James Comey is not exactly someone we’ve ever been a fan of on Techdirt. He was a terrible FBI director in so many ways. We’ve spent years criticizing the man — for his crusade against encryption, his supporting the FBI’s ridiculously aggressive impersonation of reporters, his embrace of the FBI’s program to coerce and entrap people down on their luck into fake terrorist plots, and much more. And, while the impact has been exaggerated, it is true that he took multiple actions violating DOJ procedures that likely helped get Donald Trump elected in 2016. So it’s not like I’m rushing to support the guy. He’s a bad cop and has been for some time.
But the indictment the Department of Justice handed down against James Comey on Tuesday is a truly embarrassing legal document, and everyone involved in producing it should be professionally radioactive for the rest of their careers. I would have said it’s one of the most embarrassing legal documents that this DOJ has produced, but remember, just a day earlier they filed a legal brief that was indistinguishable from a Truth Social post.
The charge, in its entirety, concerns this Instagram post from May 2025:
If you can’t see that, it’s an Instagram post from Comey showing some shells on some sand with the shells spelling out 8647 and the caption on the post saying:
Cool shell formation on my beach walk
For this — for posting a photo of arranged seashells in a slightly sassy pattern and posting it to Instagram — Comey has been charged with two federal felonies: threatening the President under 18 U.S.C. § 871, and transmitting a threat in interstate commerce under 18 U.S.C. § 875(c). (For what it’s worth Comey has claimed he didn’t arrange the sea shells, but just found them. It’s unclear if that makes much of a difference, it’s protected speech either way).
Ken “Popehat” White, who has perhaps done more than any other lawyer in America to explain First Amendment doctrine to laypeople, didn’t mince words about what this is:
The charge is preposterous and no competent or honest prosecutor would bring it. It represents a betrayal of the professional and ethical obligations of every U.S. Department of Justice attorney involved, and reflects the complete collapse of the Department’s credibility and independence in favor of a cultish and cretinous devotion to Donald Trump.
He’s right, and the way to understand just how right he is requires understanding the path that brought us here.
Because this is the second time the Trump DOJ has tried to indict Comey. The first attempt collapsed in spectacular fashion last year, after Trump — in what was apparently supposed to be a private direct message but accidentally went out as a public Truth Social post — demanded that Pam Bondi install Lindsey Halligan, a former insurance lawyer with no relevant experience, as a U.S. Attorney specifically because she had promised to indict Comey. The problem: Halligan wasn’t legally appointed. The entire indictment got tossed before the court could dismiss it for being ridiculous (which would have happened) because the person who filed it wasn’t allowed to file it.
As we noted at the time, this pattern of procedural self-sabotage is a recurring feature of an administration that treats legal procedure as an inconvenience rather than the actual point of having a justice system.
So how did the DOJ respond to that humiliation? By coming back with something substantively even worse. In theory, they tried fixing the “wrong person filed it” problem by having an actually legally appointed person file something… even if that something has no legal basis whatsoever. Progress! Sort of?
The seashell indictment was filed by W. Ellis Boyle, the U.S. Attorney for the Eastern District of North Carolina, with Assistant U.S. Attorney Matthew R. Petracca listed as the prosecuting attorney. Remember those names. They put their signatures on this. Boyle is listed as the U.S. Attorney for the Eastern District of North Carolina, but he’s serving in an acting capacity — Trump has nominated him multiple times, yet the Senate has still refused to confirm him.
The legal problem with the indictment is pretty easy to spot: to convict someone under either of the threat statutes the DOJ is invoking, the government has to prove the communication constituted a “true threat.” Under controlling Fourth Circuit precedent (this case is in North Carolina), a true threat is something “an ordinary, reasonable recipient who is familiar with the context in which the statement is made would interpret as a serious expression of an intent to do harm.”
As Ken White noted, the Supreme Court established this framework in Watts v. United States, a 1969 case involving an 18-year-old draft protester who said:
They always holler at us to get an education. And now I have already received my draft classification as 1-A and I have got to report for my physical this Monday coming. I am not going. If they ever make me carry a rifle the first man I want to get in my sights is L. B. J.
The Court found this was protected political hyperbole, not a true threat. An explicit statement about wanting a President in your rifle sights — protected.
If Watts isn’t damning enough, there’s United States v. Bagdasarian, a much more recent Ninth Circuit case where a man posted online statements about wanting to shoot then-candidate Barack Obama, including some genuinely vile racially explicit language about hoping Obama would be killed. The court held that even that did not constitute a true threat under the relevant statutes.
I’d be curious to hear from anyone defending this indictment whether they think Bagdasarian was wrongly decided. Or do we change the “true threat” standard when the target is Trump?
So the descending ladder of seriousness looks like this:
Any first-year law student who’s taken a basic First Amendment course could tell you the seashell post is constitutionally protected. Any prosecutor with five minutes of research time would know that Bagdasarian and Watts exist. But, of course, as we’ve seen over and over and over again in the Trump era, the point is not to bring a good case or a winnable case. The point is just to punish Trump’s enemies with vexatious, vindictive prosecutions in hopes of creating a chilling effect among the populace and stopping them from criticizing the President with the thinnest skin possible.
Now, “86” has had various meanings over the years — to “86” something in restaurant slang means to remove it from the menu or get rid of it. The DOJ’s theory is apparently that when used about a person, it means to kill them. No one else believes that. This is the kind of motivated reading that requires ignoring both the dictionary and how actual humans use language.
But fine, let’s grant the absolute most uncharitable reading and say “86 47” means “get rid of the 47th President through killing.” Even granting that — even doing all the work for the prosecution — it’s still obviously protected political expression, and still obviously not a true threat under the controlling case law.
Which brings us to the part that genuinely cannot be explained by anything other than pure vindictiveness. Here is a tweet from Jack Posobiec, a prominent Trump loyalist/conspiracy theorist, posted in January 2022:
That tweet is still up. I just made that screenshot minutes ago. As of this writing, it has been online for nearly four years. No FBI investigation. No federal indictment. No felony counts. Literally no one thought that was an actual threat. Because it’s not. Apparently the DOJ’s theory of criminal threats has a loyalty-based expiration date — the same numerical expression is a felony when arranged in shells by a Trump critic and a perfectly fine tweet when posted by a Trump supporter about a different President.
