Forget weather stations; this gadget lets you monitor the entire planet
Kickstarter illustrates one of the great things about crowdsourcing: The ability to garner interest in a new product from thousands of people around the world. Another part of crowdsourcing, however, is the possibility of harnessing that same crowd for large-scale collaborative projects. A new project from Kickstarter veterans Raspberry Shake fulfills both of these criteria. Called the Raspberry Boom, it’s a citizen science project which provides you with a finely tuned atmospheric monitor for your home.
Using it, you can detect infrasonic (read: extremely low frequency) sound waves, which are emitted by a huge number of natural and manmade activities. These infrasonic sounds can be caused by anything from incoming harsh weather or volcanoes erupting to planes flying overhead and potential nuclear testing.
“The Raspberry Boom brings this technology into your home at an affordable price,” Michael Hotchkiss, marketing manager on the project, told Digital Trends. “The complex circuit board and advanced sensor collect and process atmospheric infrasound readings, while the unit itself is powered by the most popular single board computer, the Raspberry Pi. As soon as you plug in your device, it connects with our Station View network and you can start monitoring your local region for activity.”
The idea is, essentially, to create a device similar to the many home weather stations around the world, but one which can extend far beyond simply climate tracking. Each device incorporates a powerful 24-bit digitizer, which samples infrasound at 100 samples per second with data transmission rates of four packets per second. It connects automatically to the team’s Earth-monitoring network, thereby allowing you to see atmospheric readings from other users all over the world. Whether it’s a volcanic eruption in Hawaii or tornado season in Florida, as long as there is a Raspberry Boom in the area, you will get to find out about it.
“Our previous campaign for our home Earth monitor, the Raspberry Shake, appealed to hobbyists, makers, creators, citizen scientists and Raspberry Pi enthusiasts,” Hotchkiss continued. “Being the most advanced and first ever atmospheric monitor of its kind, designed for home use, the Raspberry Boom is likely to attract attention from a wider audience — including early adopters who already have some kind of home monitor, to weather enthusiasts, such as fans of the popular Netatmo devices.”
While all our usual warnings about crowdfunding campaigns apply, if you want to get involved with Raspberry Boom you can pledge money over on their Kickstarter page. A complete kit will set $459, although cheaper options are available minus the Raspberry Pi board.
Editors’ Recommendations
- These Raspberry Pi 3 bundles will cover everyone, from coders to gamers
- Take a bite of the new Raspberry Pi 3 Model B+ on Pi Day
- Raspad is a portable Raspberry Pi tablet for bringing creative projects to life
- You can build your own smart home hub with Mozilla’s Project Things
- Tom Cruise honored, Michael Bay snubbed at 38th Annual Razzie Awards
This cyberpunk spartan helmet is actually a portable brainwave scanner
It may look like some crazy mashup of The Phantom of the Opera and the weird John Travolta alien from Battlefield Earth, but this sci-fi-style piece of futuristic headgear is actually a wearable portable brain scanner that is able to record neural activity while you’re on the go.
Unlike the large stationary functional magnetic resonance imaging (fMRI) scanners which are often utilized for brain imaging, this magnetoencephalography (MEG) system could open up whole new ways of studying brain activity, without requiring subjects to remain entirely motionless while this is going on.
“MEG works by measuring the weak magnetic fields that are produced outside the skull by electrical activity in the brain,” Richard Bowtell, a professor of physics at the U.K.’s University of Nottingham, told Digital Trends. “By mapping the spatial variation of these magnetic fields over a portion of the head surface, we can work out whereabouts in the brain the current is flowing. To make the first wearable MEG system, we used a new kind of quantum sensor that is very sensitive to magnetic fields. These optically pumped magnetometers have been used before for measuring brain activity, but we exploited new commercially available sensors which have been miniaturized and made robust by QuSpin Inc., which were mounted in a 3D-printed head cast.”
Bowtell said that the head cast was 3D printed to fit snugly over the subject’s head, and includes slots which can be used to site the sensors that are positioned over the wearer’s motor cortex. “It was printed by Chalk Studios in London and they made it look very cool,” he continued.
