• A key Google DeepMind engineer is leaving for Anthropic.
• He’s the latest Silicon Valley name to join the AI giant.
• John Jumper won the 2024 Nobel Prize in Chemistry for his work with Google’s AlphaFold technology.

John Jumper, a chemist and computer scientist who once won a Nobel Prize alongside Demis Hassabis, announced Friday he is leaving Google DeepMind to join Anthropic.

He’s the latest Silicon Valley name to jump ship for the AI startup darling. Jumper worked at Google for nearly a decade.

“The entire GDM team taught me so much about how to do great science,” he wrote in an X post on Friday. “GDM is a special place, and I’ll still be excited to hear about what amazing things they discover next.”

Jumper is best known for spearheading Google’s AlphaFold team, an AI system that predicts a protein’s 3D structure from its amino acid sequence.

The AI-powered technology was a major breakthrough in medical and biological research, enabling scientists to understand protein design in greater detail. Now offering over 200 million protein structure predictions, it’s a resource that cuts months and even years from the research process.

Hassabis, the cofounder and CEO of DeepMind, said that Jumper’s work with Alphafold will have a lasting legacy.

“What we achieved with AlphaFold changed the world, and showed the field what was possible with AI for science and medicine, lighting the way for how AI can benefit humanity,” he wrote.

Hassabis and Jumper shared the 2024 Nobel Prize in Chemistry for their work developing AlphaFold.

Jumper’s exit to Anthropic coincides with a series of tech leaders at companies like Google and Meta leaving for the big AI startups. Firms like Anthropic and OpenAI are the hottest tech job around — especially as both prepare for an initial public offering.

Jumper did not say what his role will be at the AI giant. First, he said, he will be “taking some time to recharge.”

• Bo Pettersson runs his socials with the help of his daughter, Emily.
• Emily suffered a brain injury playing soccer, which changed both their lives.
• Their content helps support treatments for Emily’s injury.

This as-told-to essay is based on a conversation with Bo and Emily Petterson, of Dad Advice From Bo. It has been edited for length and clarity.

Parents aren’t supposed to have a favorite child, but there’s no secret that Emily has always been mine. She’s the second youngest of my six children. She’s been the boss of our household since she was 3. She was the type of child who you’d look at and think, “This kid is going places.”

Emily got a scholarship to college. Then, in 2009, she was playing soccer when she and another player both jumped to head the ball and collided. From that moment, Emily’s life changed forever.

At the time, my wife and I had a “shake it off” mentality. No one was talking about concussions back then.

We were horrified as we slowly watched Emily’s life fall apart. Today, Emily describes her injury as being seasick 24/7. When Emily has a bad spell, she stays near the toilet, vomiting for days. There’s nothing I can do except hold her.

I never wanted to be a TikTok star

In the years since Emily’s injury, her world got smaller and smaller. First, she lost control of her body, then she lost her friends, her social life, and her ability to work. By 2020, she’d also lost hope. Emily was severely depressed, and her mother and I were worried.

Emily had mentioned making videos with me, but I had always brushed her off. I’m not the type of person who wants to put himself out there. I didn’t even know what TikTok was at the time. But I could see that Emily was desperate to have something positive in her life. So, one day when she asked, I reluctantly agreed.

Emily prompted me to give Dad advice. I thought about the things my own father had taught me and settled on a simple tip: how to back up a trailer. Emily posted the video, and the next day we were amazed by how many people had seen it. As we made more videos, I could see that our channel was pulling Emily out of the depths of despair.

I enjoy hearing about the real impact my advice has

Usually, our videos are based on something I’m already doing around our home and land. I’m shy, so I try not to think about the fact that millions of people watch me — I just speak like I’m talking to my own kids.

I love seeing the comments and messages our audience sends. One woman messaged to say that when her dishwasher sprang a leak, her husband was running around trying to figure out how to turn off the water. The woman remembered I’d posted about just that, and she saved the day.

Another time, I encouraged viewers to pick up the phone and call their fathers. I’d be thinking about my own dad, and the conversations we’d had during his final years. A bit later, a viewer messaged us to say that she had reconnected with her estranged father because of that video.

