OFFBEAT

Turns out this space thing really is hard

Virgin Galactic has delayed its return to commercial spaceflight until February 2027, and plans to raise ticket prices later this year.

The delay was disclosed alongside the company’s financial results, which showed a net loss of $56 million for the second quarter of 2026, down from $67 million a year earlier. Revenue for the quarter was $0.1 million, compared to $0.4 million in 2025.

According to CEO Michael Colglazier, demand exceeded the number of seats offered in the first $750,000 batch, prompting the company to prepare another at a higher price. Those customers will, however, have to wait a little longer. Colglazier said: “Our first ship is now expected to enter commercial service in February 2027 rather than the fourth quarter of 2026.”

He blamed the delay on the extra time needed to “complete avionics and systems installations.”

During an earnings call, Colglazier said: “No single issue is driving the schedule push. Rather, we have experienced modest time duration extensions across hundreds of relatively small but important installation tasks involved in the first build of our new spaceship.”

In response to an analyst question, Colglazier elaborated: “The number of those kind of ‘Oh, we did not expect this to not fit just perfectly,’ coming in is higher than we had allotted for. That just has started to accumulate on us. It really picked up at the tail end of July. For a bit, we thought we could manage that end, but the team just needed more time to do it the correct way.”

Integrated vehicle ground testing is expected to begin later in August, followed by flight testing in October. A second spaceship is due to join the fleet in March 2027, and the company expects to stop burning cash and “deliver positive quarterly cash flow within 2027.”

When Virgin Galactic reopened suborbital ticket sales in April, it charged $750,000 for a seat, up from the $600,000 price it cited in 2023. That was a substantial jump from the $100,000 envisaged more than two decades ago, when Sir Richard Branson announced plans for a scaled-up version of Burt Rutan’s SpaceShipOne. At the time, the company said commercial flights would resume by the end of 2026.

Virgin Galactic’s last commercial flight took place in 2024, after which the company paused operations to focus on its next generation of spacecraft. Virgin Galactic is not alone in pausing its space-tourism service. Earlier this year, rival Blue Origin announced that New Shepard flights would pause for “no less than two years” while it worked on its crewed lunar program. ®

Updated 8/18 at 10:19 GMT: A previous version of this story said that the program was “grounded,” but the more appropriate term is “paused.”

The first Ferrari Luce off the production line has sold for big money, with the Jony Ive-connected vehicle fetching a staggering $40 million at auction.

The car company’s first production electric vehicle, the Ferrari Luce, is an attempt to take on a challenging and growing market. While it has had a troubled existence since launch, the very first one off the production line has been sold for a considerable sum.

The Ferrari Luce “Tailor Made” went under the hammer at RM Auction in Monterey, California, late on Sunday. The car, bearing the chassis number 0, was a highlight sale for the event, reports The Autopian

Normally sold starting from $600,000, the opening bid for the special lot was $1 million, before it immediately shot up to $5 million and beyond. It eventually capped out at $40 million.

The price is one of the highest ever spent on a Ferrari, let alone a brand-new model. In 2025, a 1964 Ferrari 250 LM sold for $40.3 million, and that car had a history to help push up the price.

This particular edition is not just about the chassis number, as it has an exclusive Madreperla Semi-Gloss finish specifically made for the car. It also uses Le Mans metallic leather in Perla, and interior elements are finished in Grigio Corvara instead of black.

Part of the reason for the high price is that it’s a charity lot. 100% of the money raised from the sale is going to Save the Children.

It’s likely that the buyer will benefit from tax write-offs with the purchase.

Good news for a put-down car

The Ferrari Luce has been an unusual vehicle for the automaker to design and produce. Not least because of its Apple connection.

Former Apple design chief Jony Ive and his LoveFrom design studio helped design the outside and the inside of the Luce. That included the steering wheel and physical controls at the hands of the driver.

However, the car hasn’t had the best reception since its introduction. Rather than the classic Ferrari styling, the car instead looks a lot more like a typical road-going vehicle.

It’s not even Ferrari Red, with the company instead going for a more pastel color approach.