Aside from its striking appearance, Bowtell said that the wearable opens up exciting new opportunities for experiments in which subjects move their heads, such as mapping neural activity in a person who is bouncing a ball on a bat. Right now the limitation is that users are stuck moving inside a shielded box enclosure, which prohibits walking around. However, this could be changed in future iterations of the wearable, which will allow greater levels of natural movement.
A paper describing the work was recently published in the journal Nature.
Editors’ Recommendations
- The weirdest, wildest, and most wonderful wearables in the world
- Awesome Tech You Can’t Buy Yet: Magnetic bike pedals, undersea scooters, and more
- ‘Mind-reading’ A.I. produces a description of what you’re thinking about
- This A.I. literally reads your mind to re-create images of the faces you see
- Misfit Vapor Review
Wind turbines have a bird murdering problem, but Oregon might have a solution
We’re big fans or renewable energy sources, but that doesn’t mean that they don’t come without their own (sometimes unforeseen) challenges. One such challenge? The issue of birds, including federally protected avians such as bald eagles and golden eagles, flying into wind turbines. While bird deaths by wind turbine are comparatively rare next to other causes of fatality, it’s still a problem that needs to be looked into — particularly when it comes to endangered species.
While there is no foolproof solution right now, researchers at Oregon State University want to help. With that in mind, they created a system designed to ensure that wildlife can coexist with wind power generation as harmoniously as possible. Not only does their tech help gauge just how big a problem this is, but it can also take steps to help deter birds (with a specific focus on eagles) from flying into the blades.
“The system is characterized by three functions,” Roberto Albertani, a researcher on the project, told Digital Trends. “[Firstly, an] eagle detection system to alert for birds flying dangerously close to the wind turbine rotor. [Secondly,] the implementation and operation of a simple yet effective ground-based eagle deterrent to safely scare eagles away from the wind turbine, and, thirdly, a blade collision detection and species recognition system.”
The smart impact detection system works by employing vibration sensors mounted to turbine blades, a video camera, and acoustic sensors. This allows the researchers to track which birds are in the area, as well as activate a ground-based deterrent in the form of randomly moving, colored facsimiles of people, which can help scare off the birds before they meet an untimely demise. Think of it like the airport bird deterrent system which mimics birds of prey to scare off regular smaller birds — only this time the goal is to scare off the larger birds themselves.
The researchers have so far put their system through its paces by firing tennis balls at turbines to show that the system is able to respond quickly to fast-moving objects. However, testing is set to continue.
“The eagle detection has been partially tested using trained eagles and showed good potential,” Albertani said. “Birds deterrent has been fully tested and blade impact detection will be tested with dummy targets.” The completed system will be tested in real operating conditions at the National Wind Technology Center in Boulder, Colorado, between summer and fall 2018.
Editors’ Recommendations
- CES 2018 will have an extra focus on smart cities and the impact of IoT
- New disease-finding electronic nose sniffs poop so scientists don’t have to
- These performance cars prove you don’t have to spend much for speed
- Soon, you won’t have to be a Windows Insider to test Microsoft’s newest apps
- Sony’s Xperia XZ2 smartphones have better Super Slow Motion than the Galaxy S9
Replaced by robots: 8 jobs that could be hit hard by the A.I. revolution
There’s no doubt that artificial intelligence (A.I.) and other cutting edge technologies are going to change the face of employment as we know it. According to one famous study, 47 percent of currently existing jobs in America are at high risk of potential automation in the coming decades.
What are some of the professions that will suffer the effects of the A.I. revolution? And is there anything people can hope to do about it? Read on to find out.
Lawyers
Why they’re screwed: Judging from the number of movies and TV shows about the profession, being a lawyer is a pretty great job: an interesting, high-earning career with bags of social status attached. However, hiring a lawyer is also expensive and a substantial portion of what lawyers do on a daily basis turns out to be a lot more routinized than some in the profession would have you believe.
While genuinely bespoke legal work still requires humans, A.I. can help perform tasks ranging from legal discovery (the pre-trial process in which lawyers decide which documents are relevant to a case) to creating contracts. They can even argue parking fines and handle divorce proceedings.
So there’s no hope? Junior lawyer jobs may be harder to come by than ever, but studying a combination of law and computer science could be extremely fulfilling. Whether it’s advising on how best to turn laws into algorithms or investigating the legal framework around new technologies like self-driving cars, there are plenty of interesting opportunities available.
