Technology

• INIU Pocket Rocket P5 is super small
• It has 10,000mAh and multiple ports
• It comes in macaron colors but not flavors, for obvious reasons

If you like your power banks small, full of energy, and the color of your favorite macarons, INIU might have you covered.

The company, best known for constantly innovating power cell stacking to create increasingly smaller and lighter power banks, introduced this week what it claims is "the World's smallest 10,000mAh, 45W fast-charging" power bank.

The Pocket Rocket P50 (don't look at us, we didn't name it) is indeed small. Measuring 3.3 x 2.0 x 1.0 inches, the P50 weighs just 5.6 oz. Similarly configured 10,000mAh power banks on Amazon tend to weigh a few ounces more and are slightly larger.

They also generally cost a little more. The Pocket Rocket 50 lists for $32.99 (£38.99) on Amazon.

(Image credit: Iniu)

INIU achieved the P50's pleasingly small size by using its trademark TinyCell Pro technology, which the company says uses "efficient cell arrangement and space-saving thermal layers." It also come equipped with a small monochrome display that offers real-time charge status.

The P50 includes multiple charging ports, including a USB-A port and two USB-C ports. The attached lanyard doubles as a USB-C-to-USB-C charge cable that you can use to charge devices connected to the 45W power bank and to recharge the P50.

Available in a collection of macron-style colors that include pink, green, purple, and blue, the Pocket Rocket P50 can deliver a 45W charge and supports Samsung Fast Charging 2.0 for a speedy top-off.

INIU claims the P50 can charge a smartphone from 0% to 73% in just 25 minutes. Naturally, this is a claim we'll want to verify in lab testing.

(Image credit: Iniu)

The P50, according to the company, is capable of recharging multiple devices at once, and, on a single charge, can fully charge an iPhone 16 twice as well as an iPad mini or a Samsung Galaxy S24 one and a half times. INIU also claims the Pocket Rocket P50 is approved for carry-on use.

It's certainly small enough to fit anywhere, and with those tasty colors, it might attract more than a few wistful stares at the airport.

You might also like

• Fujifilm’s X half is a tiny retro compact that’s big on wacky film photography features, and I love it
• I tried Project Moohan and loved it, but only because of Google Gemini
• United Airlines just unveiled its next-gen cabin with new Polaris seats – here are the biggest changes

• Many freelancers are now thriving by offering emotional nuance which AI consistently fails to deliver
• Visual content creation roles are rising as personal style outperforms algorithmic predictability
• The backlash against AI slop is pushing businesses to revalue human creativity and depth

While generative AI tools continue to dominate headlines and reshape workflows, demand for creative freelancers appears to be growing, not shrinking.

Figures from the Freelancer Fast 50 Global Jobs Index found in Q2 2025, job postings for writers, designers, and video editors are climbing steadily - even as roles in machine learning, blockchain, and other AI-adjacent fields show marked declines.

The shifts suggest businesses are drawing clearer lines between automated output and the type of nuanced, human creativity that machines still fail to replicate convincingly.

Originality rises as slop loses appeal

The findings are based on more than 251,000 projects posted on a leading freelance site during the second quarter of 2025.

Communications jobs surged by 25.2%, making it the fastest-growing category, with freelancers in this space are being hired to craft contracts, edit manuscripts, and produce emotionally resonant writing that AI tools struggle to deliver.

This trend emerges amid what some commentators have described as widespread “AI slop fatigue”.

This is a growing pushback against the mass of bland, automated content that has flooded social media and search platforms.

The fatigue may be both aesthetic and functional, as platforms such as Google have introduced algorithm updates designed to penalise auto-generated material, putting further pressure on brands to prioritise originality.

Clients now appear more willing to invest in skilled professionals who can ensure their content maintains visibility and emotional resonance.

Many are still using AI writer programs in support roles to brainstorm ideas or speed up drafts, but final outputs are increasingly expected to pass a test of authenticity that machines fail to meet.

In video and visual production, the shift is just as pronounced, as job listings for skills such as Adobe After Effects, Instagram content creation, and 3D design using Unity have all posted double-digit gains.

Content creators are not just surviving alongside AI; they are thriving in areas that rely heavily on personal style, spontaneity, and audience connection.