It does have Ferrari speed, though, with a combined 1,035 horsepower from four motors and a 0-60mph of 2.5 seconds.

Bottom line: A study of Backblaze’s hard-drive fleet found that HGST and Western Digital drives had lower failure rates than Seagate and Toshiba drives after accounting for factors such as age, capacity, operating temperature, and form factor. The results show how raw fleetwide failure figures can obscure meaningful reliability differences when manufacturers’ drives are at different stages of their service lives.

The peer-reviewed IEEE study was written by Christoph Siemroth of the University of Essex and Yeomyung Park of Sungkyunkwan University. It reviewed data on 443,156 drives used in Backblaze data centers from 2013 through the second quarter of 2025. The sample covered more than 1.66 million drive-years.

The researchers found that HGST drives failed at about 41% of Seagate’s rate when compared under similar conditions. WD drives failed at about 52% of Seagate’s rate. Toshiba drives had a rate equal to 107% of Seagate’s rate.

HGST ranked first in the analysis, although the brand is no longer sold as a separate new-drive option. Western Digital acquired HGST in 2012 and later phased out the brand.

The study takes a different approach from Backblaze’s regular quarterly reliability reports, which compare the drives active in the company’s storage fleet at a given time. The authors said that can produce uneven comparisons because the fleets are not the same age.

Number of HDDs installed by manufacturer and year.
Backblaze’s drive mix shifted sharply over time, a factor the study accounted for when comparing manufacturer failure rates.

HGST accounted for many new Backblaze installations in 2014. Seagate was the main supplier from 2015 through 2020. Toshiba had the most new installations in 2023, and WD led in 2024. As a result, reported failure rates can compare old Seagate and HGST drives with newer Toshiba and WD models.

The researchers adjusted their analysis to account for those differences. They also excluded some of the bias created when drives leave service before they fail. About 146,943 drives, or 31% of the sample, were removed from Backblaze’s dataset without a recorded failure. Most were likely replaced as the company moved to higher-capacity hardware.

Toshiba showed the clearest change in failure behavior as drives aged. Its monthly failure rate rose more than fourfold after 60 months in service. From 50 to 100 months, Toshiba drives failed at a rate of five to six per 1,000 drives each month. The other three brands generally remained at or below 2.5 failures per 1,000 drives per month over that period.

Seagate had the highest failure rate among newer drives in the study. But its rate did not increase as sharply with age as Toshiba’s did. Backblaze has previously reported that many drive failures occur before a drive has been in operation for three years.

Operating temperature also mattered. The study found that each 1-degree Celsius increase in average drive temperature raised the failure rate by 2.1%. A 10-degree Celsius increase would raise the rate by about 23% when compounded.

Higher-capacity drives performed better in the model. Each additional terabyte of capacity was linked to a 3.4% decline in failure rates. The researchers said that may be partly tied to newer drive designs, including helium-filled enclosures and advances in head manufacturing.

The results do not necessarily apply to hard drives used in consumer PCs or home storage systems. Backblaze operates enterprise drives in data-center conditions, and the dataset does not include comparable workload data for each manufacturer.

The analysis is also weighted toward shorter service periods. Only 0.52% of drives in the study stayed in operation for more than 10 years. That makes the results more useful for evaluating reliability during a drive’s early and middle years than for estimating its full lifespan.

WD and HGST finished close together in the adjusted ranking. The authors said that could reflect technology sharing after WD acquired HGST. Backblaze’s most recent annual report put the company’s overall annualized drive failure rate at 1.36% in 2025.

If you really want one, you can make it work.

Bid farewell to OxygenOS

Adnan Ahmed for Engadget

High-risk, high-reward

Adnan Ahmed for Engadget

Live at the Honda Center in Anaheim on Friday night as part of Disney’s D23 fan convention, Lucasfilm President Dave Filoni announced some of what’s coming next in the Star Wars universe.

And perhaps the biggest reveal was a trailer for the upcoming film, Starfighter.