Journalists
Why they’re screwed: An algorithm could never write a listicle as compelling as this one, right? Guess again! Whether it’s using bots to generate sports reports and other news articles or attempts to use A.I. for more in-depth investigative journalism, there’s plenty to suggest that journalism isn’t safe from the clutches of artificial intelligence.
Heck, combine advances in computational creativity and text generation with the sorry financial state of many media companies and the results don’t add up to anything approaching optimistic for wordsmiths.
So there’s no hope? A.I. doesn’t have to be your enemy. As it turns out, bots could be the hired researcher human journalists always dreamed of, with the ability to pull up statistics and unearth interesting patterns in data which can lead to entirely new ways of telling and presenting stories.
In the future, there’s also the possibility that A.I. automated agents could be used to help personalize human-written stories for readers, based on their knowledge, location, age, or reading level. Doing so could expose human writers’ work to even larger audiences.
Drivers
Why they’re screwed: In a chapter from their 2004 book, The New Division of Labor, MIT and Harvard economists Frank Levy and Richard Murnane argued that a computer would never be able to drive a car, due to the enormous complexity of information involved with this task.
Today, we know that is categorically false, due to the thousands of miles successfully driven by self-driving cars. Fleets of autonomous vehicles owned by companies like Uber will have an enormous impact on professional human taxi drivers, while autonomous trucks will mean the same thing for long-distance drivers.
Things don’t look too hot for driving instructors either. After all, will kids born in 2018 even need to pass a driving test?
So there’s no hope? Based on the response to Uber in some major cities like London, human cab drivers may be able to resist the threat of self-driving cars for a short time. Sad as it is to say, however, human drivers should probably think about reskilling.
Chefs
Why they’re screwed: Considering that they need cooling fans at the best of times, working in a hot kitchen sounds like a terrible idea for a computer. But A.I. is always ready to surprise us. One example of a chef robot was created using IBM’s Watson technology. Called Chef Watson, it’s able to generate entirely new recipes from scratch using an astonishing knowledge of taste chemistry and flavor pairings.
Meanwhile, robots like Miso Robotics’ burger-preparing Flippy are capable of preparing meals and serving them up at speeds that human chefs struggle to achieve. Add table delivery drones into the mix and you don’t even need human waiters to deliver the food to customers.
So there’s no hope? As with a lot of areas, humans who are ready to take advantage of the technology stand to benefit. If you’re a chef, you could conceivably use robots to churn out identical dishes to your specifications in greater quantities than you yourself could cook.
Using technologies like 3D food printing will also make possible the creation of entirely new dishes that would have been previously unimaginable.
Financial analysts
Why they’re screwed: Like being a lawyer, working in the finance sector has traditionally been a high status, high income job. However, increasingly A.I. is taking over. Computers can spot patterns and make trades faster than even the most eagle-eyed of human analysts.
With billions of dollars (or more) at stake, it’s no wonder that machine learning tools are all the rage, while some estimates suggest that around 30 percent of banking sector jobs will be lost to A.I. within the next decade.
So there’s no hope? There will be fewer jobs, but there are certainly opportunities. So-called “quants” who are able to combine knowledge of the financial sector with computer science and math are highly sought after to help develop the algorithms which increasingly drive this field.
Elsewhere, the importance of “relationship banking” to help build up customer loyalty and provide personalized service will continue to grow.
Telemarketers and customer service assistants
Why they’re screwed: Chatbots are getting way smarter, as tools like Google Home and Amazon’s Alexa show us. That doesn’t bode well for a lot of telemarketers and phone-based customer service assistants, who are often speaking according to a script.
With miserable conversion rates for direct telephone sales and fewer people willing to wait customer service assistants to become available, smart chatbots can perform a lot of these tools admirably. That’s not good for people working in this field, who have already been hit by the outsourcing of many of these jobs to people in other countries like India.
So there’s no hope? Not necessarily. One A.I. company, Mattersight, uses voice recognition technology to figure out the personality type of customer service line callers and patch them through to humans with a similar personality type. Doing so can dramatically shorten the length of calls, while increasing the problem resolution rate.
Meanwhile, companies like Twitter employ human “judges” who can help machines make sense of the information they’re being shown — such as figuring out the context of a specific trending topic. In other words, these jobs may still exist, but you could be answering an A.I.’s questions instead of a human’s.