Freelancers interviewed for the report describe growing interest in projects that range from low-budget films to custom branding efforts, with clients favouring professionals who can offer “strategic thinking” and “tailored solutions.”

This growth in creative jobs also underlines a broader recalibration of the role of AI tools.

Instead of displacing freelancers, many organisations are shifting toward hybrid workflows, leaning on machines for efficiency while entrusting humans with the final creative direction.

The simple conclusion to this situation is that for now, human nuance still matters.

You might also like

• These are the best AI website builders around
• Take a look at our pick of the best internet security suites
• Is the cloud the wrong place for AI?

• Apple's just promised that future iPhone and Apple Watch models will feature US-made Glass
• It's not an entirely built-in the USA iPhone or Apple Watch
• It's part of a larger $600 billion investment in the US economy from Apple

While it’s not an iPhone that’s entirely made in the U.S.A., Apple is making some pretty major hardware-related news alongside a fresh commitment from the Cupertino-based tech giant to invest a total of $600 billion in the U.S. economy within the next five years.

Apple, in a just-announced partnership with Corning, will aim to make and produce all of the glass covers for the iPhone and Apple Watch in the United States – specifically at Corning’s facility in Harrodsburg, Kentucky. It’s part of a new $2.5 billion commitment from Apple and means that once in place, all the glass for the iPhone and Apple Watch models sold globally will be made in the United States.

Apple’s partnership with Corning is far from new. While Apple rarely explicitly names who makes which components, it’s long been known that they use some custom form of Corning Gorilla Glass. Corning has always been a US-based company. The news that all iPhone and Apple Watch glass manufacturing is coming to the US inadvertently reveals that Apple may have been using multiple glass suppliers, including some from outside the US. That all changes now, though.

(Image credit: Apple)

Most recently, this facility has been producing glass that’s named ‘Ceramic Shield’ for Apple’s iPhone lineup. The Harrodsburg, Kentucky, facility will exclusively be used for making glass for Apple devices going forward. The release notes that this decision will increase Corning’s manufacturing and engineering workforce here by 50% and that a combined Apple-Corning Innovation Center will open nearby.

(Image credit: Future)

At a joint conference held at the White House and attended by Apple CEO Tim Cook, US President Donald Trump stated that this is a "smart glass production line" and will ultimately create 20,000 new American jobs.

Cook actually gave Trump a present, well, a gift from Apple – a piece of Corning Glass with ‘Trump’ engraved on it, and a base made from 24 karat gold sourced from Utah. It might be the first unboxing on the Resolute Desk, at least performed by Apple’s CEO.

The bigger picture: Apple’s upping its promised US investment

While this is the major hardware-related news as part of Apple’s commitment, the company did promise an additional $100 billion investment United States. Previously, the total investment was $500 billion, and that jumps to $600 billion, which should be complete within four years.

Alongside the new partnership with Corning, Apple’s also committed to working further with other US manufacturers like Coherent, GlobalWafers America (GWA), Applied Materials, Texas Instruments (TI), Samsung, GlobalFoundries, Amkor, and Broadcom. This is dubbed Apple’s American Manufacturing Program and will result in a tangible 450,000 jobs created in America across 79 factories.

(Image credit: C-Span)

Beyond the fact that all glass for the iPhone and Apple Watch will be made in the United States, Apple also hopes to create an end-to-end silicon supply chain in America. Apple already expects this supply chain to build over 19 billion chips by the end of 2025 here. Speaking at the White House, Cook said, “American innovation is central to everything we do," and it’s clear that the tech giant is further investing to ensure that will be the case going forward, especially from a building perspective.

Apple's decision to shift some component manufacturing to the US may have just saved it from a 100% tariff on chips and semiconductors that Trump announced during the press conference. Trump said, for companies like Apple, "if you're building in the US or have committed to building in the US, there will be no charge."

Apple has also started construction on a 250,000-square-foot facility in Houston, Texas, that’s focused on building advanced Apple servers, and is expanding a data center that supports services like Apple TV+ and Apple Music in Maiden, North Carolina.

You might also like

The Nintendo Switch's best games include Pokemon, Zelda, Mario and Metroid games, alongside indie hits and multiplayer party games.

The offer is open to new or existing customers.

