
This BlackBerry Ripoff Is My Only Hope at Feeling Young Again
I love my phone. Guys, no… I love my phone so much. I hate that I love my phone, but I love it nonetheless. I love my phone so much that I spend all day typing stupid little words into its stupid little screen and waiting for stupid little responses. I love my phone, but it does not love me, especially when it comes to typing. I'm so bad at typing. If there were a contest to try and fail to type the least amount of words in the most amount of attempts, I'd win the triple crown platinum all-time iron chef trophy (on one leg with my eyes closed). That's a me problem for sure, but if you're old enough to remember a time before social media, it may be a you problem, too.
Don't worry, though; there's a solution to your God-awful typing, and it's been under your nose for a long time, technically. The solution: this blatant BlackBerry ripoff with a physical keyboard, Android 15, and 5G connectivity.
Introducing Titan 2 — the phone that can move forward and look back! 👀
As the latest 𝟓𝐆 𝐐𝐖𝐄𝐑𝐓𝐘 smartphone running 𝗔𝗻𝗱𝗿𝗼𝗶𝗱 𝟭𝟱, Titan 2 blends classic design with modern performance. 🚀 Its redesigned physical keyboard brings back the tactile satisfaction of real… pic.twitter.com/9H11WKMmuk
— Unihertz (@Unihertz) June 26, 2025Introducing the Titan 2, a new crowdfunded phone from Unihertz that (like its predecessor, the Titan 1) clearly takes its cues from BlackBerry. Most importantly, there's a full QWERTY keyboard with real buttons that you can press with your tired-ass thumbs to produce words that hopefully construct whole sentences. It's a story as old as time, but in a never-ending sea of touchscreens, somehow a breath of fresh air—or at least recycled, non-touchscreen air. If you're wondering how the hell you scroll on something like this, I'm also excited to relay that there's a scroll sensor built into the freaking keyboard. That means you can just swipe on the keys to ingest all the brain-rotting TikToks your internet-addled brain desires. I have my doubts about how well that feature works, but it's a nice flourish nonetheless. Also, don't worry, you can still use the display as a standard touchscreen if you so choose.
On top of all of that, there's also a screen on the back of this thing, which is bonkers. Unihertz, on its Kickstarter page, describes this feature as follows: 'Titan 2 features a 4.5-inch square primary display with a resolution of 1,440 × 1,440 pixels, alongside a secondary rear display for an even more imaginative dual-screen experience. Its flat-edge design adds a modern, sleek touch to the device.' Basically, it looks like you can see timers and notifications on the second, smaller screen, which is nice if you want to flip your phone upside down and give your equally tired eyes a bit of a rest but still keep a smaller, tired eye on the influx of Slack messages you desperately want to ignore. Camera-wise, there's nothing really to write home about—there's a 50-megapixel front-facing sensor and a 32-megapixel selfie camera—but that's probably not why you'd be interested in a phone like this anyway.
There's something really interesting about the fact that a phone like the Titan 2 or the Titan 1 can exist in this world where everything is a glass slab, and as a screen-addicted millennial, I think I understand the appeal. As much as technologies like touchscreens and autocorrect have become the lynchpin and launchpad to our smartphone experience, I think there's still something weirdly unintuitive about typing on a flat glass surface. My thumbs, no matter how much I try to train them, sometimes just don't want to cooperate, and autocorrect sometimes feels equally as unruly. There's a simplicity and tactility that I crave, and clearly others do too. As of writing this, the Titan 2 has $1,005,074 from backers already, and the fact that Unihertz was able to make a sequel to its 2019 Titan in the first place says a lot. It's going for around $271 and estimated delivery is currently October. Let's just be honest: I'll probably never feel young again, but hats off to Unihertz because a physical keyboard might at the very least make me feel less old.

Try Our AI Features
Explore what Daily8 AI can do for you:
Comments
No comments yet...