Medics
Why they’re screwed: Whether it’s algorithms which can make diagnoses about disease, computers being used to make recommendations about the best cancer treatment, A.I. pharmacists, wearable devices that can help treat physical disorders, or even robots carrying out surgery, there’s no doubt that cutting edge technology will have a big impact on a range of medical professions.
Due to their brilliance at capturing and analyzing data, artificial intelligence promises to be a major disruptor in this space, giving everyone an A.I. physician in their pocket.
So there’s no hope? A lot depends on the area of medicine you work in. For the most part, though, humans will still be in the loop. Diagnoses and even surgery could one day be managed by machine, but in the immediate future technology will augment human physicians and healthcare workers, not replace them.
Construction workers or other manual labor jobs
Why they’re screwed: Whether it’s bricklaying on construction sites, working in warehouses, or picking fruit and vegetables on a farm, there’s no doubt that a large number of manual labor jobs that once required humans can now be carried out by robots.
The advantage of these robots is their ability to work nonstop without getting tired. That’s something that’s not possibly in any line of work, but especially not in physically taxing manual labor roles.
So there’s no hope? All of the jobs we’ve mentioned are going to get more automated, but right now humans still have an advantage over robots when it comes to dexterity. For example, Amazon’s warehouses use robots developed by Kiva Systems to move around racks of shelves and bring them to stationary human workers who then pick the required items off the shelf.
Similarly, bricklaying robots can lay bricks, but require humans to do the grouting. This balance will shift as robots get more dexterous, but for now many humans will find themselves working alongside robots rather than being flat-out replaced by them.
Editors’ Recommendations
- Science fiction’s 5 most haunting A.I. villains, ranked
- Don’t be fooled by dystopian sci-fi stories: A.I. is becoming a force for good
- A.I. perfectly predicted last year’s Super Bowl score. What happens to betting?
- Truly creative A.I. is just around the corner. Here’s why that’s a big deal
- Rise of the machines: Here are the best robots we saw at CES 2018
Westworld is becoming a mobile game, and we got an early look!

You may not be able to be a guest in Westworld, but you can be something better: an employee!
I would have a difficult time finding someone that hasn’t at least heard of the HBO series Westworld, which takes place in a dystopian universe where a giant amusement park filled with truly humanistic androids are the play things of the truly inhuman humans. I don’t know about you, but before the first episode ended, I wanted to go to there.
WBGames has made this possible, at least in a digital capacity, with the Westworld mobile game. To make the simulation as realistic as possible, the game is meant to be thought of as an employee manual for a new recruit — The Delos Park Training Simulator. You learn everything there is to know about the park in the laboratory underground, while making sure the action never stops above ground in Sweetwater, Pariah, and the like.
Westworld is scheduled for launch this April on iPhone and Android devices. At GDC 2018, I was able to get a first look at the upcoming simulation game.
Everything in this world is magic … except the magician

The premise of the game is that you, as a new employee at Westworld, are given this training simulator so you can learn the ropes of your new job. Your in-game friends are considered your coworkers. Your in-game tasks are the corporate objectives. It’s pretty detailed.
The game takes place in two sections, the above ground, where you’re helping guests by matching them with hosts, and the underground, where the Delos lab is situated and you do such things as manufacture new hosts, repair and rebuild existing hosts, and build new buildings or unlock new locations.
Since I was a child, I’ve always loved a good story

Sweetwater (where you start) is where the stakes are the lowest and the tasks are the easiest and the hosts are capable of matching with guests at the lowest level.
As you move further from Sweetwater and unlock the broad spans of the Westworld landscape, you’ll come across guests that ask for more difficult situations. Hosts have to be capable of higher-level actions.
Guests and hosts have pre-assigned affiliations of black, white, or neutral. Just like when William is offered a black or white hat when he first enters Westworld, each host and guest aligns with a dark or light side, with the third option of neutral — neither good, nor bad.
If you try to pair a guest requesting a high-level situation with a low-level host or a black hat player with a white hat android, the results could end in violence or death — temporary death, at least. If the interaction between guest and host doesn’t go well, the guest may decide to end the conversation the fastest way possible, with a bullet.