• Android phones possibly under threat from worrying security threat
• Qualcomm releases fix for two major flaws in May and urged OEMs to apply it
• Google released a patch, so users should update now

Google has patched a major vulnerability affecting Android smartphones which is being actively exploited in the wild.

In June 2025, Qualcomm publicly announced discovering three vulnerabilities: CVE-2025-21479, CVE-2025-21480, CVE-2025-27038, saying they were “indications” from Google Threat Analysis Group (TAG) the flaws were being used in “limited, targeted exploitation.”

TAG specifically focuses on tracking state-sponsored threat actors, along with other highly sophisticated hacking groups, so if these were being used in limited and targeted exploitation, it’s safe to assume that these were nation-states targeting high-value individuals such as diplomats, journalists, dissidents, scientists, and similar.

CISA sounds the alarm

At the time, Qualcomm also urged OEMs (such as Google), to deploy the patch in their products without delay.

"Patches for the issues affecting the Adreno Graphics Processing Unit (GPU) driver have been made available to OEMs in May together with a strong recommendation to deploy the update on affected devices as soon as possible," Qualcomm said.

Google has now issued it August 2025 update for Android, which includes fixes for two of the flaws: CVE-2025-21479 and CVE-2025-27038.

The former is described as “memory corruption due to unauthorized command execution in GPU micronode while executing specific sequence of commands,” and was given a severity score of 8.6/10 (high). The latter is described as “memory corruption while rendering graphics using Adreno GPU drivers in Chrome,” with a severity score of 7.5/10 (high).

The US Cybersecurity and Infrastructure Security Agency (CISA) also added these two bugs to its Known Exploited Vulnerabilities (KEV) catalog on June 3, giving Federal Civilian Executive Branch (FCEB) organizations a three-week deadline to patch up, or stop using vulnerable software entirely.

Given Android’s decentralized structure, it is safe to assume that different devices (for example, Samsung’s Galaxy lineup, or OnePlus’ One lineup) will be getting these updates at different times. Pixel, being Google’s lineup of mobile phones, will most likely receive the updates first.

Via BleepingComputer

You might also like

• Major Android security update patches a host of actively exploited flaws, so download now
• Take a look at our guide to the best authenticator app
• We've rounded up the best password managers

• OpenAI's announced a livestream for August 7 at 1PM ET / 10AM PT / 6PM BST
• We're expecting GPT-5 to be unveiled as OpenAI wrote, "LIVE5TREAM"
• It comes after many rumors and hints about the next model

After countless rumors, teases, hints of a delay, and many, many thoughts from CEO Sam Altman, OpenAI has finally confirmed a livestream tomorrow, and we're expecting to see Chat GPT-5's formal unveiling.

It’s not just that we’ve been waiting for the next-generation model to arrive, but a post on X (formerly Twitter) from the @OpenAI account makes it pretty clear, as it reads, “LIVE5TREAM THURSDAY 10AM PT”. That’s a pretty clear spelling of ‘livestream’ replacing the ‘s’ with a 5, and hinting at the GPT-5 model.

As the next major model for OpenAI, GPT-5 is rummored to bring with it more speed and better efficiency, but a real spotlight might be on how we can interact with it. We’ve already seen more formal Agents debut from ChatGPT, but GPT-5 is likely going to bring in automatic selection of the right model.

LIVE5TREAM THURSDAY 10AM PTAugust 6, 2025

This means you won’t need to select the model you think is the best fit, as GPT-5 will understand your prompt and handle the specific routing for you. Hopefully, that means easier, more appropriate answers for various prompts. Just a few days ago, on August 3, 2025, Sam Altman shared a screenshot of ChatGPT with ChatGPT 5 as the selected model in the top corner.

With a planned livestream for tomorrow, August 7, 2025 at 1PM ET / 10AM PT / 6PM BST, this will turn out to be a pretty packed week for OpenAI. Yesterday, on August 5, 2025, OpenAI debuted two open-weight AI models, gpt‑oss‑120b and gpt‑oss‑20b. The latter of which is capable of running locally on a consumer PC.

GPT-5 would have a significantly more immediate impact, assuming it gets a wide rollout and could be in the hands of consumers soon after the livestream. Sam Altman did tease in a post on X on August 2, 2025, that OpenAI has “a ton of stuff to launch over the next couple of months--new models, products, features, and more” – so the August 7 livestream – err, LIVE5TREAM – could be the start of plenty of new features to try.