Indeed, the fact that Posobiec seems to have no issue keeping this tweet up is itself a sign that the MAGA world knows it’s engaged in purely theatrical vindictive prosecution — and wants you to know they know. To them, once again, nothing here is about justice or the rule of law. It’s just “will this make the people I dislike upset.” That is their only motivating factor.
The DOJ has baked the selective prosecution argument directly into its own theory of the case. Comey’s lawyers will surely refresh the selective prosecution motion they filed in the first, dismissed indictment, and the facial absurdity of this one — combined with the existence of identical, ignored expression by Trump allies — makes that motion approximately as easy to support as such motions ever get.
There’s a specific kind of institutional rot in play here, driven entirely by Donald Trump and his minions. Competent authoritarianism is dangerous in obvious ways. Incompetent authoritarianism that keeps trying anyway is dangerous in different ways: it normalizes the use of state power for personal vengeance while demonstrating that the people wielding it will stop at nothing — even on the most facially ridiculous grounds. That’s a chilling effect doubled: a politicized DOJ, staffed by people who can’t pass a First Amendment quiz.
White is right that the indictment is unlikely to survive. Comey’s attorneys can challenge it on its face, arguing that even taking every allegation as true, seashells spelling “86 47” are protected by the First Amendment as a matter of law. The assigned judge was appointed by a Republican but is reportedly not a partisan hack, and the case law here is so clear that it would take extreme judicial bad faith to let this proceed. The selective prosecution motion is also stronger now than it was the first time, with Posobiec’s untouched tweet sitting there as Exhibit A.
But as White notes, surviving the motion to dismiss isn’t actually the point:
The point of the indictment is to demonstrate that the United States Department of Justice is wholly an instrument of Donald Trump’s senescent pique, no more independent of him than a boil on his ass. The point is to show that the administration can, and will, use the Department’s mechanisms to punish enemies. The point is to show that the Department can, and will, punish protected speech. The point is to show that the Department is staffed by committed fanatics willing to do anything, however unethical and unconstitutional, to promote Trump.
The point is to show that in the war between Donald Trump and the U.S. Department of Justice, Trump has won. Now they’re on the field slitting the throats of the wounded and looting bodies.
W. Ellis Boyle and Matthew R. Petracca put their names on this indictment. They will, presumably, lose this case the way the previous Comey case was lost — embarrassingly, on grounds that any competent attorney not engaged in cult-like performative fealty to a wannabe authoritarian could have anticipated. And when this is all over, when there is some accounting for what was done to the Department of Justice in these years, the people who signed the seashell indictment should never be trusted with prosecutorial power, a bar membership, or any position requiring professional judgment ever again.
The shells, for what it’s worth, were on a beach. The tide has presumably long since rearranged them. The Instagram post was taken down fairly quickly when the MAGA world lost their minds over it. The federal felony charges, somehow, remain.
Filed Under: 1st amendment, donald trump, ellis boyle, free speech, jack posobiec, james comey, matthew petracca, true threats, vindictive prosecution
An exercise bike is one of the most effective tools for building endurance and burning calories, but it’s also one of the easiest to use incorrectly. Just because the motion feels repetitive doesn’t mean it’s foolproof; whether you’re a first-timer or a seasoned cyclist, small errors in setup or form can stall your progress. Instead of maximizing your cardio, poor habits can lead to unnecessary strain and keep you from seeing the results you’re working for.
To ensure you’re using your exercise bike with the proper form, adjustments and technique, so you can get the most out of your workouts, I spoke with several cycling experts to learn more about mistakes you’re likely making when on an exercise bike.
One of the first lessons I learned when I started taking group cycling classes was that you need to properly adjust the seat height. Aviron coach and certified indoor cycling instructor John Steventon says the right seat height is key.
“[If the seat is] too low, they won’t be able to get full extension of the legs, often needing to cycle with their knees splayed outward to make room for them,” Steventon explains. This leads to a loss of potential power and a compressed and uncomfortable pedal stroke. “If the seat is too high, the rider will risk bouncing from side to side as they lean off the side of the saddle, trying to get the right extension of the legs.”
According to Matt Wilpers, the saddle should be positioned high enough that your knee is slightly bent and not locked.
Peloton instructor Matt Wilpers offers a couple of tips to keep in mind when setting up your seat. “When riding, bring your one foot to the bottom of the pedal stroke (6 o’clock),” he says. “The saddle should be positioned high enough that there is a slight bend in the knee so that the knee is soft and not locked.” This will prevent you from feeling cramped or having to reach for the bottom of the pedal stroke.
Steventon points out that how the foot sits on the pedal is important, too. “The center spindle that the pedal rotates around should be underneath the ball of the foot because I’ve seen people who ride on their toes, and people who ride on their heels — both reduce power and efficiency of the stroke.”
Wilpers advises trying the “heel test,” which is when you unclip the foot at 6 o’clock to see if your heel can touch the pedal when your leg is straight.
You don’t want the seat too far or too close to the handlebars.
Another factor to consider when setting up your exercise bike’s saddle is its distance from the handlebars. Wilpers says the goal here is to get the knee stacked over the pedal, especially when the pedal is at 3 o’clock. “I usually start with the saddle pushed toward the handlebars and then adjust accordingly when I ride,” he recommends. The problem is that when your seat is too far forward, your knees take on the pressure, and if it’s too far back, the stress lands on the heels.
Besides the seat, make sure the handlebars are at the right height. “In general, it’s recommended to keep the handlebars at the level of your saddle or just above,” says Wilpers. “Many competitive cyclists like their handlebars at or slightly below saddle height because it better optimizes both performance and aerodynamics when cycling outdoors.”
Make sure the handlebars are aligned with the saddle on your bike.
However, Wilpers says that in some cases, such as if you have back problems, raising your handlebars even further is advised.
If you’re eager to jump on your bike and get a quick workout in, chances are you’re skipping a warm-up or cool-down. Steventon says that if you want to get your body ready for your workout, spending 10 minutes increasing resistance and cadence before a longer or more intense ride will prime the muscles.
“Rather than slowly working up from 60 revolutions per minute to 110 rpm during the warmup, 30-second bursts at gradually higher reps in between 30 seconds of easy pedalling will allow the warmup to be effective without draining energy before the main workout,” he explains.
Some warm-up exercises Wilpers recommends include:
If you skip a warm-up or cool-down, you’re missing out on some important steps.