No problem! That’s what you’re here for. If a host dies, it goes to “the body shop” where Felix and Sylvester does repairs. It is at this time that you can decide to go through a rebuild cycle. This can either give the host a complete personality reassignment or boost up a host to a higher level with more roles and stronger attributes. How do you make these rebuild alterations? By destroying other hosts, of course.
You’ll be manufacturing and adding new hosts to the park on a regular basis. What you manufacture is a bit of a random chance. Some of the hosts are very common, like Emmett. It’s highly likely that you’ll make dozens of Emmetts throughout the game. You can level up your Emmetts or use them as bio material to help level up other hosts.
Players must also build buildings around Westworld in different towns, like Las Mudas. The streets are bare when you unlock a town. Using materials you earn from completing tasks for guests, you can build additional buildings, which will bring in more and varied guests. You can also upgrade buildings, which will make them more enticing to more complicated guests, which will in turn, earn you greater rewards.
There are a few additional behind-the-scenes game features, like Critical Analysis. This is where you can earn Host Codes, which are the tokens needed to manufacturer new hosts. Three times per day, you can perform a host’s critical analysis where you select five randomly generated questions to answer from the host’s perspective. The goal is to keep the host as well-balanced as possible by selecting answers that reward the best average of increase or decrease in the different calibrations. If it sounds a bit confusing, you’re not alone.
You can’t play God without being acquainted with the devil

Westworld will feature regular events in the game. I didn’t get to participate in an event, but Jason Everett from WBGames San Francisco explained that they happen regularly and offer different rewards, depending on what’s happening. You might have the opportunity to earn rare materials or be able to access buildings that are normally only open on the weekends.
For the right type of player, there will definitely be plenty of reasons to play the game regularly and events will pull you back in if things start to seem too repetitive.
It doesn’t look like anything to me

The game is free to download and play with the ability to purchase in-game premium currency. One thing I’ve noticed in the past few years is that game developers have figured out a good balance to getting and keeping players instead of turning them off with paywalls. Everett explains Westworld as completely playable for all time without needing to spend a dime. You may want to use premium currency to speed up a task, but nothing is ever locked behind a paywall or otherwise frustratingly unplayable.
Since the game doesn’t launch until sometime in April, I can’t say for sure how well WBGames will be at making a totally fun experience for players who don’t ever want to spend money, while making a few dollars off of those players who don’t mind opening their wallets. The proof is in the pudding.
They say that great beasts once roamed this world

Presumably because Westworld is an incredibly gritty and violent world, the game’s designers went with a slightly more cartoonish look. The action moves in 2D when you’re up close and viewing an encounter. Characters have a stylized look.
Visually, it strikes a good balance between the lighthearted fun of a simulation game and the darker side of Westworld. Hosts can and will die for the simplest of reasons, like the guest just didn’t like them. Manufacturing simulated human life is as trivial as pressing a button and destroying that life is as banal as making a simple decision.
At the same time, though, they all have pleasant looks on their faces and are illustrated as flat, two-dimensional characters. I feel like there’s a deeper discussion to be had here.
This world doesn’t belong to them, it belongs to us

There’s a lot going on here, a lot more than simply making your visitors happy. You’ll also have to keep your hosts from dying or glitching out and doing something that would warrant a role reassignment. You have to build and develop the world that guests want to visit while maintaining the complex technical side of androids that can think for themselves (sort of).
It’s so complex that it’s somewhat overwhelming at first glance. You’re in charge of managing both what happens on the surface above ground and in the lab below ground. It’s not a simple game that players can drop in on for a few minutes while waiting for the bus to arrive. As any good simulation game will do, you’re going to want to keep track of what’s happening in the game as often as possible.
The entire game is a bit of a balancing act. Should you level up all 10 of the Emmets that you have? Or do you sacrifice seven of them while leveling up only three of them to satisfy the white, black, or neutral hat guests? If you decide to sacrifice, do you level any of those up so they become more useful material for a rebuild? Is that worth your time?
My limited experience with the Westworld mobile game wasn’t enough for me to complete my “training.” If this were a real-life first day on the job, I’d barely learned the importance of washing my hands before handling food. I can’t wait to really dig in and see how to game plays out over the course of days, weeks, and months.
Keep an eye out for more of our Westworld coverage as the official launch date approaches.
