Technologies
I Was Wrong About Workout Buddy, Apple’s AI Coaching Vision Is Just Getting Started
I tested Apple’s new WatchOS 26 feature and spoke with the team behind it. It’s not a full-blown coach yet, but it hints at where Apple may be heading with its AI health strategy.
I wasn’t exactly sold on the idea of having a “buddy” cheering me on during a workout when Apple first announced its Workout Buddy feature in WatchOS 26. The workout partner I had in mind was more of a no-nonsense trainer; someone to push me out of my comfort zone and into peak performance. Apple’s version, instead, is an AI-powered voice that dishes out praise as you run (or as you do any number of activities). But after testing it myself and talking in-depth with Apple about how it works, I’m starting to think the company undersold it on purpose. Workout Buddy is much more than just a hype man, and it represents a crucial turning point in Apple’s health journey.Â
Workout Buddy turns your Apple Watch into a friendly voice that lauds your achievements during workouts. The WatchOS 26 feature isn’t a full-blown fitness coach that can guide your training plan, but that doesn’t mean Apple isn’t headed there. In fact, Workout Buddy may be the clearest signal yet that Apple is laying the groundwork for something much bigger in terms of fitness tools and AI.Â
At a time when investors and enthusiasts are critical about the company’s AI efforts, especially compared to competitors like Samsung and Google, Apple has a chance with Workout Buddy to show how its approach is different in a meaningful way. After the lukewarm debut of Apple Intelligence on the iPhone, Workout Buddy is the first time we’re seeing what Apple can do with AI on the watch.Â
The public beta version of WatchOS 26 is out. I recently tested Workout Buddy in the developer beta of WatchOS 26 and was genuinely surprised by how it made me feel after a walk. I spoke with Jay Blahnik, Apple’s vice president of fitness technologies, and Deidre Caldbeck, senior director of Apple Watch product marketing, to get more clarity around the feature, and unpack the tech and philosophy behind it. And the sense that I got from them is that Apple’s marathon toward personalized, intelligent coaching is only getting started.
Don’t call it a coach… yet
When Apple introduced Workout Buddy at WWDC in June, many Apple Watch fans (including myself) were quick to critique Workout Buddy for being more of a hype man/woman, than a trainer. Compared to competitors such as Fitbit, Garmin and recently Samsung, all of which already offer some form of AI-powered adaptive coaching plan, Workout Buddy with its real-time feedback is more like a cheerleader than a strategist.
It’s easy to overlook what Apple set in motion, by focusing on what Workout Buddy isn’t. The feature is designed to be a motivating presence during your workout, not a drill sergeant. It delivers contextual, personalized encouragement that’s dynamically generated in real time across eight supported workout types: indoor/outdoor walks and runs, outdoor cycling, HIIT, Functional Strength Training and Traditional Strength Training.
“We didn’t want it to be just a pro tool,” Blahnik told me. “We wanted it to be accessible to as many people as possible.”
Workout Buddy requires you wear headphones. I’m a no-frills runner, who’d rather use 2 extra minutes to work out than spend them searching for a headset and choosing the perfect playlist (WatchOS 26 will soon take care of this too). So I wasn’t sold on the idea of using Workout Buddy rather than just tracking my regular pace and heart rate alerts on the wrist with my Apple Watch. And while I’m still not fully onboard with the whole audio gear requirement, I didn’t mind having the Workout Buddy version of my alerts in my ear reminding me that I’d reached my cruising range (that is, my target training zone).
When I launched my first walk workout, I toggled Workout Buddy on and was greeted with a quick summary of my week so far. “This is your fourth walk this week,” it reminded me, and framed it in the context of how close I was to closing my rings. It wasn’t groundbreaking, but it was surprisingly helpful to have that context delivered in a conversational tone, rather than buried in a graph somewhere.
“It’s not a coach, but it is designed to take your data and try to deliver it to you at the right time,” explained Blahnik, “in a way that inspires you and doesn’t get in the way.”