Related Articles


Motor 1
18 minutes ago
- Motor 1
We Tested Apple CarPlay Ultra. It's Great—But There's a Catch
"Does it have CarPlay?" It's one of the first questions anyone—driver, passenger, or buyer—asks about a new car. The frustrating menus, clunky user interfaces, and overall poor experience surrounding modern car infotainment systems have driven the majority of people to Apple CarPlay (or Android Auto, Google's native equivalent), which effectively mirrors your phone on the infotainment display. Since CarPlay's introduction in 2014, it's been almost exclusively for the center display. Now, though, with the arrival of CarPlay Ultra, the entire gauge cluster and dashboard turn into an Apple-mimicking iPhone projection. But of course, it's not without a bit of controversy. Photo by: Chris Rosales / Motor1 Fourteen automakers originally signed on for CarPlay Ultra when Apple announced it back in 2022. But some, like Audi, Ford, and Mercedes-Benz, have since backed away from the deal , while General Motors continues its assault on CarPlay entirely. Genesis, Hyundai, Kia, Porsche, and Aston Martin, though, still welcome the new technology. On the surface, it's a debate between the popularity of CarPlay and the automaker's ability to tightly control its user experience. Beneath the many consumer layers, automakers are worried about handing over user data to Apple—data that the automakers find extremely valuable. If they aren't selling your data to insurance companies, they're selling it to advertisers or aggregators looking for valuable details on consumer habits. Cluing Apple into that data isn't something automakers will do readily. In the case of Aston Martin, it recognized the popularity of CarPlay and made the decision "for [its] customers." As a small-volume luxury carmaker, Aston stands to lose if it doesn't get with the times. For much larger brands, this isn't as much of an issue. Nonetheless, I got to sample and stress-test CarPlay Ultra for a few hours in a 2025 DBX and lightly grill Aston about how Ultra integrates into existing systems. First and most importantly, CarPlay Ultra only works with an iPhone 12 or newer running iOS 18.5. Upon first connection, the phone uploads 50 MB of data to the car, basically a folder full of graphical assets specific to each vehicle. Ultra runs alongside the automaker's infotainment system as a sort of plug-in program, interfacing with the onboard systems for advanced driver assistance, air conditioning, radio, and drive modes so that most functions can be controlled through CarPlay. CarPlay Ultra does not deeply interface with the vehicle network. Instead, the onboard systems offer the relevant information to CarPlay and nothing more, at least according to Aston Martin. In nerd speak, CarPlay doesn't even communicate on the controller area network bus (CAN bus) of the DBX. Effectively, it's Apple graphics and UI over Aston's systems. It does run as a small operating system within the DBX, but it's not an entirely new infotainment system, and it still operates as a phone projection. Photo by: Chris Rosales / Motor1 Effectively, it's Apple graphics and UI over Aston's systems. Curiously, Aston did not have to pay a licensing fee either. All the automaker had to do was foot the development costs on integration, and Apple simply provided support. As Aston tells it, integration was relatively painless, with Apple listening to feedback from Aston and vice versa. But customers beware: As the old saying goes, 'If the product is free, you are the product.' Still, the user interface is classic Apple: Easy and intuitive. All major functions were easy to find under the new "Vehicle" menu, with specific push notifications while using physical controls. For example, clicking the ADAS off button on the center console prompted a CarPlay notification confirming the system's deactivation. Even drive mode selections were displayed when switching from Sport to Sport+, and specific drive mode settings were available deeper within CarPlay, allowing adjustments to things like the dampers, steering, and engine tuning, among others. The controversial piece, the gauge cluster, was also a highlight. It ran faster and looked slicker than Aston's native stuff, though that's more of an Aston issue than a CarPlay improvement. Photo by: Chris Rosales / Motor1 Photo by: Chris Rosales / Motor1 The ability to customize colors and backgrounds on a variety of layouts was lovely, and Aston's collaborative gauge cluster was crisp, with the Apple-specific stuff existing between the speedometer and tachometer. A full-screen Apple Maps projection also exists in the gauge cluster with turn-by-turn navigation. It's all undeniably Apple, but extremely nice to look at and use. There were a few caveats, though. Ultra is wireless only, which is difficult enough for standard CarPlay at times. Using Aston's provided iPhone 15, it ran smoothly for the two hours I tested it. Yet, using it with my personal iPhone 13 Pro Max running iOS 18.5 was excruciatingly laggy and slow. I attempted to use it for 15 minutes, but it never smoothed out, so I reverted back to the iPhone 15. That