Samsung is ‘looking into’ Galaxy S9 touchscreen issues
Tonight Samsung has responded to growing reports of Galaxy S9/S9+ phones with touchscreen problems. Android Police points to a few threads on Reddit and Samsung’s own support forums where owners have noted issues with their phones registering touches. Where on the device they see this issue has varied, and some said they were able to resolve it with a factory reset or turning up the sensitivity but most needed a new phone. A Samsung spokesperson said to Engadget that “We are looking into a limited number of reports of Galaxy S9/S9+ touchscreen responsiveness issues,” and encouraged any owners with problems to contact the company directly.
Of course, we’ve seen worse hardware issues, and iPhones have had their own bouts with things like “touch disease” but we’ll be watching carefully to see if this issue is truly widespread or limited to just a few devices.
At Samsung, customer satisfaction is core to our business and we aim to deliver the best possible experience. We are looking into a limited number of reports of Galaxy S9/S9+ touchscreen responsiveness issues. We are working with affected customers and investigating.
We encourage any customer with questions to contact us directly at 1-800-SAMSUNG.
via bobdurfob
Source: Android Police, 9to5Google, Samsung support forum
Hands-on with the sci-fi game that falls apart as you play
You’re never going to play Clunker Junker in your living room or on your desktop PC, no matter how many GPUs it has. Hardware is the issue here, but it’s not a matter of processing power — Clunker Junker requires two LED-adorned arm cranks, plus four crates with glowing lights and doors that come crashing down when they’re unlatched during gameplay. The game itself, to be fair, runs on a laptop, but that’s about as traditional as this thing gets.
Clunker Junker is a cooperative space shooter for two players. Each crew member has a handheld, gun-like controller that stretches from palm to elbow; it’s made to be held in the left hand, with a crank on the right side, a navigation cube on the base, and an LED spout on the front. The spout fits into openings in the four crates, representing different aspects of the spaceship — movement, shields, weapons and engine.

Players stick their navigation cubes into squares cut in the top of two of the boxes; one person controls where the ship goes while the other manages the direction of the turrets. When the on-screen ship takes damage (and, trust me, all Clunker Junker players will take damage at some point), flaps on the front of the affected boxes slam down, their innards glowing red and demanding repairs. Fix up the ship by sticking the LED gun into the appropriate hole and turning the crank until the box glows green again. Then close the flap and get back to steering the ship or shooting with the turret — or fixing whatever else has broken in the interim.
Clunker Junker is a fast-paced, communication-driven video game that doubles as a solid arm workout. Its creators at experimental studio HNRY know how to build physical experiences with real impact: Last year, they took home the alt.ctrl.GDC award at the Independent Games Festival for their Fear Sphere project. Not bad for a group of folks making weird games in their spare time, just for fun.
That’s the heart of the alt.ctrl.GDC space — strange experiences built to push the boundaries of video game development, rather than turn a profit. Clunker Junker may not be heading to the Xbox One or PlayStation 4 any time soon, but the ideas powering its development may, one day, seep into traditional development pipelines.
Click here to catch up on the latest news from GDC 2018!
Researchers develop device that extracts water from desert air
Researchers at MIT and UC Berkeley have developed and now tested a device that can extract water out of the air even in the driest of climates. The team proposed the device in a Science article last year and now they’ve improved the design and tried it out in Tempe, Arizona. While there are a few ways to pull water out of the air, most come with significant limitations. They usually require humidities upwards of 50 percent and some need a lot of energy input to make them work. The research team’s latest design, however, works passively, without the need for energy input, and can work in places with humidity as low as 10 percent.
The core of the design is a material called a metal-organic framework, or MOF. It’s made up of linked molecules that create a super porous material with lots of surface area. And depending on what you make the MOF with, it can be very hydrophilic, meaning it attracts water. See the video below for more on this type of material. The MOF used here is able to pull water from the air and store it in its pores, which it does during the night. Then come daytime, sunlight is used to release the water and a condenser is used to harvest it. The system is completely passive and the research team showed it can work even in the arid Arizona climate.
Tested over five daily cycles in May 2017, the water output was estimated to be around a quarter of a liter per kilogram of MOF. Additionally, the team tested the water for impurities and found that the MOF wasn’t leaching into it. Next, the researchers want to scale up the system and make it more efficient. “We hope to have a system that’s able to produce liters of water,” Evelyn Wang, who lead the research, said in a statement.