Starfighter

Corinne Reichert/CNET

Disney’s D23 event gave fans a peek at the new chapter in the Star Wars franchise during its entertainment presentation on Friday, revealing a trailer for Lucasfilm’s Star Wars: Starfighter film. The movie was announced in early 2025 and is due to hit theaters on May 28, 2027. The film features Ryan Gosling and Flynn Gray portraying the lead characters. 

Gosling, in a jacket with a skull and crossbones, was on hand to reveal fresh details about his upcoming Star Wars film. The emblem is the signature of his character, whose name was revealed to be Kade Auberon — not one franchise fans should recognize, but certainly one that sounds like it came from a galaxy far, far away.

Gosling called it a life-changing role before introducing the trailer.

Set five years after the events of 2019’s The Rise of Skywalker, the space-faring tale teases Gosling riding “the fastest starfighter ever built.”

Gosling plays a talented pilot, while Gray is the young boy he’s charged to protect. The packed cast also includes Aaron Pierre, Mia Goth, Amy Adams, Daniel Ings, Jamael Westman, Simon Bird and House of the Dragon’s Matt Smith. 

Shawn Levy is directing the adventure flick. At the Star Wars Celebration in Japan last year, he explained how this story is in a lane of its own. “This is a stand-alone,” he said. “It’s not a prequel, not a sequel. It’s a new adventure. It’s set in a period of time that we haven’t seen explored yet.”

It’ll hit theaters on May 28, 2027.

Ahsoka season 2: Hayden Christensen is back as Anakin

Corinne Reichert/CNET

Joining Filoni on stage were Ahsoka live-action actress Rosario Dawson; Eman Esfandi, who plays Ezra Bridger; and, surprisingly, Hayden Christensen, who played Anakin Skywalker in Star Wars Episodes II and III.

Anakin was a big part of the original Ahsoka Tano story in the animated Clone Wars series, and Christensen’s return to the character in a pivotal episode during the first season of Ahsoka was arguably its best. It makes sense for him to return in season two, seemingly in flashbacks of Ahsoka’s past and as a guiding Force ghost in her present.

The teaser trailer showed Christensen leading Ahsoka to, perhaps, become a Jedi once again.

Ahsoka season 2 will stream on Disney Plus on Jan 20, 2027.

Star Wars IV coming back to theaters

To celebrate the 50th anniversary of Star Wars, which was released in 1977, the original movie will be rereleased in its original form in theaters next year.

Kourtnee Jackson

Senior Editor

Kourtnee covers streaming services and entertainment. She previously worked as an entertainment reporter at Endgame 360, where she wrote about film, television, music, celebrities and streaming platforms.

See full bio

Corinne Reichert

Senior Editor

Advertisement

Corinne Reichert (she/her) grew up in Sydney, Australia and moved to California in 2019. She holds degrees in law and communications, and currently writes news, analysis and features for CNET across the topics of electric vehicles, broadband networks, mobile devices, big tech, artificial intelligence, home technology and entertainment. In her spare time, she watches soccer games and F1 races, and goes to Disneyland as often as possible.

See full bio

• America’s aging nuclear missiles depend on data scattered across dozens of systems
• Minuteman III could remain operational decades beyond its original retirement schedule
• The US Air Force wants one AI interface to organize fragmented missile data

The US Air Force is searching for an artificial intelligence agent capable of centralizing the fractured data behind its aging nuclear missile fleet.

An active industry notice indicates that the Minuteman III program relies on roughly 60 disparate, disconnected systems.

Cryptocurrency hardware wallet provider SafePal is warning of a data breach affecting about 39,798 customers after a flaw was exploited to steal customer order information, and a threat actor is now claiming to be selling the stolen data.

SafePal says the breach impacts customers who placed orders between March 2, 2025, and April 11, 2026, exposing their names, email addresses, shipping addresses, phone numbers, and purchase information.

The company says the breach did not expose customers’ wallet seed phrases, private keys, passwords, bank account information, payment card numbers, government-issued identification numbers, or other credentials.

“No evidence has been found that the incident itself compromised access to SafePal wallets or funds,” SafePal said in a security advisory published Sunday.

The company says it notified all impacted customers via email on August 16 with the subject “[Important] Your SafePal Order Information Has Been Affected.”