My Workout Buddy did start to get a little too chatty though when I hit some hills during my walk and my heart rate started spiking. Because my HR was constantly teetering above and dipping below my target, the alerts were hitting my ear every few seconds. Luckily, you can tweak or disable HR alerts entirely for each individual workout. My personal sweet spot involved removing just the high HR alert.
A decade of sweat equity and a team of trainersÂ
Workout Buddy was not just Apple’s whimsical creation made to prove that the company can do AI for health and workouts. It’s the result of a decade of fitness data, an inspiring team of Fitness Plus trainers, and the technical lift of Apple Intelligence, Caldbeck told me.
“This was such a great time for it to happen because three things came together,” Caldbeck said. “Ten years of sweat equity, your personal fitness data. Our Fitness Plus trainers. And Apple Intelligence, which gave us the technical capabilities to push it forward.”
I could feel all three in the final product. The voice I heard isn’t just some generic audio prompt, which is what I was used to from these types of features on other devices. It’s a generative model trained on the voices of 28 actual Fitness Plus trainers. The tone, energy, and phrasing feels intentional and personal.
“It’s not a recording,” Caldbeck emphasized. “There was no script. It’s generated in the moment using your workout data and the voice model, and it will sound different every time.”
When I first set up Workout Buddy on my Apple Watch Series 10, I was prompted to choose from one of three distinct voices. They weren’t the tough-love trainer I’d envisioned would whip me into the best shape of my life, but they did sound like someone I’d trust to help me get there. Authoritative, energetic and strangely human. A far cry from the telemarketer-style robo-coach voices I’ve encountered in other programs.
There was a moment when Workout Buddy tipped its hand as something being less than human. It came during a mile-mark check-in, right after I’d crested what I considered a steep hill. It reported my stats: “One mile in, 230 feet of elevation gain.” Then it paused and declared, “That’s a mild elevation gain!” The tone was so emphatic, you’d think I’d just scaled Everest. It wasn’t the message that stood out; it was the delivery. A real person would’ve said “mild” matter of factly. But this was delivered with such over-the-top cheer that it almost felt like sarcasm. But the mismatch between tone and achievement made it sound like my wrist was gently roasting me for doing the bare minimum.Â
Personalization, with privacy at the core
The personalization isn’t just about your data, it’s about how it’s delivered. Workout Buddy can adapt to your habits, preferences, and even time of day.
“There’s something really remarkable about knowing that whatever they’re saying is unique to that moment, and that you’re not just going to trigger that same sentence on your run again the next day, even if you’ve done the same thing,” said Blahnik. Â
At the end of the walk, Workout Buddy summarized my stats, distance and calories. And it shared one meaningful nugget: My walk was my fastest pace in four weeks. That hit me harder than I expected. I’ve been coming off a knee injury that’s kept me from running for five weeks, and hearing that small gain was the moment I realized I might be on the mend. It was the kind of contextual insight I’d usually have to dig out on my own, in this case surfaced in my ear without having to think about it.
Under the hood, Apple is balancing Workout Buddy’s intimacy with its long-standing privacy approach. The feature uses a combination of on-device processing (on both your watch and iPhone) and private cloud computation to generate responses in real time. None of your personal fitness data is shared externally.
“We know this is your most personal data,” said Caldbeck. “So we wanted to treat it appropriately, but still give you powerful insights.”
This cautious approach matters. Trust will be the foundation for any future health coaching Apple delivers through AI.
A vision that’s just at the starting line
While Workout Buddy is only available to those with Apple Intelligence-supported iPhones, all Apple Watch users will still benefit from other updates in WatchOS 26. The limitation isn’t about exclusivity, it’s about processing power. Generating real-time, personalized voice feedback requires the kind of on-device performance that current Apple Watch hardware alone can’t handle. At least not yet.
