The predictive powers of AI could make human forecasters obsolete


Thousands of years ago, oracles read the future through divine inspiration. Today, we’ve still got Oracle making predictions (along with many other forward-thinking tech firms), but it uses something a little more grounded. Artificial intelligence and its capacity to assess approaching events are pretty awe-inspiring even without the supernatural flair.

Many industries are looking to artificially intelligent software to help make predictions on everything from a customer’s buying decisions to which medical treatments will be most effective for a sick patient. Though we live in a world that still depends on the educated guesses of experts, it is becoming increasingly clear that next generation of prognosticators will be more silicon-based than carbon-based.

AI is a prediction technology at its very essence. With the ability to evaluate data exponentially faster than any person, machine learning programs can assess patterns, make connections, and test hypotheses in less time than it takes their human equivalent to pour a cup of coffee. Thanks to its advanced capabilities, AI’s predictions are already taking shape, with strong implications for retail, health care, and the way we understand the world around us.

Retail

Retailers are going all-in on predictive software, with an eye on better serving their customers’ needs even before they manifest themselves. While the oft-repeated story about Target’s marketing software revealing a teenage girl’s pregnancy based on her buying patterns turned out to be a myth, that kind of foresight is not completely out of the realm of possibility for retailers, given enormous leaps in machine learning capabilities.

In an IBM survey, 91 percent of retail executives said that AI is set to disrupt their industry. There’s no reason to believe they’re wrong, either. In a field that rewards efficiency and ingenuity, smart computing offers both in a rapidly improving package.

Amazon trademarked predictive stocking all the way back in 2014 (an eon in software development time). We haven’t yet seen it come to life, but Amazon technology like Dash buttons is part of the data-collecting process that will one day make it a reality. Today, you press a button to tell Amazon when you need detergent, razors, or cat food. Tomorrow, the data gleaned from those button presses will get those goods to you before you even realize you need them. Creepy? Maybe. But our AI-inflected future will be so full of this kind of interaction that those uneasy feelings are unlikely to last.

Health care

While the consumer front is undergoing its own AI revolution, there’s a parallel one happening in the health care field. While big-picture changes usually come more slowly in medicine due to entrenched hospitals and insurers and the necessary diligence required when dealing with a person’s well-being, the potential to improve and save ever more lives ought to make AI a priority for decision makers as it develops.

Advanced software has already been able to predict heart attacks and strokes better than traditional methods and create end-of-life treatment paths to better alleviate symptoms of long-suffering terminal patients. It’s not always pretty, but when AI can make life-or-death decisions more assured, there’s no question it’s worth pursuing.

Predictive AI can even stem health problems on the largest scale, allowing scientists to halt highly contagious diseases before they go global. A team of researchers from the University of Georgia, Massey University, and the University of California was able to use AI to model potential hot spots for outbreaks by monitoring the movements of likely disease vectors in bats. The next epidemic may be over before it ever takes root, thanks to a healthy dose of AI.

Meteorology

If there’s one field that stands to make a quantum leap forward with AI, it’s the millennia-old practice of weather forecasting. Ever since ancient Babylonians looked to the clouds to determine what was coming next, we’ve struggled to accurately and consistently make reliable weather predictions. Computer modeling represented a major step in the process, and AI looks poised to make another leap.

In 2016, researchers were able to identify complex atmospheric phenomena, formerly something that could only be accomplished by teams of human experts, through AI software. Artificial intelligence’s power here is in its ability to assess information at a scale far beyond what even the brightest scientists are capable of. We can now analyze patterns in air pressure by the petabyte, meaning weather-prediction models can lean on the entirety of weather-observation history to make predictions.

The same software currently used to identify images and language can do the same for weather, and while it might not yet be able to predict when hurricanes will arise, it can rapidly model and forecast scenarios for their growth and movement, potentially saving lives in the process.

These are just a few illustrative examples, but globally, we can expect massive changes in how we look at driving, law enforcement, and even human interaction itself, arising from the prediction power of AI. For any of us hoping to look into a machine learning crystal ball to see the future, that may be impossible for now. But the coming years are sure to usher in a new level of predictive power in spaces that we encounter every day as consumers. Whatever mystical powers the ancient oracles may have had, they definitely didn’t see this coming.

Bennat Berger is the cofounder and principal of Novel Property Ventures in New York City.

Bookmarks Is A New Way To Save Tweets On Twitter


Twitter first talked about its Bookmarks feature last year when it started testing the feature. It lets users privately save tweets for later without having to engage with them. The only way to save tweets for later previously was to “like” them by tapping on the heart icon. This let the poster know that you had engaged with their tweet and may even convey the impression that you actually liked it. Bookmarks eliminates the need to engage with tweets just because you want to save them for later.

The big difference between Bookmarks and the heart button is that with the former, it’s only you who knows what tweets have been saved for later. The account that posted that particular tweet doesn’t know which of their tweets have been bookmarked by which user.