SafePal has also launched an online verification tool that lets customers enter their order number and shipping country to determine whether the details of that order were stolen.

The company warns that the stolen information could be used to conduct targeted phishing and other social engineering attacks, with customers reporting SafePal phishing emails and phone calls as early as May.

Order-tracking flaw exposed customer data

A threat actor now claims to be selling the stolen SafePal customer data on a cybercrime forum.

As spotted by DarkWebInformer, the seller referenced the same affected order period and approximately 39,798 customers disclosed by SafePal.

For potential buyers, the threat actor is also willing to share order ID and shipping country information from stolen orders, which can be confirmed on SafePal’s online verification tool as proof that the sale is legitimate.

“Not interested in low balls , please come correct and with a good price or do not message me at all,” reads the forum post.

BleepingComputer has not independently verified that the threat actor possesses the stolen data.

SafePal says it first received a report consistent with the incident in early May 2026, which it initially treated as an isolated case.

While it is unclear whether this report is related, a customer posted on X that they received a SafePal phishing email and a phone call from someone claiming to be a company employee in May. The phishing email claimed that a security vulnerability had been discovered in the SafePal X1 hardware wallet and that a firmware update was required to fix the flaw.

“We first received a report consistent with this issue in early May, and treated it as an isolated case at the time, but escalated it into a formal security investigation and introduced additional protections,” reads the advisory.

“As our e-commerce system involves multiple interconnected components and external integrations, as well as third-party logistics partners, we could not immediately rule out several possible explanations.”

In July, SafePal began what it described as a “full review and rebuild” of its order-processing system and discovered an authorization flaw in the order-tracking function of a plug-in that allowed unauthorized access to another customer’s order information.

SafePal says it fixed the vulnerability and implemented additional security measures. The company is also working with a third-party security firm to validate the fix and conduct a broader review of its order-processing systems.

However, as part of this investigation, SafePal determined that a threat actor exploited the flaw to steal order information belonging to approximately 39,798 customers.

During the investigation, SafePal also discovered a separate configuration error that caused a data-cleanup process to stop functioning correctly between September 2025 and April 2026, resulting in order data being retained as far back as March 2025.

For affected orders, SafePal says it has purged personal data from active e-commerce servers, although it is retaining an encrypted offline copy for potential law-enforcement investigations.

SafePal also warns customers to watch for targeted phishing emails and phone calls about firmware upgrades, product returns, refunds, or legal investigations.

The company says it has already taken down more than 30 fraudulent websites and phishing links tied to this incident.

Customers whose order information was exposed do not need to replace their hardware wallets or move cryptocurrency because of the breach, according to SafePal.

However, if a customer already shared their seed phrases or private key in response to a phishing email or text, they should treat their wallet as compromised and transfer any assets to a new wallet on a trusted SafePal device or official application.

Overall prevention scores can hide what happens after initial access. Once attackers are using valid credentials, prevention drops sharply.

The Blue Report 2026 measures defenses technique by technique across 338 million simulations run in customer production environments.

Advertisement

Get the report

Most teams building retrieval augmented generation (RAG) systems for high stakes classification make the same architectural bet: Route every ambiguous case straight to the language model and trust the retrieved context to sort it out. This works fine in a demo. It falls apart the moment the system has to survive an audit, a regulator, or a compliance officer asking why a specific decision was made six months ago.

I have spent the last year building RAG based classification systems in regulated enterprise settings, where the cost of a wrong answer is not a bad chatbot reply. A decision has to hold up to scrutiny long after the model produced it. This environment forces a different design philosophy than most AI engineering content assumes.

Here is what changes when you cannot afford to be probabilistic about everything, and how a cascade architecture solves it.

The invisible cost of an all LLM pipeline

The appeal of routing everything through a large language model (LLM) is obvious: Fewer moving parts, faster iteration, the model handles unanticipated edge cases. The problem shows up later, in three places.

First, auditability. “The model decided based on retrieved context” is not an acceptable answer. You need a decision path a human can reconstruct without rerunning inference and hoping for the same output.

