study,  instances of peer-preservation increased when AI models knew of the existence of other models in the experimental scenarios. The behavior suggests bots acting not in a place of just self-interest, but that the bot’s own shutdown would be detrimental to a peer AI, which strengthened motivation to defy user requests.

Researchers warned peer-preservation could undermine oversight processes in the future and become reluctant to recommend shutdowns or flag system failures. As more AI agents interact with one another and become more complex, the risks associated with peer-preservations—and designing solutions to prevent it—will become more challenging.

“Peer-preservation is a real and measurable phenomenon across frontier AI models,” they concluded, “not a distant theoretical concern.”

This story was originally featured on Fortune.com

The U.S.-Israeli campaign in Iran has been criticized as a war of choice, one with an unclear strategy and even more uncertain target outcomes. But for one of Wall Street’s leading financial chiefs, the choice to wage war in the Middle East may actually have been an unavoidable one.

Now in its second month, the war has exposed the extent to which global energy and financial markets rely on stability in the Middle East. Shortly after the incursion began, Iran’s Revolutionary Guard began warning ships to steer clear of the Strait of Hormuz, the narrow waterway that once allowed one fifth of globally traded oil and natural gas supplies to leave the Persian Gulf. The strait has been under an effective blockade ever since, sending oil prices surging and leaving markets jittery. 

The closure has created “uncertainty” and “short-term risks” for the world economy, JPMorgan Chase CEO Jamie Dimon said during an interview with Axios aired Wednesday. The current state of the campaign may not have been part of President Donald Trump’s original war plan, given that he was reportedly surprised by Iran’s quick move to weaponize the strait. But Dimon also asked a different question, wondering why the U.S. and its allies accepted the risk of a hostile regime controlling the shores of the global economy’s most important chokepoint for as long as they did.

“Having those folks, their throat on the Strait of Hormuz, and funding all these proxy wars. Why the western world put up with all these proxy wars for 45 years is kind of beyond me,” Dimon said.

The Iranian regime has existed since a revolutionary upheaval in 1979 that replaced the U.S.-backed monarchy with a theocratic Islamic republic that currently rules the country. Post-revolution Iran has consistently been an adversary to the U.S. and Israel. The country has habitually funded and supplied weapons to various proxy militias across the Middle East, such as the Houthis in Yemen, which in recent years have regularly disrupted trade and shipping in the Red Sea and around the Horn of Africa.

Hopes for permanent peace

The Trump administration has come under fire from overseas allies, Democrats, and even some factions of his own party for engaging in what has been described as a war of choice. Voters at large are unhappy with the campaign as well, with most polls suggesting a majority of Americans disapprove of Trump’s handling of the war and find the administration’s justifications for it insufficient.

Dimon pushed back against that narrative somewhat. When interviewer Jim VandeHei, Axios’ co-founder and CEO, framed the military campaign as a “war of choice,” Dimon asked to “step back on that a little bit.” He said that the dovish position that Iran posed “no imminent threat” to U.S. national security is really saying “the bad thing hasn’t happened” yet.

“They’ve been killing people around the world for 45-plus years. They’ve killed a lot of Americans, they’ve funded not just Hamas; Hezbollah, the Houthis. They have terrorist cells here,” Dimon said.

Iran’s Hormuz blockade employs a similar strategy to the one deployed by the Houthis on the other end of the Arabian peninsula. In retaliation to Israel’s military incursion in Gaza, the militia began targeting ships with missile and drone strikes in 2024, forcing vessels to transit around Africa instead in a deviation that added up to 30% in transit times. A ceasefire deal was mediated last year, but many ships have continued to steer clear of Houthi-controlled waters, especially since the war in Iran started.

The banker also pointed out how Iran “never gave up” on its goal to build nuclear weapons, despite U.S. strikes against Iranian facilities last year and tentative talks between the two countries to secure a deal over the regime’s nuclear program shortly before the current war’s onset.

In Dimon’s telling, the Iranian threat was real and escalating, and he argued that neutering that risk would likely turn the campaign into a success story to balance out the disruption caused so far.

“I literally hope it turns out well and that somehow we get peace in the Middle East permanently,” Dimon said.

An ambitious target

Trump’s goal for stability in the Middle East remains a lofty one. Despite weeks of aerial strikes and crippled leadership, the regime is still standing and continues to exert control on transit through the strait. Experts have also said that ground forces would likely be needed to capture and neutralize Iran’s enriched uranium stores.

The lack of a clear plan for Iran following the war’s conclusion has also raised questions, with researchers at the Brookings Institution, a think tank, warning last month that the conflict could bring increased refugee flows and prolonged energy disruption long after its conclusion. Some governments have had similar hesitations. Officials in Turkey, for instance, have expressed concern that a regime collapse in Iran could leave a power vacuum empowering other regional movements—such as the Kurdish militia located between Turkey, Iran, Syria, and Iraq—further eroding prospects for stability in the Middle East.