Of course, Altman also used that post to warn about capacity issues or ‘hiccups,’ so similar to other launches with longer lead times, it could be a bit of a wait before trying GPT-5 for yourself.

Either route, stick with TechRadar as we’ll be reporting live on whatever OpenAI announces during its livestream tomorrow, and we’d bet on GPT-5. Like previous OpenAI announcements, we're expecting the event to be livestreamed on the brand's YouTube channel here.

You might also like

• Guess the movie from just emojis – try the TechRadar AI Challenge!
• Sam Altman hints at ChatGPT-5 delays and posts about ‘capacity crunches’ ahead for all ChatGPT users
• The next generation of ChatGPT is just around the corner - here’s why GPT-5 could transform the way you use AI

AI hacks show how Gemini can be secretly used to control lights, heaters and much more, signaling a new evolution in digital vulnerabilities.

• Consolidating printer security under one vendor may save time, though not without integration risks
• Canon’s new security subscription arrives just as its printer drivers face critical vulnerability exposure
• Root access to Canon’s firewall allegedly sold online amid new enterprise protection service launch

The lines between traditional hardware providers and cybersecurity vendors are beginning to blur as printer brands enter the cybersecurity field, but hackers can still use your business printer as an easy backdoor into your corporate network.

Canon, long associated with cameras and office printing hardware, is now offering a tiered cybersecurity subscription aimed at protecting endpoint devices, documents, and data.

The offering includes two tiers: Enhanced and Premium - the former covers basics such as firmware updates and data backup, while the latter introduces proactive monitoring, threat detection, and rapid device recovery.

Canon security concerns

The launch follows closely on the heels of serious security concerns related to Canon's print infrastructure, including high-severity driver vulnerabilities and a possible network breach advertised on underground forums.

Just days before the new subscription service was announced, Microsoft’s offensive security team disclosed a critical vulnerability, CVE-2025-1268, affecting Canon’s printer drivers.

The flaw, which scores 9.4 on the CVSS scale, could enable attackers to halt printing or execute arbitrary code under certain conditions.

Canon issued advisories and urged users to update vulnerable drivers, particularly those tied to several production and office printer models.

While patching is essential, the persistence of such flaws highlights the broader risks that poorly secured print infrastructure can pose.

Adding to this unease, Canon has reportedly become the subject of underground listings offering root-level access to its internal firewall systems.

Though the company has not confirmed any such breach, security analysts continue to monitor claims circulating on dark web forums puporting to offer access allowing attackers to create backdoors or move laterally through the corporate network

Against this backdrop, Canon’s new Subscription Security Services may be seen as both a response to reputational risk and an attempt to reposition itself as more than a printer supplier.

Though these services resemble endpoint protection platform (EPP) features, they are focused solely on Canon’s device environment.

Whether this strategy gains traction depends on more than just Canon’s execution, as there is still a fair amount of skepticism around traditional hardware companies taking on roles typically reserved for antivirus and cybersecurity providers.

For businesses managing large fleets of print devices, consolidating protection through the hardware vendor may offer convenience, but it raises questions about scope, integration, and oversight.

If others in the hardware sector begin offering similar subscriptions, the market could see a gradual expansion of what constitutes EPP.

Via Cybersecurity News and Security Week

You might also like

• The printer market may have been cornered by a single brand, but a surprising number of people just don't have one
• These are the best small business printers to keep your SMB going
• And here are the best monitors for programming we've found

The tech company lost a massive privacy case involving the Flo app that has raised huge questions about how health apps are secretly being used.

Here are hints and answers for the NYT Strands puzzle for Aug. 7 No. 522.

Here are hints and the answers for the NYT Connections: Sports Edition puzzle for Aug. 7, No. 318.

Here are some hints and the answers for the NYT Connections puzzle for Aug. 7, #788.

Here are hints and the answer for today's Wordle No. 1,510 for Aug. 7.

• Google's latest Pixel software update fixes a few high-level security flaws
• It's rolling out now for eligible Pixel phones and tablets
• The update also fixes a glitch with the 'Back Button'

If you have a Google Pixel phone, make sure you’ve downloaded the latest security patch, as it includes several important updates that fix some potentially critical issues with your smartphone that you might not even be aware of.