Second, cost at scale. If your system processes tens of thousands of cases a day and every one hits an LLM call with several retrieved documents in context, your inference bill and latency both scale with volume in a way that rule based logic does not.

Third, and least discussed, model drift on the easy cases. LLMs are excellent at nuanced judgment calls. They are inconsistent, in ways that are hard to detect, on cases that should have a deterministic answer. A clear structured match against known criteria should never depend on a language model’s mood.

The cascade approach

The fix: Stop treating the LLM as the front line and start treating it as the escalation path. In practice this means a three stage pipeline.

Stage one is deterministic. Exact matches, structured field comparisons, and anything with a clear rule get resolved here with no model call at all. This stage should clear the majority of volume, often more than half depending on your data quality, and every decision is fully explainable because it is a lookup, not an inference.

Stage two is where retrieval earns its keep. For cases that survive stage one — and I mean survive as in they were not clearly resolved — you build a retrieval layer that pulls the specific evidence relevant to the ambiguity: Prior reviewer decisions on similar cases, contextual documents that explain an apparent conflict, or historical precedent that clarifies an edge case. The retrieval step matters more than the generation step here. If you retrieve the wrong context, even the best language model in the world will produce a confident, well reasoned, wrong answer.

Stage three is the LLM call, and it should only see the residue that stages one and two could not resolve. This is the part people skip when they design their first version, and it is the single biggest lever for both cost and quality. In one system I worked on, routing only the genuinely ambiguous 10 to 15% of cases to the LLM cut inference cost by roughly 6X compared to an all LLM baseline, while improving consistency on the deterministic majority to effectively perfect.

Designing the prompt for asymmetric risk

Once a case reaches the LLM stage, most teams default to a neutral prompt: “Assess whether this case should be approved or flagged.” That framing is wrong for high stakes classification because the cost of the two error types is not symmetric. Missing something that genuinely needed attention can mean real harm downstream. Incorrectly flagging something that was fine costs a reviewer’s time and a delay. Those two outcomes are rarely equally bad, yet a neutral prompt asks the model to treat them as if they were.

An asymmetric risk prompt makes that tradeoff explicit to the model rather than letting it guess at your risk tolerance. Concretely, this means instructing the model to treat uncertainty as a reason to escalate rather than clear, providing calibrated examples of both error types with their consequences spelled out, and asking for a confidence score alongside the classification rather than a binary answer. The confidence score becomes your second cascade point: Anything below a certain threshold goes to a human reviewer instead of being auto resolved, no matter what the model’s classification says.

This sounds like a small prompt engineering detail. In practice it is the difference between a system that reduces reviewer workload and one that quietly increases risk while looking like it is working.

Evaluating a system like this properly

Standard RAG evaluation metrics were not built with this use case in mind, and using them without adaptation will give you a false sense of confidence. A few adjustments that matter.

Retrieval quality needs to be measured separately from final classification accuracy. A system can have excellent retrieval ranking scores and still make bad final decisions if the generation step misweights the evidence. Track them independently.

Your evaluation set needs deliberate oversampling of the cases that reach stage three, since that is where your system’s judgment actually gets tested. If your eval set mirrors your production distribution, it will be dominated by the deterministic cases your cascade already handles well, and you will be blind to exactly the failures that matter most.

LLM as judge evaluation works for this domain but only if the judge prompt encodes the same asymmetric risk framing as your production prompt. A judge that treats both error types equally will systematically favor the wrong tradeoff when you are tuning your system.

Finally, build a feedback loop from confirmed outcomes back into your retrieval corpus. When a human reviewer overturns a model decision, that case and its correct resolution should become retrievable context for future similar cases. Without this, your system’s handling of ambiguous cases never improves, it just keeps making the same category of mistake at the same rate.

The broader lesson

The instinct to reach for the most capable model for every decision is understandable, but in domains where wrong answers have real consequences, the more valuable engineering work is deciding what should never touch the model at all. Cascade architecture is not a workaround for LLM limitations. It is what a mature RAG system looks like once you have actually had to defend its decisions to someone whose job is to find the flaw in your logic.