The cool-down shouldn’t be neglected either. “In some cases, after a particularly hard workout, stopping suddenly can cause the blood to pool in the legs, and the cyclist may get dizzy,” warns Steventon. He recommends spending 5 minutes slowing your cadence to help the muscles keep pumping blood and oxygen as you ease out of the strain of the main workout.
You should also stretch the muscles used during a ride, such as your hamstrings, quads, hip flexors and calves. “Triceps, shoulders and wrists shouldn’t be ignored either,” Steventon says. “These muscles are soaking up the mass of the upper body, bouncing up and down, left and right on the handlebars.”
If you’re riding a bike with a screen, the screen’s position can sometimes lead to neck pain. In those circumstances, Steventon recommends gentle neck stretches to ease off the tension after a workout.
If you don’t have a general workout program or structure, you may be using your bike inefficiently. It’s important to make sure that you’re riding with intention if you want to get the most out of it. “Everyone has the same three training variables to manipulate in order to get what they want out of their training: frequency, duration and intensity,” Wilpers says.
Frequency relates to how often you work out, since it’s important to have a consistent riding schedule. “Cycling is great because it’s low impact, meaning it’s easier to recover from it and therefore you can ride more often without much fear of injury,” Wilpers says.
If you’re a newbie, it’s best to focus on easy rides. Steventon recommends new riders aim for two to three rides per week and take a rest day between each cycling workout to let their bodies recover from the workout the day before. “Not only for the muscles, but depending on the length of the ride, there’s an element of saddle discomfort that rest will help,” Steventon explains.
Once you have a riding schedule established, you can focus on duration. This is when you start increasing the length of your sessions.
Finally, once you’ve become comfortable with your bike, you can increase the intensity to improve your fitness. Steventon says more experienced riders are capable of putting in four to five workouts a week.
“The importance of foundation building zone 2 heart rate [exercise performed at 60% to 70% of your maximum heart rate] during longer rides is the same for all riders, but with three more workouts to play with,” Steventon explains. “They can add in some tempo and sprint interval workouts too, pushing the cardio harder, and working at a mix of cadence and resistance through these workouts.”
Below are different ways you can structure your workout depending on your goals:
You should have a plan, whether cycling is your main form of cardio or your choice of cross-training.
If cycling is your main focus, Wilpers recommends riding three to five days per week, with one to two intense days, one long day and the remainder as easier rides. Aim to ride for 30 minutes to 2 hours or longer.
If you’re prioritizing strength training but want to include cycling as your preferred form of cardio, Steventon and Wilpers advise aiming for two to three sessions at a zone 2 heart rate for 20 to 45 minutes.
You can also use an exercise bike for cross-training. This could be ideal for runners or other athletes who want a hybrid workout schedule during the week. Steventon says, “Cross-training utilization of cycling can be a very effective way to keep cardio topped up without the impact problems of constantly running.” Wilpers recommends cross-training once or twice per week for 20 to 45 minutes, as long as you’re healthy.
If you have injuries or certain aches from running, you can shift your training to cycling two to four times per week for 20- to 45-minute sessions. Steventon recommends that runners cycle whenever their bodies need to recover from running. “Long, slow rides will keep your fitness foundation strong, with harder sprint intervals tapping into VO2 max improvements,” Steventon says. VO2 max is the maximum rate at which your body consumes oxygen during exercise.
Riding distracted is just as bad as doing too much during a ride.
One of the biggest mistakes people make when riding an exercise bike is either doing too much too soon or coasting while distracted. “Many people seem to think that unless they wake up sore, nothing was accomplished, and on the other end of the spectrum, I see people just pedaling and texting,” says Wilpers.
When riding, remember to focus on the workout at hand. Distractions like texting or answering emails can get in the way of making progress. “Have a purpose and plan before getting on the bike so that your time is structured,” Wilpers suggests. “This will help you stay focused and get the most out of your time.”
Steventon notes that the best way to use a bike or any fitness machine is to mix up the intensities to avoid hitting a plateau. “Long, slow rides, short sprint intervals and tempo rides (moderately hard ride) including hard, long intervals will improve the mitochondria-building, zone 2, foundation end of fitness, while the shorter sprints will help improve the VO2 max end, and the tempo rides are where mental resilience is built,” he says.
Wearing the right cycling shoes can improve your riding experience. Depending on the bike, you may need specific cycling shoes, like carbon-fiber cleats, road bike shoes, clip-in shoes, mountain bike shoes or even everyday shoes.
When choosing a cycling shoe, you want to make sure it’s comfortable because they don’t come cheap. Steventon recommends mountain bike shoes if you aren’t sure what to pick. “These are a little bit more flexible and have recessed cleats, making it a lot easier to move around, providing stability in situations where ‘off bike’ moves like squats or weights are included in a cycling class.”
The right shoes can improve your cycling experience.
Wilpers favors clip-in cycling shoes because they’re stiff and lead to better energy transfer from the body to the bike. However, the shoe you pick will depend on the type of bike you’re using, and most importantly, you want the shoe to fit well.
Steventon says everyday gym sneakers are acceptable, but that they may not be the best option. “The pedal efficiency may be compromised even with toe clips that hold these shoes in place on the pedal,” he explains, pointing out that because these shoes tend to be soft-soled, it can be uncomfortable to use for long periods of time.
Wilpers notes that some cycling shoe brands are much narrower than others, so it depends on your needs and comfort. “Lastly, I think it’s worth noting that a good set of road cycling shoes can cost hundreds of dollars, but these shoes typically last up to five years,” he says.
Usually, when you stop an exercise bike, you’re either pressing down the emergency brake or using the resistance knob to slow it down. “Trying to suddenly stop the flywheel or unclip feet while the heavy flywheel still wants to turn has great potential for injury,” warns Steventon. “Always use the resistance knob or emergency brake to slow things down properly, and wait until the flywheel has stopped before unclipping.”
If you aren’t properly braking, you risk injuring yourself.
It’s easy to forget that you need to do maintenance and clean your bike every so often. Wilpers reminds us that all bikes need to be cleaned and maintained. “Cyclists are known for constantly cleaning and occasionally replacing parts on their bikes because this is what it takes to keep your bike working great for many years,” he explains.
Keeping your bike in good condition is important if you want it to last a long time.
One thing people often overlook is adjusting their bike settings annually. “Think about how much your body can change in a year,” explains Wilpers. “You may get stronger, weaker, heavier, lighter, tighter, more flexible and so on.” By adhering to bike settings, you’ll not only have a more comfortable ride but also one that’s customized to your changing needs.