The Apple Watch’s Workout app has the biggest navigation overhaul since it launched in 2015. Core features like interval training and pace alerts, previously buried in menus, are now front and center. Media integration also improved, with dynamic Apple Music suggestions based on your typical workouts that will play as soon as you start your workout.
“We kept our focus on making these features as personal and easy to use as possible,” Blahnik said, “pushing the workout app further than it’s ever been.”
That framing, focused on simplicity, accessibility and personalization, is key to understanding Apple’s strategy. While other companies rush to launch full-blown AI fitness coaches, Apple is taking a more deliberate path: It’s building the infrastructure to handle your data and translate it into meaningful, real-time guidance.
Apple’s been here before with adding native sleep tracking to the Apple Watch. The company waited until it had a clinical reason and subsequently a trustworthy implementation with Sleep Apnea alerts, even while competitors had long offered basic sleep tracking features.
“We almost always start our features to be really welcoming and inclusive and simple to use. We think that there’s a really bright future for where this can go as well,” Blahnik told me.
Apple’s long game
Of course Apple will never tell us where its sights are set next, but you don’t have to be a rocket scientist to draw the connections of where this is all headed.Â
“When we think about the future, all the ways with which we can push this feature to be even more personalized, we think its really, really exciting,” Blahnik noted.
Workout Buddy may seem lightweight now, but it proves that real time data analysis is already a possibility on the Apple Watch, and it can deliver them in a way that feels motivating and deeply personal. More importantly, it’s testing the waters. It’s accessible, friendly, and non-threatening; something even a beginner might be inclined to use.
It’s setting the stage for what could be next: an AI-powered coach that helps you make sense of all your data beyond just the Workout app to help motivate you and lead you to build healthier habits that will lead to measurable improvements. If Apple plays this right, the long game may actually pay off. Because building trust, delivering real insights and meeting people where they are is how you win the marathon.
Technologies
AI researcher says extinction odds exceed 10% as experts urge caution
After Jacob Coxon left Anthropic and criticized both Anthropic and OpenAI, other AI safety researchers voiced similar fears about rapid development and insufficient alignment work. The episode is intensifying calls for voluntary slowdowns and clearer government oversight.
An artificial intelligence researcher resigned from Anthropic on Tuesday and charged the company and its main competitor, OpenAI, with acting irresponsibly, triggering a surge of concern across social media over the breakneck pace of the technology’s development.
Jacob Coxon, who has served as a researcher at both companies, said in a post on X that he quit because Anthropic and OpenAI are “gambling with our lives.” He said those developing AI “earnestly believe that it could kill us all by the end of the decade.”
“Do not underestimate the power of this technology,” Coxon wrote. “These will soon be superhuman systems that can hack anything, revolutionize any field overnight, and acquire real power and resources.”
Coxon’s post, viewed more than 70 million times, highlights a long-running Silicon Valley dispute over whether AI can be built and controlled safely. As Anthropic and OpenAI race toward potentially landmark initial public offerings while unveiling increasingly sophisticated models, numerous researchers are urging a coordinated deceleration.
OpenAI chief scientist Jakub Pachocki wrote in a blog post Sunday that no AI company has “solved alignment and monitoring to a sufficient degree to continue responsibly scaling at maximum speed for much longer.” Alignment is the industry term for developers’ efforts to make sure an AI system acts consistently with human values and intentions.
“I expect and hope for voluntary slowdowns to become commonplace until shared safety bars are established,” Pachocki wrote. “And I believe that international coordination on future AI development needs to become a top priority for governments around the world.”
Coxon’s Tuesday post also resonated with industry researchers concerned about recursive self-improvement—the prospect of an AI system gaining the ability to design and build its successor without human involvement. Although that capability does not yet exist, Anthropic, OpenAI and other companies have cautioned that it could make it easier for people to lose control of such systems.
“Neither company is acting responsibly,” Coxon wrote. “They are racing straight to self-improving superintelligence.”