said, it's still in the early stages, so there are probably some bugs that need ironing out, and there's a new switching menu to smoothly choose between paired phones. At any rate, CarPlay Ultra is here—and it's very good. While the DBX does lose a little bit of Aston Martin charm, it still has a reasonable amount of brand identity beneath the Apple onslaught. But even Aston's new infotainment system, for as good as it is, benefits from legibility improvements and additional ease of use. That makes CarPlay Ultra an easy win. But what happens once Ultra goes up against Hyundai and Porsche's excellent systems? With automakers dropping out of Ultra and dragging their feet on integration, we will have to wait and see. More On Apple CarPlay Chevy's Infotainment System Is Excellent—Especially With CarPlay: Review GM Swears Ditching Apple CarPlay Was Still the Right Move Stop Trying to Reinvent the Infotainment System Apple CarPlay iOS 18: All the Updates You Should Care About Share this Story Facebook X LinkedIn Flipboard Reddit WhatsApp E-Mail Got a tip for us? Email: tips@ Join the conversation ( )
Yahoo
19 minutes ago
- Yahoo
China Mobile monetises one of the world's most advanced 5G networks
As one of the first telcos in the world to deploy 5G-Advanced networks, China Mobile Shanghai is blazing a trail in deciding where to upgrade and how to operate its new network to serve its VIP users and value-added service tiers. Verdict has previously laid out how China Mobile Shanghai designed its services to encourage multiple tiers of upgraded users. The story of how China Mobile rolled out its new network while monetising their network as well, is equally as useful to other telcos starting to plan for their own 5G-Advanced networks. Since 5G nomenclature can be confusing, a brief review: the majority of 5G networks today are 5G-Non-standalone (5G-NSA), which means that they run 5G New Radio base stations, but are controlled by a 4G/LTE core network. The non-standalone standard was designed to get the higher-capacity 5G radio network into the world quickly, but does not support 5G network slicing. 5G-NSA is therefore a kind of '4G 2.0' and has led to some of the popular dissatisfaction with 5G networks as 'more of the same' since it does not support any dramatically new services. 5G-Standalone (5G-SA) marries 5G New Radio with a 5G core network and can therefore support more finely differentiated quality of experience. Via network slicing, it can allocate network resources on a deterministic basis (meaning that the performance is guaranteed), not just best-effort. Roughly a third of global communications service providers (CSPs) are operating 5G-SA networks, but this 'mature 5G' level is already widely deployed in China. 5G-Advanced (5G-A) is a new mid-cycle upgrade analogous to 4G's LTE-Advanced. It introduces features that improve connection quality, reliability, and ease of management. It can support higher levels of service performance, but the big improvement is in the reliability of that performance. 5G-A is for power users that are willing to pay for guaranteed, tailored network experience: not only premium consumers, but IoT customers, emergency services, and so on. China Mobile selected the Shanghai region to work out 5G-A service design and marketing, but also to iron out technical issues regarding coverage and operations. China Mobile has established multiple tiers of 5G-A service plans: not only the top, 'platinum' tier, but a gold tier that receives network performance that is not as outstanding as the platinum group, but still better than subscribers to ordinary 5G plans. 5G-A's ability to provide precisely guaranteed network performance with not only a lower limit (faster than x) but an upper limit (no faster than y) allows for multiple such service tiers. Another benefit of this deterministic experience is that network resources can be released for others to use when not required by the premium tiers. Making the 5G-network a success requires two types of non-operational analytics: identifying users who are ready to upgrade to a better service tier, and prioritising new regions to upgrade to 5G-A by tracking the movements of existing 5G-A users and by finding new clusters of users who would be likely to upgrade. Shanghai is a dense urban area with many offices, but also many educational and scientific institutions. It is therefore rich in potential power users as well as consumers that can afford to pay more for better performance. China Mobile has already completed its 5G-A rollout to Shanghai's outer ring road, encompassing over 5000 sites. Now that it has covered the core business areas of the city, the company is expanding its 5G-A network by analysing user behaviour and demographics to determine the best areas for upgrades. The granular analysis looks for areas where high-ARPU users congregate as well as where network demand is heavy, for example central business districts and universities. It also tracks travel patterns for its existing 5G-A subscribers, which has led to it upgrading to 5G-A coverage along heavily trafficked roads and metro lines. China Mobile also tracks