The venture world is still digesting the eye-popping debut of London-based Ineffable Intelligence, the new startup from DeepMind legend David Silver, which announced $1.1 billion in funding at a $5.1 billion valuation this week.
The deal — which CNBC called the largest-ever seed round for a European startup — drew heavyweight backers including Sequoia Capital, Lightspeed Venture Partners, Google and Nvidia.
But tucked inside the cap table is a less obvious name: Flying Fish Partners, a Seattle-based firm with less than $250 million under management that wrote the first check for the nascent company just four months ago.
So how did a relatively small and scrappy Pacific Northwest venture capital firm break into one of the most competitive AI deals in years?
Turns out it was less about check size and more about old-school hustle, networking and smarts, including years of relationship-building, a well-timed early bet, and a lot of flights to London.
I caught up with Flying Fish Managing Director Frank Chang — who had just landed in London — to get the scoop on how his tiny Pacific Northwest firm managed to secure a front-row seat in one of the most ambitious moonshots in the race toward artificial general intelligence.
“We’ve been laying the foundations for an investment like this for years,” Chang said via email.
The story starts with Phaidra, a Seattle-based AI startup co-founded by a team that includes former DeepMind engineers Jim Gao and Vedavyas Panneershelvam. Flying Fish wrote one of the first checks for the company in 2020. The investment gave them more than a compelling portfolio company — it opened doors into the world’s most elite AI talent pool.
Chang and Flying Fish partners Heather Redman, Geoff Harris and others worked those connections, essentially making the firm a fixture in the London AI scene, where DeepMind started in 2010 before Google acquired it in 2014.
They hosted enough coffee meetings and dinners with AI researchers that Flying Fish apparently became a recurring character in the “Ex-DeepMind” WhatsApp group.
“There were so many gatherings over the years that I’ve been told founders would post on that group about which Flying Fish dinner they got invited to,” said Chang, who worked at Amazon and Microsoft before co-founding Flying Fish.
In fact, it was Heather Gorham, then a principal in the VC firm, who initially reached out to Silver after he published a paper titled “The Era of Experience,” which posited that a new approach was needed to AI since the “knowledge extracted from human data is rapidly approaching a limit.”
Chang and Gorham developed a strong relationship with Silver, discussing company goals, strategies, recruiting and a complementary philosophy on where AI is headed. The Phaidra investment also gave them credibility with Silver and his circle.
Gorham has since joined Ineffable’s founding team, according to her LinkedIn profile.
By the time Silver — the mind behind AlphaGo and a leading figure in reinforcement learning — was ready to build his new startup, Flying Fish wasn’t just a random VC from Seattle; they’d established themselves at the forefront of the field. And they were ready to write a check.
In the most recent round — which included Google, Nvidia and the U.K. Sovereign AI Fund — Flying Fish wrote the biggest check in the firm’s 10-year history. Chang declined to disclose the full size of Flying Fish’s investment in Ineffable. However, it was so large that they established a special purpose vehicle, a separate fund set up specifically for a single investment, to fund a large portion of it.
The big bet comes from a mutual understanding with Silver on how AI is changing, going beyond large language models (LLMs).
“Lots of investors are searching for the company that will deliver AGI or superintelligence, and there is a popular belief amongst many that LLMs can get there,” Chang said. “However, we hold the position that LLMs, powerful as they are, cannot. Many prominent AI researchers have the same view.”
The thesis behind Ineffable is that LLMs have a ceiling. To reach AGI, or true superintelligence, you need a “superlearner.” That means discovering knowledge from its own experience, rather than just vacuuming up the internet’s existing data.
The company’s approach, as Silver described in a Wired profile this week, involves placing AI agents inside simulations where they can learn from experience, achieve goals and collaborate with one another.
Chang acknowledges this is an outlier in the “frontier lab” arms race where Anthropic has reached a $1 trillion valuation on secondary markets. But he argues the capital is a necessity.
“If Ineffable achieves what it set out to, and far surpasses the capabilities of LLMs, the returns, even at a seed round value of over $5B will be well worth it,” he said.
Beyond the math and the GPUs, what most impressed Chang about the investment was the entrepreneur behind Ineffable.
“David is a genuinely good human being, down to earth, and likeable, but he is clearly driven by the mission, so much so that he has said he will give away what he makes from Ineffable equity to charity,” he said. “Combine all this with his pedigree, and he will be able to attract the very best talent to the company.”
In fact, in the Wired profile this week, senior writer Will Knight wrote, “Silver’s reputation as being both a top researcher and frankly, not an asshole, may work in his favor when it comes to recruiting talent.”
That also stands out to Chang at Flying Fish.
“Add it all up and you have one of the most respected and prominent AI researchers on the planet who is mission oriented, a good person, has built a rock star team, and is going after a huge swing that could change the world,” said Chang. “What’s not to like?”
Healthcare, as we’ve known it for decades, has largely been built around a simple premise: you seek help when something goes wrong. Symptoms appear, you visit a doctor, you run tests, and then you treat the issue. But that model is increasingly being challenged by a new category of companies – platforms that aim to predict, prevent, and optimize health before disease ever sets in.
At the center of this shift is a rapidly growing industry often referred to as preventive health, longevity tech, or health optimization platforms. And within this space, Superpower is positioning itself as one of the most comprehensive solutions available today.
This article breaks down what this industry is, why it matters, who it’s for, and how Superpower compares to its closest competitors.
Superpower operates in the preventive healthcare and longevity technology space – a sector that brings together advanced biomarker testing, AI-driven health insights, continuous monitoring, and personalized medical guidance into a unified system. Rather than treating these as separate services, this industry merges them into a cohesive experience designed to give users a deeper and more ongoing understanding of their health.
Unlike traditional healthcare systems, which are reactive, this industry focuses on early detection, long-term optimization, and data-driven decision-making. The underlying philosophy is straightforward but transformative: if you can measure your body deeply and consistently, you can identify risks years before they develop into full-blown diseases.
This space has grown rapidly in recent years as consumers have become more proactive about their well-being, wearable technology has entered the mainstream, and artificial intelligence has made it possible to interpret complex biological data in a way that was previously inaccessible.
Modern lifestyles have created a paradox. We have access to better medical technology than ever before, yet chronic conditions such as diabetes, cardiovascular disease, and hormonal imbalances continue to rise at an alarming rate.