There are a few high-level security flaws the patch solves, as well as one “critical” System vulnerability. According to Google, this flaw can be executed remotely (in combination with other bugs), and what’s more, it can be activated without any user interaction.

Yikes.

Google didn’t go into specifics about the hack beyond these details, but it doesn’t sound like one it would want to leave unaddressed.

(Image credit: Future)

Beyond security improvements, Google has also seemingly solved a Back Button bug, which saw it not work at times for users.

Here's a demo of the back button randomly not working on Android 16.I grabbed a bug report and submitted it to Google engineering along with this reproduction screencast. Hopefully, they'll figure it out. pic.twitter.com/nEmifqQRvbJune 14, 2025

As you can see in the video above, users would swipe back on their Android 16 Pixel phone and nothing would happen – which isn’t ideal if you want to exit out of an app or conveniently return to a different screen.

It might have taken close to two months, but after beta users got the fix in July, the back button glitch should now be solved on all devices running Android 16’s stable version on their phone.

Are you ready to update?

If you want to upgrade your Pixel device, the patch is rolling out now to all Pixel tablets and phones launched since the Pixel 6 and Pixel 6 Pro. Those two phones launched in 2021.

With automatic updates enabled, you might have already updated. However, to find it manually, you can head to your Settings app, then search for System Update and hit the Check for updates button to see if you’re up to date on your software.

If you have the August patch, then you’re all set, though it can take up to a week for updates to be made available to everyone – so if you are still on July’s update and see no option to install August’s, don’t worry, you’ll just have to wait a little longer for a fix.

You might also like

• 6 of the best Android phone accessories for photography, charging, tracking, and more
• I'm a die-hard iPhone fan, but switching to Android has shown me what Apple Intelligence is missing
• 'I think you see the future first on Android' – Google's Android leader Sameer Samat

• Shanling CD80 II is a small hi-fi CD player costing $359 (about £269 / AU$550)
• 3.5mm and 4.4mm balanced headphone outs delivering up to 850mW/32Ω
• Stream LDAC over Bluetooth from your phone

You've been told a million times about how wonderful vinyl is, but you hear a lot less about CDs – and that's a good thing, because the relative lack of trendiness means that the cost of good-condition CDs is often a fraction of what you'd pay for the same record on vinyl. If, like me, you like saving money as much as you like listening to music, then a CD player is still a smart addition to your system.

Chinese firm Shanling makes some impressive CD players, including ones with integrated amplification. And it's just brought out a new player called the CD80 II (via Darko Audio), with high-powered headphone amplification for wired over-ears and IEMs – and with high-quality Bluetooth streaming so you can play music from your phone, computer or tablet.

(Image credit: Shanling)Shanling CD80 II: key features and pricing

The Shanling CD80 II takes the compact CD80 and delivers a new DAC system and a much improved CD loader too. The 4th-generation CD loader comes from the more advanced CD-S100 model, with an HD450 laser and a familiar tray-style mechanism.

Inside the ESS DAC of the previous model has been replaced by a Cirrus Logic CS43198, and it's teamed with dual SGM8262 headphone amps to drive the 3.5mm and 4.4mm balanced outputs. They deliver 215mW and 850mW into 32 ohms respectively, with an in-ear monitor-friendly output impedance of less than 1 ohm.

Bluetooth input is Bluetooth 5.0 with support for LDAC as well as the familiar AAC and SBC, and there's also a USB input with support for 2TB drives to play files, at hi-res audio quality up to PCM 384kHz and DSD256.

There's a lot going on here, but Shanling has managed to pack it into a very small 28 x 20 x 5cm case, so it's small enough for even the tightest setups.

The new Shanling CD80 II has a US recommended price of $359; other pricing hasn't been announced but in the UK the first-gen model had a typical price of £339.

You might also like

• I didn't know I wanted a super-cool, wireless, audiophile CD player until I saw FiiO's
• I heard Rotel's ultra-luxe CD player and Hi-Res Audio DAC, and I'd buy it in a second if I could
• Technics SL-G700M2: a high-achieving network streamer that's also a high-end CD player

• KTC dual-mode 5K display delivers clarity or speed, depending on how you work
• Delta E below 2 and broad color coverage hint at pro-level image accuracy claims
• The 120Hz 2K mode offers smoother visuals for multitasking or low-friction gaming setups

The new KTC H27P3 monitor enters the market with a proposition which is hard to ignore: a 5K-resolution display priced at just $570.