The channel has helped Emily access treatment and community

A good dad is always there for his family, but he’s not perfect. I’ve always been clear about that with my kids, and I make a point of saying it on social media. I make mistakes. As a parent, you’re going to mess up millions of times. The best you can do is always try to make the next right decision.

I made a mistake with how I initially responded to Emily’s injury. But deciding to get onto social media with her was definitely the right decision. We have so much fun together making videos. We often laugh until we’re on our knees. It’s been amazing to have this much time with my daughter.

It’s also helped Emily directly. When we started, she didn’t know anyone else with a traumatic brain injury. Now, she has a network. Because of funds and connections on social media, Emily has had access to new treatments. We started making content because we were desperate. Now, because of everything that content has given us, we have hope.

• Chris Rojas barely touched alcohol until he entered the restaurant trade and drank with colleagues.
• His drinking got worse as he got older, and he had bad hangovers after consuming beer and whisky.
• The dad of four finally realized he wasn’t being present for his children and got sober.

This story is based on an interview with Chris Rojas, 41, of Tucson, Arizona, where he works in food distribution. It has been edited for length and clarity.

Some of my friends growing up were into recreational drugs and drinking, but I was big into sports and stayed away from all that.

Then, at 18, I got my first job in the restaurant industry. My colleagues and I would work hard all evening and then party when we were done.

We’d hit the bar scene around 10 pm or go to someone’s house. I drank heavy beers, whisky, tequila, gin, vodka, and anything else I could get my hands on.

I used alcohol to soothe myself

Then I got hit by a drunken driver. I was left with a lot of soft tissue damage, and I couldn’t stand up for more than 30 minutes for about a year.

There was a lot of pain involved, and I used alcohol to soothe myself. I got married and divorced within two-and-a-half years, which didn’t help.

Still, I met my second wife, and we got married in 2013. It wasn’t long before our first child arrived. We went on to have three more.

I didn’t go out so much because I had kids, but I would drink whisky and beer in my home office after helping them with their homework and eating dinner.

I was a bit of a workaholic with several side projects that I was passionate about. I’d stay up late working and drinking every night or every other night.

Alcohol dominated my thoughts. I’d drive to the supermarket to buy 12-packs of beer as soon as my supply got low in the fridge.

Stress got to me

The hangovers were awful, and I felt slow and sluggish. I was constantly tired from lack of sleep and wasn’t as connected to my little ones.

In 2022, I tried to reduce my drinking and was successful for about 18 months. Then I was demoted twice over the course of six months and earned 45% less than my previous salary.

The stress got to me, and I suffered from my prior injury to my back. My solution was drinking again.

Then, on November 15, 2024, I woke up with yet another hangover after a binge. I finally realized I couldn’t keep punishing my body.

I was giving up my energy and attention to my kids to a substance that was parasitic. I downloaded an app that offered advice on both reducing and eliminating alcohol consumption.

It was a life-changer, and I haven’t touched a drop since then. I read about the neurological and physiological effects of long-term abuse and what it was doing to my relationships.

I feel so much closer to my children

I created a routine of waking up, lifting weights at home, and committing to not drinking.

Unfortunately, my second marriage ended. It was traumatic, but it would have been 10 times worse if I’d been drinking.

Best of all, I feel so much closer to my children. I spend a whole lot more time with them and am fully present.

They’re young, but I’ve talked to my kids about my past through a filter. My 11-year-old recently told me that she was proud of me. It felt like the ultimate praise.

• Yale Budget Lab researchers are tracking the impact of AI on the job market.
• They found that AI isn’t causing major unemployment, and its impact is similar to the internet’s early days.
• The new technology is changing the nature of work, making companies reevaluate their bottom lines.

It’s easy for job seekers to feel stuck in an AI doom loop.

Chatbots have overhauled white-collar 9-to-5s, agents are rewriting the rules for basic tasks, and C-Suite leaders can’t stop talking about productivity gains. If you’re struggling to land a new role right now, however, new research found that technology likely isn’t the main culprit.