The issue is not the availability of treatment but the timing of intervention. Most conditions are diagnosed only after symptoms become visible, at which point treatment becomes more complicated, more expensive, and often less effective.
Preventive platforms like Superpower are designed to address three interconnected challenges. The first is the lack of visibility into one’s own body, as most individuals rely on infrequent and limited testing that fails to capture early warning signs. The second is the fragmentation of health data, where medical records, fitness metrics, and lab results exist in isolation without a unified view. The third is the complexity of medical information, which makes it difficult for individuals to interpret results and take meaningful action without expert support.
By bringing these elements together, Superpower creates a system that continuously tracks, analyzes, and guides health decisions, effectively turning scattered data into a coherent and actionable narrative.
At its core, Superpower combines multiple layers into a comprehensive health ecosystem that goes well beyond traditional testing platforms.
Deep Diagnostics
Users undergo testing across more than a hundred biomarkers, spanning 13 distinct health categories that collectively map the body’s internal state. These include key systems such as metabolic health, hormonal balance, inflammation, nutrient status, and long-term disease risk. This level of depth allows Superpower to establish a meaningful baseline for every user, turning what would otherwise be a generic health check into a highly personalized starting point.
This baseline membership experience is critical because it defines the user’s entry into the platform. Rather than offering fragmented insights, Superpower builds a structured understanding of the body from day one, enabling early detection of imbalances long before they manifest as symptoms.
The Superpower Baseline Panel is a comprehensive entry-level diagnostic test designed to give users a detailed snapshot of their health through the analysis of over 100 biomarkers across key systems like metabolism, hormones, inflammation, and disease risk. It serves as the starting point for Superpower’s broader preventive health platform, establishing a personalized baseline that can be tracked over time. The blood draw can be completed in a lab visit or an optional at-home appointment, with results typically delivered within a week. These insights are then used to generate tailored health recommendations and ongoing guidance within the Superpower ecosystem.
But the Baseline Panel is called “baseline” for a reason. Superpower offers a suite of advanced add-on panels for users who want to go deeper into specific systems, including expanded hormone, advanced cardiovascular, heavy metals, micronutrient, and toxin panels. These optional tests layer onto the baseline to give users a more granular view of the areas most relevant to their goals, symptoms, or family history.
Unified Health Platform
Superpower integrates lab results, medical records, wearable data, and lifestyle inputs into a single, continuously evolving dashboard. Instead of forcing users to piece together their health story across multiple apps and reports, it presents a unified view that reflects how different systems in the body interact with one another.
This integration extends beyond diagnostics into actionable pathways. Based on biomarker insights, users can access a built-in marketplace that offers curated supplements, often recommended directly in response to their test results. This creates a seamless transition from insight to intervention, removing the guesswork that typically follows health testing. Another important point we would like to mention here is that the marketplace has member-exclusive discounts of up to 25 per cent.
AI + Human Expertise
Superpower’s approach is not limited to data interpretation. It combines artificial intelligence with real medical oversight to deliver guidance that is both scalable and clinically grounded. The platform goes a step further by integrating access to a pharmacy, allowing members to obtain prescription treatments such as GLP-1 therapies, NAD+ support, and testosterone optimization when medically appropriate.
This combination of diagnostics, recommendations, and access to treatment creates a closed-loop system where users are not just informed about their health but are also equipped to act on it immediately.
Superpower is not alone in this space. Several companies are attempting to redefine preventive healthcare, each approaching the problem from a different angle.
Function Health stands out as one of the closest competitors, offering extensive biomarker testing and detailed reports. While its strength lies in diagnostics, Superpower provides more tools for members to turn their results into action plans.
InsideTracker has built a strong reputation among athletes and fitness enthusiasts by focusing on performance optimization. Its insights are valuable, but they are often centered around fitness outcomes, while Superpower expands the conversation to include long-term health and disease prevention.
Mito Health offers biomarker testing alongside biological age tracking, giving users a sense of how their body is aging. Mito Health offers a more premium experience (such as a 1:1 consultation about results), but at a more expensive price point ($349 vs $199), and does not have an at-home blood draw option as Superpower does.
Empirical Health takes a specialized approach by focusing on cardiovascular biomarkers. This depth is useful within its niche, but it lacks the broader, multi-system perspective that Superpower provides.
Vitals Vault emphasizes the scale of data, offering a larger number of biomarkers. While this appeals to users seeking detailed insights, the platform leans heavily on data collection, whereas Superpower focuses on translating that data into clear and actionable recommendations.
SelfDecode approaches preventive health through DNA analysis, offering insights based on genetic predisposition. While valuable, genetic data is inherently static, and Superpower complements this by focusing on dynamic, real-time biological signals.
SiPhox Health prioritizes convenience through at-home testing kits, making health tracking more accessible. However, it does not offer the same level of integrated care and continuous guidance that defines the Superpower experience, which includes a team of trained medical professionals for blood draws instead of home test kits, so that the results are as reliable as possible. The reason being self-administered at-home test kits are at a higher risk of being contaminated or done incorrectly.
WHOOP brings wearable-based insights into recovery, sleep, and strain. While useful for daily monitoring, it lacks the diagnostic depth and medical-grade analysis that Superpower integrates into its platform. Also, members can also integrate data from Whoop wearables into the Superpower platform.
Across this competitive landscape, most platforms excel in specific areas, whether it is deep diagnostic testing, fitness tracking, genetic insights, or accessibility through at-home solutions. Superpower distinguishes itself by combining these capabilities into a single, unified system that prioritizes both depth and usability.
Its differentiation lies in three core strengths. The first is its ability to pair extensive data collection with meaningful context, ensuring that users are not overwhelmed by numbers but guided by clear explanations of what those numbers mean. The second is its focus on continuous health management, transforming what is typically a one-time testing experience into an ongoing journey that evolves with the user. The third is its integration of artificial intelligence with human medical expertise, creating a system that is both scalable and deeply personalized.
Preventive health is no longer confined to niche audiences such as biohackers or elite athletes. It is becoming increasingly relevant for a wide range of individuals who want greater control over their well-being.
High-performance professionals, including entrepreneurs and executives, can use these platforms to optimize energy levels, focus, and productivity. Individuals with a family history of chronic conditions such as diabetes or heart disease can benefit from early detection and proactive risk management. Fitness enthusiasts who already track their activity can gain deeper insights into underlying biological processes, moving beyond surface-level metrics. At the same time, a growing number of people who simply want to understand their bodies better are turning to these platforms as a more informed alternative to reactive healthcare.