Users will now see a new share icon on every tweet which will let them bookmark a tweet, share it via Direct Message or share it off of Twitter in different ways. Users can easily bookmark tweets by tapping on this icon and selecting “Add Tweet to Bookmarks.” Tweets can be removed from Bookmarks at any time.

The Bookmarks is now being rolled out globally on Twitter for iOS, Android, desktop, and mobile.

Bookmarks Is A New Way To Save Tweets On Twitter , original content from Ubergizmo. Read our Copyrights and terms of use.

AT&T says repealing net neutrality and implementing “paid prioritization” are matters of safety


AT&T supports all of you who are out there right now fighting to have the net neutrality repeal overturned. It’s just dissapointed. And if you don’t stop, it’s going to take away your video games and your car. In a company blog post, Bob Quinn, AT&T’s Senior Executive Vice President of External & Legislative Affairs, claims people have it wrong when it comes to fast lanes and throttling. “But no discussion of net neutrality would be complete without also addressing the topic of paid prioritization. Let me start by saying that the issue of paid prioritization has always been hazy…

This story continues at The Next Web

Dangerous storm targets Northeast, Mid-Atlantic with high winds, coastal flooding, inland snow

TwitterFacebook

For weeks, weather experts have been expecting that early March would live up to its reputation of coming in “like a lion,” with a pattern conducive to major East Coast storms. That expectation is going to become a reality, first during the next three days, and then again next week. 

However, the specifics of the weather forecast, particularly what form of precipitation will fall in areas like New York City, Hartford, and Boston, are turning meteorologists’ hair white, and causing weather Twitter to resemble a therapy session.

SEE ALSO: Drastic Arctic warm event stuns scientists, as record-breaking temperatures reach the North Pole Read more…

More about Climate, Science, Extreme Weather, Jet Stream, and Nor Easter

Quick Takes: Apple Watch Apps Called ‘Frustrating’ to Develop, iOS 11 Jailbreak With Cydia, and More

In addition to our standalone articles covering the latest Apple news and rumors at MacRumors, this Quick Takes column provides a bite-sized recap of other headlines about Apple and its competitors on weekdays.


Wednesday, February 28

iOS 11 jailbreak released with Cydia, which turns 10 today: Cydia was released by Jay Freeman, better known as saurik, on February 28, 2008. The unofficial App Store is the gateway to installing apps, tweaks, themes, and other files on jailbroken iPhone, iPad, and iPod touch devices, outside of Apple’s control.

Just yesterday, a new jailbreak called Electra was released for iOS 11 through iOS 11.1.2. It is the first iOS 11 jailbreak to include Cydia, although it is a modified version that may cause issues, so proceed at your own risk.

Commentary: While jailbreaking has faded in popularity over the years, Cydia remains a centerpiece of the community after a decade. Freeman is still in the process of updating Cydia and its frameworks to be fully compatible with iOS 11, so it might be a good idea to hold off on using the Electra jailbreak to avoid problems.

Waymo’s self-driving cars have now covered five million miles on public roads: Waymo says the first million miles took six years to complete, while the fifth million took just under three months, as its testing rapidly speeds up.

To celebrate the milestone, Waymo shared a 360-degree video today that reveals how its self-driving vehicles view their surroundings, recognize objects like cars and pedestrians, and predict what those things will do. The video combines footage and real-time data from a trip around Phoenix, Arizona.



Commentary: A recent report claimed Apple is accelerating development of its own self-driving software to compete with the likes of Waymo, but it sounds like the iPhone maker may have significant ground to make up still.

Flutter enters beta, Sketch 49 released with iOS 11 design templates: Bohemian Coding’s popular app design tool Sketch has been updated with Apple’s official iOS 11 design templates, including tab bars, status bars, buttons, and other user interface elements for developers to incorporate into their apps.



In related news, Google’s new mobile UI framework Flutter entered beta yesterday. Flutter aims to help developers more quickly craft high-quality native interfaces for both iOS and Android, with support for both iOS 11 and the iPhone X on the Apple side. Beginners can read Flutter’s Getting Started guide.



Commentary: Sketch and Flutter can help developers to more quickly design iPhone and iPad apps that are consistent with iOS 11’s design language.

Marco Arment says developing Apple Watch apps is “extremely frustrating and limited”: Apple doesn’t give developers access to the same watchOS frameworks that it uses for its own Apple Watch apps. Instead, it offers WatchKit, which Arment argues can only be used to create “baby” apps.

Image Credit: MacStories


Commentary: There has certainly been a trend of some developers discontinuing their Apple Watch apps in recent months, including notable brands like Amazon, eBay, Google Maps, and Slack. Apple providing developers with expanded watchOS resources could encourage those companies to reconsider.

For more Apple news and rumors coverage, visit our Front Page, Mac Blog, and iOS Blog. Also visit our forums to join in the discussion.

Tag: Quick Takes

Discuss this article in our forums