An analysis by Yale Budget Lab found that AI has had a modest impact on America’s job market since the release of ChatGPT in 2022. So far, AI has changed jobs more than it has eliminated them, the researchers said — a pattern similar to the impact of other major advances, such as the internet and computers.

Yale’s team put it bluntly: AI usage has “no connection” to changes in employment or unemployment.

AI is changing work, but not eliminating it

While it may not be obliterating jobs just yet, AI has undoubtedly changed the nature of those jobs. Business Insider has heard from Americans without a tech background who vibecoded solutions to their biggest problems, and business leaders who are using chatbots to streamline their workflow.

A solid benchmark for AI’s actual impact on jobs, Yale’s researchers found, is to compare it with other tech advances, such as the introduction of computers in the 1980s and the dawn of the internet in the 1990s. AI’s effect is slightly sharper in the months after launching, but not the work revolution some Silicon Valley leaders have heralded.

Some sectors have been hit harder than others. Finance and business are more vulnerable than a profession like nursing. Occupational churn, which measures growth and decline in the job market, however, is following a similar trend line to these other moments in tech history — not causing a massive reset.

The Yale report also found that high AI exposure doesn’t have a stark impact on how long job seekers are unemployed — those who have been out of work for less than 5 weeks have a relatively similar trend line to those who have been unemployed for 27 weeks or more. The number of unemployed workers whose jobs were automated is also fairly static.

It’s not to say that the job market is rosy. A lack of vacancies, widespread hiring freezes, and layoffs — which some CEOs say are somewhat related to AI — have boxed people of all ages out of offices. And relatively low quit rates mean that open positions have been few and far between. Jobs numbers are recovering a bit this summer after months of disappointing results, though that dip may have had more to do with high interest rates than tech disruption.

Giants like OpenAI and Anthropic are also reevaluating how they price their products, meaning companies will have to shell out a lot more money if they want their employees to use AI regularly. And, as Business Insider has reported, much of the current AI use in the corporate world isn’t translating to major profits or productivity gains.

It’s still early days for chatbots at the office, and the tech is evolving rapidly. But, at least for now, it’s unlikely to cause a sudden wave of unemployment.

• The Education Department announced a one percentage point interest-rate reduction for borrowers who enroll in autopay.
• Borrowers who are already enrolled in autopay do not need to take any action.
• It comes just weeks before Trump’s sweeping student-loan changes take effect.

Want a lower interest rate on your student loans? Enroll in autopay.

On Thursday, the Department of Education announced that federal borrowers who enroll in autopay by September 30, or who are already enrolled, will receive a one percentage point interest-rate reduction through June 30, 2028.

The reduction provides moderate savings: A graduate program borrower with $50,000 in student debt and a 7.94% interest rate could save nearly $23 per month over the two-year period.

Autopay is a feature available to all federal borrowers that allows the servicer to automatically deduct their monthly student-loan payment from their bank account. Borrowers currently enrolled in autopay receive a quarter percentage point interest-rate reduction.

“This interest rate reduction will help borrowers as they consider new, affordable repayment plans and work to repay their loans on time,” Undersecretary of Education Nicholas Kent said in a statement. “We expect this temporary incentive to drive up repayment rates and significantly improve the overall health of the federal student loan portfolio.”

According to the department’s press release, borrowers who are already enrolled in autopay do not need to take any action — their servicers will reduce their interest rate by an additional three-quarters of a percentage point. Borrowers in default are not eligible for the benefit until they return to good standing.

Interest rates on federal student loans range from 6% to nearly 9%. High interest rates have been a primary reason for surging student-loan balances; if a borrower does not maintain consistent monthly payments, their balances will grow due to interest, sometimes exceeding the amount they originally borrowed.

This announcement comes just weeks before President Donald Trump’s sweeping student-loan repayment overhaul will take effect on July 1. The changes include new borrowing caps and new repayment plans, including the Repayment Assistance Plan, which waives unpaid monthly interest. Borrowers are expected to still see their monthly payments increase on that plan, some by hundreds of dollars.

Have a story to share about your student loans? Contact this reporter at asheffey@businessinsider.com.