handset penetration in these areas. Newer 5G handsets are capable of a technology called Component Carrier Aggregation, which is abbreviated as '3CC' because it allows the handset to connect to three different carrier frequencies simultaneously. Once 3CC is enabled, users can achieve very high network speeds as well as the extra layer of reliability provided by simultaneous connections to multiple wireless cells. Handset penetration is therefore an important aspect of site selection for 5G-A upgrades. Operationally, it is a challenge to ensure that the network fulfills all of its service promises all of the time. This requires continuous real-time tracking and optimisation to ensure that users get the experience that they pay for; China Mobile's analytics platform tracks user experience on a per-application basis to make sure that the network is serving the different needs of gamers, streamers, TV watchers, and so on. China Mobile is currently using a Quality-of-Service flow control mechanism to provide this experience, but is planning to replace that mechanism with one using the NWDAF (the main analytics function of the 5G core network) and the AM-PCF (access and mobility management tied to the policy control function). It says that the NWDAF will support more granular customisation by user and/or application. Today, it uses all of 5G-A's network resources to optimise service delivery: 3CC aggregation, device-network synergy, differentiated resource scheduling, and pre-allocation of wireless network resources through network slicing. On a non-real-time basis, China Mobile also analyses the per-application experience of premium 5G-A users versus regular 5G users to make sure that the former retain a performance advantage. Similar analysis supports the marketing side by identifying users to target for upselling to a higher service tier or add-on package. China Mobile Shanghai's success in monetising 5G-A shows that proper network planning is essential, but that tracking and managing the experience of all groups of premium users is the real key to monetising the new network technology. "China Mobile monetises one of the world's most advanced 5G networks" was originally created and published by Verdict, a GlobalData owned brand. The information on this site has been included in good faith for general informational purposes only. It is not intended to amount to advice on which you should rely, and we give no representation, warranty or guarantee, whether express or implied as to its accuracy or completeness. You must obtain professional or specialist advice before taking, or refraining from, any action on the basis of the content on our site. Error in retrieving data Sign in to access your portfolio Error in retrieving data Error in retrieving data Error in retrieving data Error in retrieving data


Forbes
21 minutes ago
- Forbes
Your Boss Will Soon Grade You On AI Usage - Here's Why
Here's a number that should terrify every CEO: Daily AI usage has doubled in the past 12 months - from 4% to 8% of employees. Yes, you read that correctly. After billions in investment and endless hype, 92% of employees still don't use AI daily. At first glance, this seems to capture the spectacular failure of corporate AI adoption. Yet, this belies a fundamental truth that Microsoft' CEO Satya Nadella recently admitted: "The hardest part of AI isn't the tech - it's getting people to change how they work." "The hardest part of AI isn't the tech - it's getting people to change how they work." - Microsoft ... More CEO Satya Nadella And yes, companies need to push to change how they work in the AI era. According to the PwC 2025 Global AI Jobs Barometer, the report finds that since GenAI's proliferation in 2022, globally, productivity growth has nearly quadrupled in industries most exposed to AI (e.g. financial services, software publishing), rising from 7% from 2018-2022 to 27% from 2018-2024. In contrast, the rate of productivity growth in industries least exposed to AI (e.g. mining, hospitality) declined from 10% to 9% over the same period. It's not a question of 'if', but 'how' to drive AI adoption. The Adoption Mirage The headlines scream success. 78 percent of respondents say their organizations use AI in at least one business function, according to McKinsey's latest State of AI report. Companies are patting themselves on the back for "digital transformation." However, the success or failure of technology has always been at the mercy of adoption by people. This is a fundamental principle in business transformation that consulting firms have studied for decades and have created countless presentations and papers to help enterprises with organizational change. If we dig deeper into the data, a different story emerges. Despite two years of AI hype, only 8% of U.S. employees use AI daily - up from a tiny 4% just 12 months ago. Even 'frequent' usage (a few times a week) sits at just 19%. The vast majority (60%) still don't use AI regularly at all, according to Gallup's latest comprehensive workplace study. Think about that - we're celebrating going from terrible to slightly less terrible. Regular AI Use Growing