Healthcare is undergoing a fundamental transformation that is reshaping how individuals interact with their own bodies. The shift from episodic care to continuous monitoring is changing expectations around how often and how deeply people engage with their health. Similarly, the move from generalized advice to personalized insights reflects a broader demand for precision and relevance in medical guidance.
This evolution is being driven by advances in diagnostics, the growing capabilities of artificial intelligence, and a cultural shift toward proactive health management. As these forces converge, platforms like Superpower are emerging as central tools in this new ecosystem, acting as the interface through which individuals understand and manage their health.
The preventive health industry is no longer a niche category but an essential part of the future of medicine. While competitors such as Function Health, InsideTracker, and WHOOP each offer valuable capabilities, most remain focused on specific aspects of the health journey.
Superpower’s advantage lies in its ability to bring these elements together into a single, cohesive system that prioritizes clarity, continuity, and action. It does not merely present data but transforms it into a guided experience that helps users understand their bodies, make informed decisions, and improve their health over time.
In a world where health is increasingly recognized as the most valuable asset, this shift from information to action is not just meaningful – it is essential.
Microsoft’s Azure cloud business accelerated in the March quarter, growing 40% and topping the company’s own forecast, giving the tech giant a new answer to questions about its ability to translate record capital spending on AI infrastructure into stronger financial results.
The company’s revenue rose 18% to $82.9 billion, beating the $81.4 billion analyst consensus, and earnings per share jumped 23% to $4.27, above the $4.06 expected by Wall Street.
AI run rate: In its earnings news release, Microsoft also disclosed that its AI business has reached an annual revenue run rate of $37 billion, up 123% from a year ago. It’s the first time the company has updated the figure since it reported a $13 billion run rate in January 2025.
Capex trends: Capital spending came down to $31.9 billion from $37.5 billion the previous quarter. Microsoft had said the decline would come and that it reflected the timing of data center construction and hardware deliveries, not a slowdown in demand for cloud and AI services.
Copilot: For the quarter, Microsoft 365 Copilot now exceeds 20 million paid seats, up from 15 million in January. That means about 4.4% of the company’s commercial base is on its paid enterprise AI plan.
Cloud overall: Microsoft Cloud revenue, which includes Azure, commercial Microsoft 365, LinkedIn, and Dynamics 365, rose 29% to $54.5 billion. The company’s remaining performance obligations, a measure of contracted future revenue, was $627 billion, with a significant part of that backlog tied to OpenAI.
Elsewhere in Microsoft’s business:
The results come three months after Microsoft’s stock dropped 10%, wiping out $357 billion in market value, despite the company beating expectations on revenue and earnings.
Investors focused on the record capital spending, a Copilot product that had reached just 3.3% of Microsoft 365’s commercial base at that time, and a revenue backlog heavily dependent on OpenAI.
The OpenAI relationship has shifted significantly since then.
This week, the two companies restructured their partnership, with OpenAI ending its exclusive commitment to Microsoft’s Azure cloud and gaining the ability to run its products on other platforms, notably Amazon Web Services. Microsoft, in turn, locked in its revenue-sharing arrangement and removed a clause that could have ended it if OpenAI had declared artificial general intelligence.
After suggesting a wood-burning stove, and a mini bellows, you should have seen this coming. What you need to complete the full-fire package is Cooking On Fire, a gorgeous book of recipes and techniques for cooking over an open flame. Cooking on Fire has a good mix of recipes, ranging from simple and delicious veggies to slow-cooked meats that require hours. There’s also plenty of background on different types of fires and cooking techniques, as well all the equipment you might want to cook various things (for example: spits, forked sticks, cast iron pans, and so on). It’s everything you—er, sorry, your outdoorsy friend—need to get started cooking on fire.
What I really want to try is the fire inside a log technique pictured on the cover, but I haven’t gotten around to that yet. So far I’ve only had a chance to make the grilled pork belly, with grilled carrots and “Krabbelurer” griddle cakes for desert. All of them were excellent, though of course, perhaps that universal rule applies more so here than with any other form of cooking: Your results may vary. In the end, though, this isn’t really a gift about cooking. It’s gift to remind us all to slow down and take your time, with food and everything else.
Gamers who remember the old days understand why CRT screens are so popular. Instant responsiveness to every command, no screen tearing, and images that feel alive in a way that newer flat screens cannot match. Found Tech has taken that magic and pushed it all the way to nearly 4K on something you wouldn’t expect, a dusty old IBM 275 monitor that’s been around since the turn of the century.
Found Tech began with a vintage IBM 275, which is roughly 20 years old and fairly strong for an office monitor constructed around a cathode ray tube. The problem is that modern graphics cards cannot generate the necessary signal to drive such high resolutions, therefore he relied on the “free” Intel graphics in his system. An RTX 4080 Super handled all the heavy rendering work while the Intel chip generated the final video feed. Custom Resolution Utility software let him dial in a 2880-by-2160 interlaced mode that standard drivers refuse to allow. After testing dozens of driver versions, the combination finally locked in.
Sale
CRTs can support this high-resolution setup because they function differently than modern monitors. Instead of a set grid of pixels, a CRT projects an electron beam across a layer of illuminating phosphors. The beam may draw lines as close together as the signal desires, and at this very high density, interlacing concerns simply vanish. The impression is that the lines mix together so smoothly that it resembles progressive scan. The tube’s analog nature takes care of the rest, providing edges with a softness that digital sharpening cannot imitate.
However, at this resolution, motion truly shines. LCD and OLED screens maintain each frame on the screen for a fraction of a second, resulting in a slight but consistent blur during fast panning and quick character movements, whereas a CRT lights up each phosphor for as long as the beam is over it and then fades away quickly. The result is incredibly crisp action, even with things racing across at full speed, and games feel much more immediate because you can see every frame just as the graphics card created it.
The depth is difficult to understand without seeing it for yourself, as the image looks to be sitting inside the tube rather than on its surface. Dark places remain pitch black, with no light leaking in. Bright highlights are super intense, and when speeding through the large open world of Crimson Desert at this resolution and 60 frames per second, the scenery begins to feel like it has depth and distance to it.
[Source]
Amazon Web Services growth accelerated to 28% in the first quarter — its fastest pace in nearly four years — pushing Amazon’s results past Wall Street’s expectations and validating, at least for now, the company’s controversial $200 billion capital spending plan.