If you are building AI systems for any regulated or high stakes domain, the question worth asking before you write a single prompt is not “How do I get the model to handle this well.” It is “Which parts of this decision should never have been the model’s job in the first place.”

Vineet Vijay is a Lead AI and machine learning engineer.

Welcome to the VentureBeat community!

Our guest posting program is where technical experts share insights and provide neutral, non-vested deep dives on AI, data infrastructure, cybersecurity and other cutting-edge technologies shaping the future of enterprise.

Read more from our guest post program — and check out our guidelines if you’re interested in contributing an article of your own!

• Private drone companies can now request access to US Army testing ranges
• Companies no longer need existing government partnerships before requesting range access
• Five military facilities across America and Morocco are joining the programme

Private drone companies without existing government contracts can now request access to the US Army test ranges nationwide.

The arrangement removes a requirement for companies to have existing government partnerships before requesting access to military ranges.

Quick summary? 

Do you believe in the power of positive journaling, but for some reason, you can’t take out the time or the effort to put pen to paper? If the answer is yes, and the above question describes your predicament, Adiaro may have a neat solution for you. The company makes jewelry in a whole bunch of styles and colors, but the center of it all is the charm gemstone that comes equipped with an NFC tag. 

All you need to do is tap your phone against the gem, and it opens a digital journal for you that is also linked to your Google or Apple account. Does it work? Yeah, if you mind the end result, and not the medium. It comes in a variety of gem choices, too. So, if you are someone who believes in the healing power of certain gems, consider this bracelet as the perfect hybrid solution that asks about your mood, lets you vent, use it as a diary, and build an emotional calendar. 

What is this? 

It’s a bracelet that wants to blend the spiritual side of gem-based healing with the scientifically proven benefits of journaling. How? Well. at the heart of the gemstone is an NFC tag, and when you tap your phone on it, you get access to a web-based journal app. It works pretty reliably, though, on a few occasions, I’ve found myself inadvertently triggering the notification by simply placing the phone near the Adiaro bracelet in my pocket, or on the table. 

I tried the Adiaro Tiger Eye Balanced Chain bracelet, but you can also pick between amethyst, obsidian, white and golden jade, rose quartz, and turquoise options. You can also try an entirely different style with the bead-heavy grounded bracelet, and separately purchase the gem charms in different shapes and colors at $19 apiece. 

Attaching them is pretty easy, thanks to a magnetic clasp system that looks like a chrome bead and separates into two halves. As far as the build quality goes, the Balance chain feels pretty high-end. The metal finish is pretty sharp, and there are no loose ends or poor craftsmanship visible on this one. 

How does it work? 

Well, think of this bracelet as more of a hidden digital diary than a piece of jewelry. As soon as you tap your phone against the gem charm, it triggers a notification that subsequently opens a web-based digital journal app. At the top, you see a thoughtful quote that varies on a day-to-day basis.

Underneath, you can pick between five different moods to reflect your emotional state in that moment. The spectrum starts at “overwhelmed” and slowly progresses to “amazing.” And finally, there’s a dedicated text field where you can describe your mood in 800 characters. 

It’s a tad limiting, especially considering the fact that voice-typing on a phone is a lot quicker and far more convenient than writing in a diary using a pen. Once you are done with your digital journaling, it is saved as a log that is synced with your Google Drive or iCloud Drive. 

All your mood entries and ruminations are recorded. The Adiaro app creates a progressive map that shows how your mood has changed over the past few days and what kind of thoughts occupied your mind. I like the approach. It’s almost like taking a look into the past few days and assessing how your mental state and energy levels have been at a certain time of the day, or midway through a certain activity.

The broad idea is that, aside from just journaling through a day, you can also check the mood calendar and try to find patterns, depending on the details you have shared in your notes and how your mental state has been. It may not work for everyone. But for some, it can prove to be beneficial.

For example, over a span of three weeks, I noticed that my mood affirmations were usually amazing at noon, and I was high on energy and willing to take on challenging assignments. As the evening passed, I was generally tired, even though my energy levels were consistent, but I felt a certain sense of stress. 