Rapidly Among U.S. Workers How often do you use artificial intelligence in ... More your role — daily, a few times a week, a few times a month, a few times a year, once a year, less than once per year, or never? The Real Numbers Nobody Wants to Share In another recent study, KPMG surveyed 48,340 people across 47 countries to understand attitudes toward AI and its usage in the workplace. But wait, it gets worse. Despite the 58% usage rate, only 47% of the surveyed employees reported any form of training or education in AI. Translation: We're asking employees to adopt technology they don't understand, for purposes that aren't clear, with no training or support. The Mandate Approach: Use It or Lose (Your Good Review) According to Business Insider, Microsoft is making AI usage mandatory for performance evaluations. Julia Liuson, president of the Microsoft division responsible for developer tools such as AI coding service GitHub Copilot, recently sent managers stating "AI is now a fundamental part of how we work. Just like collaboration, data-driven thinking, and effective communication, using AI is no longer optional - it's core to every role and every level." To drive AI use across the company, managers must now evaluate employees based on their internal AI tool adoption. Microsoft is even considering formal AI usage metrics for future performance reviews. The Evangelist Approach: Train the Trainers While most companies struggle with single-digit daily usage, Moderna achieved near-100% voluntary AI adoption within six months of deploying OpenAI's ChatGPT Enterprise. The difference? They didn't mandate - they taught. Moderna set their objective high - achieve 100% adoption and proficiency of generative AI by all its people with access to digital solutions in the six months period. Brad Miller, Moderna's Chief Information Officer, was pragmatic about the task at hand. '90% of companies want to do GenAI, but only 10% of them are successful, and the reason they fail is because they haven't built the mechanisms of actually transforming the workforce to adopt new technology and new capabilities.' For this, Moderna assigned a team of dedicated experts to drive a bespoke transformation program. Their approach combined individual, collective and structural change management initiatives. 'We believe in collective intelligence when it comes to paradigm changes,' said Miller, 'it's everyone together, everyone with a voice and nobody left behind.' The pharmaceutical giant's results speak for themselves: Moderna has successfully driven AI adoption across the company in six months. The Wall Street Journal reports that Moderna went so far as to merge its technology and HR departments, creating 3,000 GPTs while restructuring roles with regulatory oversight. How did they do all of this? By solving real problems. Clinical teams built "Dose ID GPT" for trial analysis. Legal departments created contract review assistants. HR streamlined onboarding. Most importantly, Moderna didn't force adoption through performance reviews. They created tools so useful that employees couldn't imagine working without them. That's the difference between compliance and transformation. The Path Forward: What Actually Works Amid this push for AI adoption, some patterns of success emerge from the Gallup data: Clear Strategy Matters: When employees strongly agree that their leadership has communicated a clear plan for integrating AI, they are three times as likely to feel very prepared to work with AI and 2.6 times as likely to feel comfortable using AI in their role. Experience Changes Everything: Sixty-eight percent of employees who had firsthand experience using AI to interact with customers said it had a positive effect on customer interactions; only 13% of employees who had not used AI with customers believed it would have a positive effect. Leaders Use It More: Frequent AI use is also more common among leaders (defined as managers of managers), at 33%. They are twice as likely as individual contributors (16%) to say they use AI a few times a week or more. Based on successful implementations at Moderna and others, here's what measurably works: The Uncomfortable Truth After analyzing data from thousands of employees across multiple studies, the conclusion is inescapable: The AI revolution is failing because companies are missing the most crucial steps: leadership from the top and training for everyone. They're buying tools instead of building capabilities. They're mandating usage instead of demonstrating value. They're measuring adoption instead of impact. Most importantly, they're treating AI like software when it's actually a fundamental shift in how work gets done. In five years, asking if someone uses AI will be like asking if they use email. The companies that win won't be those with the strictest mandates. They'll be the ones where that 8% of daily users becomes 80% - not because they have to, but because they can't imagine working without it. The question isn't whether your employees will use AI. It's whether you'll lead that transition thoughtfully or force it desperately. The race is on. Where does your company stand?