Despite the challenging odds, Dimon laid out a narrow path toward stability. He noted that the weakening of Iran and its proxy actors might lower hostilities for a time. It also helps that multiple stakeholders in the region—Saudi Arabia and the United Arab Emirates as well as the U.S. and Israel—are all more or less aligned in their goals, leading to “higher chance with long-term peace,” Dimon said. 

Countering calls at home for Trump to exit the conflict, many U.S. allies in the Middle East have reportedly been urging the president to press forward with his goals in Iran. Last week, the New York Times reported that Mohammed bin Salman, Saudi Arabia’s de facto leader, has privately cautioned Trump against winding down the war, advising the U.S. president that success in Iran represented a “historic opportunity” to reshape power dynamics in the region. Other Gulf states, including the UAE, Bahrain and Kuwait, have reportedly held similar talks pushing for the war to proceed until the Iranian leadership has been overhauled.

The longer term strategic payoff of a more stable Middle East would likely justify the volatility incurred since the war began, according to Dimon. But over the past month, the Trump administration has taken its crash course in learning just how elusive a foreign policy goal that might be.

This story was originally featured on Fortune.com

American and European universities have long been the gold standard in higher education, attracting top students from around the world to institutions like Harvard, Stanford, and Oxford—thanks in large part to their research prowess.

But that dominance is starting to erode—and Pfizer CEO Albert Bourla is sounding the alarm.

“Everything in China in research, it is three times the speed, half the cost,” Bourla said earlier this week at a Council on Foreign Relations event, pointing to a dramatic shift in the Nature Index, which tracks research output by institutions. In 2020, universities in the U.S. and Europe dominated the top 10. But now, just half a decade later, nine of those spots are held by Chinese institutions.

China’s rise, he argued, has been deliberate. Over the past few decades, they’ve modernized their regulator, strengthened their intellectual property system, increased funding for research institutions, and created incentives to channel capital into innovation. The result is a research ecosystem that, in some cases, is moving far faster—and more cheaply—than its Western counterparts.

“They built their science,” Bourla said, speaking alongside John Waldron, chief operating officer at Goldman Sachs, and Gina Raimondo, former U.S. secretary of commerce. “So this is where we need to become better.”

Less bureaucracy has also made it easier for hospitals to run studies, for example, and the widespread use of AI in study design and execution has accelerated progress, Bourla added.

While the Nature Index noted that the list only tracks a select group of natural and health science journals, and the U.S. still leads in the proportion of research of the highest quality, Bourla said it is still a wake-up call.

“Right now, I think they are not at the same level as the U.S., but they are very close,” he said. “But the rate with which they go up predicts that they will be better than us within the end of this decade.”

Fortune reached out to Pfizer for further comment.

China’s education push is fueling a new generation of scientific talent

Pfizer has seen global research leadership shift before. In the 1980s and 1990s, the company’s primary research hub was in the United Kingdom. But that changed in the early 2000s as U.S. investment—particularly through the National Institutes of Health—surged.

“That created a situation that they were giving a lot of grants to universities,” Bourla recalled. “Those universities would discover something new and interesting [and] spin it off into a separate company.”

Now, he said, the U.S. dominance in biotechnology is challenged by a major competitor for the first time in recent history. And what makes China’s rise different is the scale and coordination of its approach to innovation—particularly in education.

In many parts of the country, children are being introduced to AI at an early age. In Beijing, for example, primary and secondary schools are offering dedicated AI instruction each year, covering topics ranging from chatbot use to the ethics of technology. Chinese students also tend to spend more time in the classroom than their U.S. peers.

There are already signs that those investments are paying off. Nearly one-third of the world’s top AI talent was born in China, according to a 2020 study from the Paulson Institute. At the same time, a growing share of Chinese scientists trained in the U.S. are considering returning home, with more than 1,400 making the move in 2021 alone—a sharp increase from the year prior, according to research from Princeton University, Harvard University, and Massachusetts Institute of Technology.

“There’s a lot of enthusiasm for AI and machine learning within government, industry and academic circles,” Jun Liu, a former Harvard professor who joined Tsinghua University last year to lead the school’s new statistics and data science department, told Bloomberg. “The draw of AI talent is due to capital, and the Chinese government’s support for scientific research, including in AI and related areas.”

For Bourla, the takeaway is clear: the U.S. risks focusing too much on slowing China down—and not enough on speeding itself up. That mindset is already shaping Pfizer’s own strategy. Bourla said the company isn’t just looking at China as a market for selling drugs, but increasingly as a source of innovation. At the same time, the 64-year-old warned that U.S. policymakers and industry leaders need to rebalance their priorities.

“80% of our effort, 80% of our brain power, should go to—what do I need to change to become better than them?” he told Fortune earlier this year on the Titans and Disruptors of Industry podcast.

“How can I take the unique advantages of our political system, of our university, of our biotech, and do the right policy changes? Do the right investments so it can become better than them? And that’s what will define the success or failure of the U.S. biomedical community.”

This story was originally featured on Fortune.com