Acting on the data at hand, I tried to adjust my work hours and shifted low-priority tasks to the evening slot. It was not a magical cure, but it did help me accomplish a few pending tasks that I kept postponing and felt stressed about. Your mileage might vary depending on how you journal and what kind of actionable insights you can get from the mood calendar.

What about the science? 

I have never maintained a diary for journaling, nor have I poured my heart out on an obscure blog. But I know a lot of people who do it, and I see them feel good about it. One of my friends says it’s an escape route from the events that give them anxiety and offers a safe space to vent it all out. Broadly, they see it as a channel of relief. 

Adiaro is building on that relief pathway. It might sound like a placebo, or make-believe kind of solution, but there is scientific backing behind the approach. A research paper published in the peer-reviewed JMIR Mental Health journal, positive affect journaling (PAJ) can tone down mental distress and trigger and enhance well-being, serving as an effective intervention. 

As part of a randomized trial, patients living with different medical conditions and elevated anxiety symptoms were asked to perform a 15-minute web-based journaling task. The task was performed thrice a week, and the whole experiment lasted 12 weeks. At the end of the trial, the team found that web-based positive affect journaling was linked to lowered mental distress among the participants, lower anxiety, and perceived stress. 

It also said to have triggered greater perceived personal resilience and encouraged social integration. “Overall, the findings from this study suggest that PAJ has potential utility as an intervention for managing mental distress, particularly elevated anxiety symptoms, and other aspects of well-being among general medical patients. This is consistent with, and extends, prior research on positive writing interventions as a way to improve aspects of health and well-being,” says the research paper that was published in 2018.  

Another paper published in the Family Medicine and Community Health journal arrived at a similar conclusion after an analysis of multiple clinical trials and peer-reviewed papers that explored the link between journaling and mental health. Broadly, it sees journaling as a low-cost, low-risk method to manage mental health symptoms, but not as a replacement for proper medical care. 

“This meta-analysis provides further affirmation that journaling as an intervention has merit and can be an efficacious adjunct when prescribed and implemented properly. The findings of this study can be applied in primary care practice by using journaling as a low-risk, low-resource intensive adjunct to standard therapy for patients with mental health concerns,” notes the research paper.   

Does that mean Adiaro’s smart jewelry is one of those helpful solutions described in scientific research on the benefits of journaling? Not explicitly, but it gets pretty close to the fundamental. At the end of the day, it’s a journaling solution that also comes with the added benefit of a mood calendar. Only this time, instead of opening a diary, you tap a phone against the bracelet’s jewel and tap on a screen. 

Should you buy? 

If you are someone who is not averse to the idea of wearing a bracelet on a daily basis and want to give journaling a try, Adiaro’s bracelet is worth a try. It looks gorgeous, doesn’t ask you to pay for a subscription, and doesn’t even require an app download. All your journals are neatly synced to your Google account, and the mood calendar comes in handy without slapping you with a learning curve.

If you believe in the power of positive journaling, you should get one. The only caveat is that it costs nearly as much as a fitness band, and some of them actually come with guided meditation exercises. They don’t quite let you do the journaling part, but they offer a lot more value, especially if you’re someone who also wants to keep an eye on your body vitals and track your exercises. Alternatively, you can just try Apple’s Journal app on the iPhone, a mood tracker app, or have an AI chatbot like Gemini build your own mood dashboard in a few minutes.

How we tested 

I tested the Adiaro bracelet as a smart journal for a period of three weeks. In that span, I wore it on a daily basis and used it to log my mood and jot down my thoughts. I did not bind myself to a fixed routine for journaling, and I only did it when I felt the need to note down my positive thoughts and tired ruminations. I used an iPhone 17 Pro and used my Google account to sync and save my journal notes.

FAQ (Frequently Asked Questions) 

Is it customizable? 

Yes, you can get it in a variety of styles, colors, and gemstone choices. 

What sizes does it come in? 

It offers free size adjustment. The metallic band can be adjusted to fit wrist sizes from 15 to 22 centimeters (5.9 to 8.7 inches).

Can you change the charm gemstone? 

Yes, the bracelet comes with a magnetic stainless steel charm system that lets you easily swap the gemstone.