• OpenAI says more than 230 million people use ChatGPT for health and wellness advice each week.
• Researcher Karan Singhal leads the effort to grow that number by improving the company’s technology.
• Singhal spoke with Business Insider about his work in providing medical context in ChatGPT.

OpenAI is pushing further into its health research as more people turn to ChatGPT for pressing medical questions.

More than 230 million people use the tool for health and wellness advice each week, according to OpenAI. That growth is partially thanks to researcher Karan Singhal, who spoke exclusively with Business Insider about the company’s lofty healthcare ambitions.

Singhal leads a high-stakes goal: make ChatGPT so good on health that it changes people’s lives for the better, avoids calamity, and sways the skeptics. He wants to aid a shift he already sees underway, in which more patients trust OpenAI’s latest model as a “protector in their care journey.”

OpenAI’s GPT-5 model family is the company’s first to be trained specifically at every stage of development to be better at health advice, he said.

“You definitely want the models to be ahead of everything else,” Singhal said.

At OpenAI, healthcare has grown into a top priority

Before joining OpenAI, Singhal made his name as a researcher at Google, helping develop a series of AI models known as Med-PaLM, specifically designed for medical questions. Since then, Google has cut investment in Med-PaLM, Singhal said, because AI developers favor general-purpose models.

In the middle of 2024, when Singhal joined OpenAI, GPT-4o was the company’s flagship model. It would later come under fire in lawsuits alleging that it had encouraged suicidal ideation and given harmful advice. Those lawsuits are still unfolding — OpenAI has denied liability and wrongdoing.

In the meantime, the company hasn’t shied away from health-related use. In fact, it has dug in further.

Singhal said that when he joined, he felt a “responsibility” to improve the quality of the models’ health answers. He quickly set about building a new team of health researchers, and kicked off partnerships with more than 200 physicians — a bet, as he put it, on “aggregating the wisdom of the crowd.”

About a year later, he helped launch HealthBench, a series of evaluations that the company created with the physician group to measure AI systems’ health capabilities.

“Once you know how to evaluate it, it becomes a lot easier to improve it,” Singhal said.

OpenAI’s latest free model, GPT-5.5 Instant, scored better than both physician-written answers and GPT-4o in tests, the company said Thursday. Comparing billions of anonymized messages about health, they also said they found a 71% drop over the last two months in responses that were flagged for inaccuracy.

There’s pressure to keep those improvements going for both patients and clinicians who use the tools. Singhal said he’s seen doctors rapidly adopt ChatGPT for Clinicians and other AI tools, and he doesn’t feel that hospitals and clinics are resistant to AI.

“If you think about the adoption of technology in healthcare broadly, it’s actually incredibly, insanely fast,” Singhal said.

Singhal wants ChatGPT to get to know you better

Google Search has, for years, been the dominant peddler of healthcare information online, connecting users to websites like WebMD. Singhal sees chatbots as an upgrade, where back-and-forth conversations give people more specific advice.

One of the biggest challenges to getting valuable health information from a chatbot is how little it knows about the patient. A doctor might have your medical records in their hand or know you from a yearslong relationship. OpenAI is trying to simulate both.

For one, in January, the company announced a health-focused product within ChatGPT that connects to health apps and lets users upload medical records. Singhal gave the example of uploading his sleep data from his Apple Watch. He let the app analyze it, and learned that he was missing out on deep sleep because his bedroom was too warm.

ChatGPT Health still has a waitlist more than five months after its launch.

Singhal’s team sees the effort to make an AI model seek additional information as a top priority: a chatbot should ask questions like a doctor would, so it can say the right thing.

The team also wants to make the case that AI can bring value to everyone in health, not just “power users.”

“People’s adoption will only move at the speed of people’s readiness in practice, and so you have to guide people towards that, especially as the technology improves,” Singhal said.

Have a tip? Contact this reporter via email at scouncil@businessinsider.com, or over text, Signal, Telegram, or WhatsApp at 415-757-8198. Use a personal email address, a nonwork WiFi network, and a nonwork device; here’s our guide to sharing information securely.