The work was published today in Nature Communications.
Via: MIT
Source: Nature Communications
Case Maker’s Rendering Depicts iPad Pro With No Home Button, Slimmer Bezels [Update: Fake]
Noted leaker Evan Blass this evening tweeted a mysterious render of an iPad that features slimmer bezels, no Home button, and no notch with space for a camera.
Blass did not provide any information on the render, aside from the cryptic message “Seems to be something missing from this (encased) iPad…” which perhaps hints that he received the render from a case maker. Blass typically only shares leaks that he believes are from credible sources.
Seems to be something missing from this (encased) iPad… pic.twitter.com/n0esVAOBkq
— Evan Blass (@evleaks) March 23, 2018
Case manufacturers often create renders and dummy models of devices ahead of their debut in order to be the first to come out with new accessories and cases for freshly launched devices. Much of the time, these kinds of leaks are accurate because there’s a lot of money at stake, but there can be serious misses.
In this case, with no accompanying information, there’s no way to determine whether or not this design is indicative of what a future iPad might look like. There is no notch or cutout for the camera, and the device is depicted running an older version of iOS, but for a simple rendering to show off a case, these features wouldn’t matter. The lack of a front-facing camera and no visible notch in this particular render, if accurate, are not indicative of Apple’s plans.
Rumors have indeed suggested Apple is working on an iPad Pro that’s similar in design to what’s pictured in the rendering. Apple is said to be planning to launch an iPad Pro with thinner bezels, a faster processor, a custom Apple-built GPU, and most importantly, Face ID.
With Face ID, the next-generation iPad Pro will not need a Home button, providing more room for the display. Future iPad Pro models are expected to come in physical sizes similar to the existing 10.5 and 12.9-inch iPad Pro models, and based on icon spacing, the rendering appears to depict the larger iPad Pro.
Exact display sizing for future iPad Pro models is unknown as of yet, but a recent rumor from Taiwan’s Economic Daily News has suggested the smaller of the two tablets could feature an 11-inch display. Apple would not need to increase the actual size of the iPad Pro to introduce a larger display if the bezels are smaller.
Apple is believed to be planning to introduce new iPad Pro models sometime in the second half of the year. Bloomberg has suggested the devices will debut sometime after June, perhaps in September, while other sources have predicted a June WWDC launch.
Update: It appears the “rendering” is a fake photoshopped image, as it spotted in an auction listing for a 10.5-inch iPad Pro case along with the original image that was photoshopped. Evan Blass has deleted his original tweet.
Gimme an L! Gimme a O! Gimme a L! 😂 #Photoshopped #LOLeak pic.twitter.com/9MAtknSU9M
— Steve H. (@OnLeaks) March 23, 2018
Good find. I need to have a talk with someone. https://t.co/c41dHZdjEn
— Evan Blass (@evleaks) March 23, 2018
Discuss this article in our forums
Tesla says it’s being underpaid because its batteries are too fast
It looks like Tesla’s batteries are too fast for their own good. As the Sydney Morning Herald reports, the company claims it’s not being properly paid for the electricity its South Australia battery farm is generating for the country’s power grid. And Tesla says its because its batteries supply electricity faster than the Australian Energy Market Operator (AEMO) can register. The current standards are based on response rates for fossil fuel generators, but Tesla’s batteries can respond much more quickly than they do. “Tesla estimates that the Hornsdale Power Reserve battery has delivered 30 to 40 percent of its services to frequency markets without being paid due to existing AEMO technical specifications being written based on fossil fuel generation assets,” Tesla said.
The South Australian battery farm went live last December and it has already proven its worth. In January, the AEMO asked energy companies to provide power to the grid, which it does when there’s a system outage or planned maintenance. That typically drives up energy prices and during this particular incident, prices were expected to reach around AU$9,000/MW. But Tesla’s batteries and an adjoining wind farm kicked in and kept prices around AU$270/MW, saving the energy market millions of dollars.
In the US, Tesla’s batteries are used in California, Hawaii, Vermont, Massachusetts and Puerto Rico. They’re also used to run the island of Ta’u and are being incorporated into a Canadian pilot project that will test a hybrid of battery and wind power.
Via: Electrek