The company is targeting professionals and general users who need a high-resolution panel without the premium price typically associated with 5K monitors.

The monitor is already available for preorder on KTC’s website, with shipping expected to begin in mid-August 2025.

Targeting creators with high specs at a modest price

The H27P3, which we first flagged back in April 2025, offers a 5120×2880 resolution IPS display @60Hz, designed to cater to creative tasks such as photo editing, graphic design, and color-critical work.

It also includes a 2560×1440 mode @120Hz, allowing users to switch between high clarity and smoother motion, depending on their needs.

KTC describes this as a “dual-mode” experience, combining visual precision with responsive performance.

The panel is factory-calibrated with a Delta E of less than 2 and supports 100% sRGB, 99% Adobe RGB, and 99% DCI-P3.

These are specifications that align with the expectations for a monitor for video editing or digital content creation.

HDR400 certification, a 500-nit brightness rating, and a 2000:1 contrast ratio suggest support for high dynamic range content, at least on paper.

(Image credit: IT Home)

In terms of design, the monitor features a thin-bezel frame and an aluminum stand offering tilt, swivel, pivot, and height adjustments.

At 3.75kg and with a compact form factor, it could also serve as a desktop-friendly portable monitor solution for professionals on the move or those working in tight spaces.

The stand is described as minimalist and sturdy, although its visual appeal and desk footprint may divide opinion.

Connectivity is broad, with DisplayPort 1.4, HDMI 2.0, USB-C with 65W charging, and two USB-A ports.

This gives the monitor flexibility to integrate with a wide range of setups, from desktop PCs and laptops to even gaming consoles.

A headphone jack is also included, and KTC ships the unit with the necessary cables and even a screwdriver.

Still, while the feature set is promising, it remains to be seen how the H27P3 performs in real-world scenarios - as specs like HDR400 and 8-bit+FRC can look good in marketing but often fall short in practice.

For now, the H27P3 stands out as an ambitious, budget-friendly business monitor that could appeal to a wide audience, provided it lives up to its claims once in use

You might also like

• We've rounded up the best monitors for photo editing on the market today
• And here are the best monitors for programming we've found
• Microsoft says it will spend billions this year as AI moves pay off

Feed posts and public reels can now be reshared on the popular social-networking app.

Jamie Lee Curtis and Lindsay Lohan are back for a freakishly fun follow-up.

• Report finds humans spot malware with an 88% accuracy rate
• Additional tools like Task Manager can help them identify it even more
• There are still some pretty common misconceptions surrounding certificates

A new report from researchers at the University of Guelph and the University of Waterloo has uncovered a slight improvement in human detection of potential cybersecurity threats, but has warned we're still missing too many signs.

The small study of 36 participants (split equally between basic, intermediate and advanced PC users) had them face six separate software samples, half of which included malware, with varying levels of assistance.

The participants already successfully scored an 88% malware detection accuracy when faced with the potential threats, but this improved even more to 94% with the use of an enhanced Task Manager interface, showing details like CPU usage, network activity and file access.

Humans aren't too bad at detecting malware

Despite relatively strong detection, the researchers observed three key misconceptions.

Users commonly misinterpreted the UAC shield icon as a sign of security while also demonstrating a lack of understanding of digital certificates. They also noted an overthrust in file names and interface aesthetics.

Users' detection techniques varied depending on their experience levels, with basic users relying heavily on superficial cues like icons, typos and aesthetics.

Intermediate users were able to improve their accuracy with additional system data, but advanced users often took a backwards step by over-analyzing threats, leading to false positives.

In this particular test, the researchers were able to identify 25 separate secondary indicators users use to determine whether something is a threat or not, on top of four primary indicators.

One of the paper's limitations mentions the fact that the participants knew they were looking to identify malware – unsuspecting victims downloading files from the web aren't often so lucky to have a heads-up.

Still, the research is especially valuable for developers, who can use the findings to tweak their software "to eradicate misconceptions and improve security related interfaces and notifications."

You might also like

• New malware avoids antivirus detection, unleashes a "plague" on your devices
• We've listed the best endpoint protection software for a cybersecurity boost
• Check out the best ransomware protection