Overall, Amazon posted sales of $181.5 billion, up 17%, and operating income of $23.9 billion, up 30%. Both topped guidance and exceeded Wall Street’s expectations of about $177 billion in revenue.
Profits were $30.3 billion, or $2.78 per diluted share. However, that included a $16.8 billion pre-tax gain on Amazon’s investment in Anthropic, which inflated the bottom-line numbers. Excluding that one-time gain, adjusted earnings per share would have been $1.61, just shy of analyst expectations of $1.62.
Amazon’s advertising business grew 24% to $17.2 billion in the quarter, and the company said advertising revenue topped $70 billion over the past 12 months.
In the core e-commerce business, unit sales grew 15%. Amazon CEO Andy Jassy called it the strongest growth rate since the waning days of the COVID-19 lockdowns. It was boosted in part by faster delivery, with more than 1 billion items shipped same-day or overnight in the U.S. so far this year.
Amazon is in “the middle of some of the biggest inflections of our lifetime,” Jassy said in a release.
The company spent $147.3 billion on property and equipment over the past 12 months, nearly doubling from $88 billion a year earlier and leaving just $1.2 billion in free cash flow.
In other words, Amazon is making more money than ever but plowing nearly all of it back into building out capacity, mostly for AWS and AI infrastructure. The company has said it expects to spend about $200 billion in capital expenditures for the full year.
Shares were down about 2% in initial after-hours trading.
Update: On the earnings call, Amazon disclosed that its AWS revenue backlog jumped to $364 billion, up from $244 billion last quarter. That figure does not include a recently announced deal with Anthropic valued at more than $100 billion. Jassy said the backlog has reasonable breadth beyond just one or two customers.
Jassy also disclosed that Amazon now has more than $225 billion in revenue commitments specifically for Trainium, its custom AI chip. He called Amazon’s custom silicon business one of the top three data center chip businesses in the world and said the chips business grew nearly 40% from the prior quarter.
For the second quarter, Amazon expects revenue between $194 billion and $199 billion, with operating income between $20 billion and $24 billion. The guidance assumes Prime Day will take place in Q2 in most countries, a shift from Q3 last year.
Microsoft will take a $900 million charge in its current quarter for its one-time voluntary retirement program, the company disclosed in its earnings report Wednesday.
Just to put that in context, it’s roughly equal to one day of revenue for the company at its current rate. Microsoft brought in $82.9 billion in its most recent quarter, ended March 31.
The retirement program is part of a broader reshaping of Microsoft’s workforce, which numbered 228,000 employees globally as of mid-2025, the last publicly released count.
Without giving specific numbers, Microsoft said Wednesday that its headcount declined year-over-year in the most recent quarter and will decline again in fiscal 2027, which starts in July.
On the earnings conference call, CFO Amy Hood said the company is focused on “building high performing teams that operate with pace and agility.”
At the same time, Microsoft plans to spend more than $40 billion on capital expenditures in the current quarter — a new record — primarily on data centers and AI infrastructure.
The voluntary retirement program, announced last week, is open to U.S. employees at the senior director level and below whose age and years of service add up to 70 or more.
Eligible employees are expected to get details May 7. They’ll have 30 days to decide.
At last count, Microsoft had about 125,000 U.S. employees, and the company has said about 7% are eligible for the program, so that would translate into about 8,750 eligible employees.
The program will include a financial payout and extended healthcare, but Microsoft hasn’t disclosed the specific terms of the plan yet.
The $900 million charge, split between $350 million in cost of revenue and $550 million in operating expenses, reflects the company’s estimate of the cost, including its assumptions about how many employees will accept, which is also number it hasn’t disclosed.
The program has sparked a range of reactions on LinkedIn and other social networks. Some employees and HR professionals have praised it as a more humane alternative to layoffs, noting that it gives longtime employees a choice rather than a pink slip.
But others have warned that Microsoft risks losing experienced engineers and leaders who built the systems the company still depends on, and some eligible employees have noted that being told they qualify for “retirement” in their late 40s doesn’t feel like a benefit.
Hackers are exploiting two authentication bypass vulnerabilities in the Qinglong open-source task scheduling tool to deploy cryptominers on developers’ servers.
Exploitation started in early February, before the security issues were disclosed publicly at the end of the month, according to researchers at cloud-native application security company Snyk.
Qinglong is a self-hosted open-source time management platform popular among Chinese developers. It has been forked more than 3,200 times and has over 19,000 stars on GitHub.
The two security problems impact Qinglong versions 2.20.1 and older and can be chained to achieve remote code execution:
The root cause in both flaws is a mismatch between middleware authorization logic and Express.js routing behavior.
“Both vulnerabilities stem from a mismatch between the security middleware’s assumptions and the framework’s behavior,” Snyk researchers explain.
“The auth layer assumed certain URL patterns would always be handled one way, while Express.js treated them differently.”
Snyk reports that attackers have been targeting these two flaws on publicly exposed Qinglong panels to deploy cryptominers since February 7.
This activity was first spotted by Qinglong users, who reported about a rogue hidden process named ‘.fullgc’ utilizing between 85% and 100% of their CPU power.
The name deliberately mimics “Full GC,” an innocuous but resource-intensive process, to evade detection.
According to Snyk, the attackers exploited the flaws to modify Qinglong’s config.sh and injected shell commands that downloaded a miner to ‘/ql/data/db/.fullgc,’ and executed it in the background.
The remote resource located at ‘file.551911.xyz’ hosted multiple variants of the binary, including for Linux x86_64, ARM64, and macOS.
The attacks continued with multiple confirmed infections across various setups, including behind Nginx and SSL, while the Qinglong maintainers only responded to the situation on March 1.
The maintainer acknowledged the vulnerability and urged users to install the latest update. However, the mitigation in pull release #2924 focused on blocking command injection patterns, which Snyk says was insufficient.
The researchers report that the effective fix came in PR #2941, which corrected the authentication bypass in the middleware.
AI chained four zero-days into one exploit that bypassed both renderer and OS sandboxes. A wave of new exploits is coming.
At the Autonomous Validation Summit (May 12 & 14), see how autonomous, context-rich validation finds what’s exploitable, proves controls hold, and closes the remediation loop.
For a few hours recently, global cloud platform Cloudflare went down – and a noticeable chunk of the internet went with it.