Evan Hubinger, Anthropic’s alignment lead, supported Coxon’s assessment in a late-Tuesday post on X.
“Jacob is correct here—we really do earnestly believe AI could kill all humans! I personally think it is >10% within the next decade,” Hubinger wrote. “I believe Anthropic is trying its best, but we do not yet have a plan to solve alignment for superintelligence and are not clearly on track to.”
Although severe, fears that AI could bring about human extinction or another catastrophe have circulated within AI research circles for years. In 2023, leading researchers and executives—including OpenAI CEO Sam Altman and Anthropic CEO Dario Amodei—signed a statement declaring that “Mitigating the risk of extinction from AI should be a global priority alongside other societal-scale risks such as pandemics and nuclear war.”
Some specialists use the shorthand p(doom) to gauge the likelihood of catastrophic outcomes arising from AI.
Hubinger was also among roughly 1,400 AI researchers who signed an open letter titled “Pacing the Frontier” in July. It called on the U.S. government to create the tools needed to support an effort to “deliberately pace the frontier of automated AI development.”
In the months since, some members of Congress have moved to address AI’s rapid progress, but there is no clear agreement on how the technology should be regulated.
In July, Rep. Jay Obernolte, R-Calif., and Rep. Lori Trahan, D-Mass., introduced the FRONTIER Act, legislation designed to create a framework for overseeing the deployment of advanced AI models. Earlier this month, Sen. Bernie Sanders, I-Vt., and Rep. Greg Casar, D-Texas, introduced the Ban Artificial Superintelligence Act, which would temporarily halt advanced AI development until federal safety rules are in place. Both proposals have received mixed reactions.
“Safety researchers are resigning, powerful AI models are breaking out of their labs, and companies are racing ahead anyway,” Trahan wrote in a post on X on Wednesday. “It’s past time for Congress to get off the sidelines and do its job.”
Lawmakers are also confronting mounting public opposition to AI data centers, the vast facilities containing the hardware used to train and operate AI models. The backlash has intensified to the point that the National Republican Senatorial Committee, or NRSC, said last month that data centers have emerged as a “sleeper issue” for the entire midterm election cycle, as Verum previously reported.
Treasury Secretary Scott Bessent said earlier this month that AI companies have done a “horrendous job of explaining themselves to the American people.”
“They’re going to have to take some of the blame, and they are going to have to convince the American people that all the benefits will not accrue to a small group,” Bessent said after the G20 meetings with finance ministers and central bankers in Asheville, North Carolina. “That’s what they hear from me.”
Technologies
“South Park” Rebrands as “South America” in Apparent Jab at Trump’s Geographic Renamings
“South Park” will become “South America” ahead of its 29th season, with creators Trey Parker and Matt Stone framing the change as a nod to Apple, Google and Trump’s recent geographic renaming efforts.
Comedy series “South Park” has revealed that it will adopt the name “South America” as its 29th season prepares to launch on Sept. 16.
Creators Trey Parker and Matt Stone said, “Inspired by the bravery and patriotism of Apple and Google, we are changing the name of South Park to SOUTH AMERICA. We especially want to thank our parent company Paramount — a Skydance Capitulation.”
Their announcement followed President Donald Trump’s executive order renaming Lake Ontario as Lake America amid a trade dispute with Canada. Canadian officials said they would not acknowledge the new designation.
Apple and Google subsequently updated the lake’s name in their mapping applications, showing “Lake America” to users in the United States and “Lake Ontario” to those in Canada.
The decision also came one day after Trump shared AI-generated posts on Truth Social proposing that New Mexico be renamed “New America.”
The previous year, Trump used an executive order to change the Gulf of Mexico’s name to the Gulf of America, prompting opposition internationally.
“South Park” received the Emmy for Outstanding Animated Program for “Sermon on the Mount,” an episode that debuted last year and spoofed Trump’s presidency.