Services stalled, platforms froze, and businesses were left dealing with failures they didn’t trigger and couldn’t control – from broken checkouts to lagging internal tools.
Chief Information Officer at Tribe.
From the outside, it looked like the usual internet wobble. A quick refresh, a short wait, and most people moved on. Inside the organizations that depend on those upstream layers, the mood was very different.
Article continues below
Teams watched systems slow down or stop entirely, and it became clear – again – how easily a single fault in one provider can spill across the wider digital ecosystem.
And these incidents aren’t isolated.
Late last year, another Cloudflare issue hit – this time a configuration change that went wrong – and it took major platforms offline within minutes. Outages at global tech platforms and hyperscalers such as AWS have shown how disruption in one region can ripple outward at speed.
Even external events, like conflict in the Middle East, have knocked data centers offline and forced providers to reroute traffic under pressure. In addition, AI‑assisted changes have introduced unexpected instability.
The triggers vary, but the pattern is becoming familiar: one layer stumbles, and everything built on top of it feels the shock.
Most companies don’t think of themselves as heavily dependent on the likes of Cloudflare, AWS, or any other upstream provider in any meaningful way. In reality, they’re embedded across almost every part of modern digital services.
Today’s systems are built on stacked infrastructure: cloud platforms, routing and security layers, CDNs, authentication systems, and third-party APIs. Each layer adds capability, but also dependency.
When something breaks upstream, it doesn’t stay contained. It moves outward through those layers, affecting systems several steps removed from the original fault.
What catches people out is how far this travels. A failure in one provider doesn’t just hit its direct customers; it hits the platforms built on top of it, the tools those platforms rely on, and the businesses relying on all of them to operate. By the time the issue becomes visible, the blast radius is already far and wide.
Payments tend to expose these failures faster than most. A single transaction touches a long chain of systems: cloud infrastructure, fraud engines, authentication services, and processing networks. If one link snaps, the impact shows up immediately. Transactions hang, checkouts stall, and customers abandon carts.
That has a direct business cost within minutes, or even seconds.
But payments aren’t uniquely fragile; they’re just where people feel the outage the hardest – it’s their money on the line. The same dependencies run through ecommerce platforms, SaaS tools, logistics systems, customer support and internal operations. Payments just make the problem impossible to ignore.
That’s what makes the concentration of infrastructure such a risk. A relatively small number of providers now sit at the core of a huge share of digital services. That scale brings reach and reliability, but it also means the impact isn’t isolated: when something goes wrong, it rarely affects just one company.
Despite all this, outages are still treated as ‘we’ll deal with it when it happens’ moments, instead of something you actively design for.
Redundancy gets weighed against cost; dependency mapping is incomplete, and continuity plans sit untouched. And the assumption persists that major providers ‘don’t go down’, even though recent history shows that they most definitely do, and they will again.
There’s another issue behind all of this, and it’s one most organizations don’t say out loud: they no longer control the infrastructure they rely on.
Over the past decade, companies have handed more of their core systems to hyperscalers and a mix of third-party providers. It’s been great for speed and scale. But the trade-off is obvious every time an upstream service wobbles.
When something breaks several layers above you, there’s very little you can do except wait and watch the impact roll downhill.
Most organizations already have frameworks that are meant to cover this. But when the internet goes down, it’s not security or compliance that gets tested: it’s availability. And that’s the part that has to hold up in the real world.
In financial services especially, there’s a quiet move toward hybrid setups: keeping the big clouds in play but adding regional or specialist providers as backup routes.
Part of that shift is down to geopolitics as much as technology. Most European businesses still run on US cloud and AI platforms, and that’s fine until global politics get jumpy. Banks in particular are asking a simple question: what happens if those services get caught up in a political row they have no control over?
When a single outage can knock out authentication, routing or payments across a whole region, having another route stops being a nice-to-have and starts being basic self-preservation.
Different organizations are moving at different speeds, whether because of regulation or risk. But the message is clear: staying online isn’t luck. It’s something you build for.
The conditions that make these incidents so disruptive (shared infrastructure, tightly connected systems, real-time digital services) aren’t going anywhere. If anything, they’re getting more intense.
So, the real question for businesses isn’t whether another outage will happen, it’s how exposed they are when it does – and how quickly that exposure turns into lost revenue, operational disruption, and customer frustration.
Given how the modern internet behaves, staying online isn’t luck; it must be designed for. So, remember… do the basics to soften the blow: backup routes, a clear view of their dependencies, and no hidden single points of failure waiting to trip you up.
We’ve featured the best customer experience (CX) tool.
This article was produced as part of TechRadar Pro Perspectives, our channel to feature the best and brightest minds in the technology industry today.
The views expressed here are those of the author and are not necessarily those of TechRadarPro or Future plc. If you are interested in contributing find out more here: https://www.techradar.com/pro/perspectives-how-to-submit
Register Renaming | Hackaday
Weekend Open Thread – Corporette.com
Hyperliquid $HYPE Rally Builds Momentum as AI Sector Enters Prove-It Phase
Patterson-UTI Energy, Inc. (PTEN) Q1 2026 Earnings Call Transcript
IPL 2026: Ruturaj Gaikwad registers slowest fifty of the season, enters all-time unwanted list | Cricket News
Drax board avoid their own AGM, accused of greenwashing & environmental racism
LK Bennett closes all stores after entering administration
Michael Saylor says BTC winter is over. Market analyst disagrees, says bitcoin was in a pullback
Kylie Jenner’s KHY Enters a New Era with ‘Born in LA’
Images of Samsung’s rumored smart glasses have leaked
Mariah Carey Slams Deposition Claims In Brother’s Lawsuit
Is Algorand One of the Few Quantum-Resistant Blockchains? Here’s What the Data Shows
Most Commercial Energy Audits Miss the Real Losses
Why Blue Badges Disappeared From Toyota Hybrids
Trump threatens to review UK’s claim to Falkland Islands and punish Nato allies over Iran war disagreement
Jeanine Pirro announces closure of Federal Reserve building cost probe
(VIDEO) Charlize Theron Climbs Times Square Billboard to Promote New Netflix Thriller ‘Apex’
Kraken Builds Beyond Crypto With Futures, Tokenized Stocks, and Payments
Microsoft to roll out Entra passkeys on Windows in late April
CFTC’s AI will review U.S. crypto registration applications, chairman tells CoinDesk
You must be logged in to post a comment Login