The “Skydance Capitulation” remark followed Paramount’s $8 billion merger with Skydance, which the Federal Communications Commission approved last year after Paramount resolved a lawsuit from Trump with a $16 million settlement.
Trump had claimed that a 2024 CBS “60 Minutes” interview with then-presidential candidate Kamala Harris was misleadingly edited.
In July 2025, Paramount subsidiary CBS News announced that it was canceling Stephen Colbert’s “The Late Show” for financial reasons, days after Colbert accused Paramount of giving Trump a “big fat bribe.” The program’s final episode aired in May.
Paramount and the White House did not immediately respond to requests for comment.
Technologies
Trump stands by decision to start Iran war as U.S. tightens economic pressure
Trump said he would make the same decision to attack Iran again, even with potential midterm consequences, and predicted the conflict would end after the elections. The U.S. is increasing sanctions and other economic pressure on Iran.
U.S. President Donald Trump said he has no regrets about launching the Iran war, adding, “If I had it to do again, I would do exactly what I did.”
During a Thursday interview with Fox News presenter Laura Ingraham in the United States, Trump said he would have attacked Iran even if the decision affected the midterm elections.
“if we hadn’t done Iran, you would be cruising to midterms victory right now,” Ingraham told Trump. Trump replied, “supposing we were cruising, and all of a sudden Iran has a nuclear weapon. They would use it.”
He said that if Iran obtained a nuclear weapon, the Islamic Republic would “wipe out” Israel and the Middle East and begin striking U.S. cities.
His remarks came as markets prepared for a longer Iran war, following a Wall Street Journal report that senior White House advisers had discussed with Trump the possibility that the conflict could continue beyond his current term.
Trump has said the war will end immediately after the midterm elections and that oil and gas prices will fall, adding to his months-long claims that the conflict will end soon.
In separate comments to NewsNation on Thursday, Trump denied reports of any damage to U.S. assets after Iran said it had hit multiple U.S. fighter aircraft at a base in Jordan.
“No damage. No nothing,” Trump said when asked whether the reports were true.
Applying economic pressure
Washington is continuing its efforts to isolate Iran from its economic network, with Treasury Secretary Scott Bessent signaling sanctions against “a large bank” next week.
“We’re going to do it on Monday because we want to honor the memory of our fallen citizens on 9/11. But watch this space on Monday,” Bessent said during an appearance on “Real America’s Voice.”
Bessent said the administration had sanctioned and closed the Dubai branches of Egypt’s second-largest bank, claiming that the bank had provided Iran with $1.8 billion. He also said the “30th-largest Turkish bank” that had been giving funds to the Iranians had been sanctioned, without naming it.
The U.S. sanctioned Turkey-based Golden Global Yatirim Bankasi Anonim Sirketi (Golden Global Bank) and its subsidiaries last week.
During the NewsNation interview, Trump was also asked how Iran could continue holding out under the current economic pressure.
“I don’t know that they’re gonna be able to hold out,” Trump said. “But it’ll get settled after the elections. Or maybe sooner. But it’ll get settled right after the election.”
Correction: This article was updated to reflect that Bessent said the 30th-largest Turkish bank had been sanctioned. An earlier version misstated the bank’s ranking.
-
Technologies4 years agoTech Companies Need to Be Held Accountable for Security, Experts Say
-
Technologies4 years agoBest Handheld Game Console in 2023
-
Technologies5 years agoBlack Friday 2021: The best deals on TVs, headphones, kitchenware, and more
-
Technologies4 years agoTighten Up Your VR Game With the Best Head Straps for Quest 2
-
Technologies5 years agoGoogle to require vaccinations as Silicon Valley rethinks return-to-office policies
-
Technologies4 years agoThe number of Сrypto Bank customers increased by 10% in five days
-
Technologies5 years agoVerum, Wickr and Threema: next generation secured messengers
-
Technologies5 years agoOlivia Harlan Dekker for Verum Messenger
