HomePod Diary: The best party gadget ever, and the difference between ‘really good’ and ‘great’

Friday morning, my HomePod arrived; Friday evening, we had a small party. And the belle of the ball was the HomePod.

We had a group of friends round for drinks and pizza. That seemed a good opportunity to put HomePod to the test – both to see what others thought of it as a speaker, and to allow friends to try Siri control …

more…

9to5Mac

We Know You Don’t Really Read Privacy Policies. This AI Can Do It For You.

Visualized Summaries

“I have read and understood…” has got to be one of the biggest lies people commit on a regular basis. It’s the typical ending for the long-winded customer agreements or privacy policies attached to every online service, which few ever read. Or at least, not in their entirety.

When humankind finds something difficult, we typically build a gizmo to do it for us — and this case is no different. It turns out, reading lengthy fine-print is the expertise of a machine-learning artificial intelligence (AI) designed by researchers from the Federal Institute of Technology at Lausanne, Switzerland (EPFL), the University of Wisconsin, and the University of Michigan. Their research began with a question, said lead researcher Hamza Harkous from EPFL.

Privacy and the Internet of Things
Click to View Full Infographic

“What if we visualize what’s in the policy for the user?” Harkous said in an interview with WIRED. “Not to give every piece of the policy, but just the interesting stuff.”

The question led to Polisis, which is short for “privacy policy analysis.” It works as a website and browser extension that can read and make sense of terribly long privacy policies in web services that cost users both time and, ultimately, money. Polisis can pull a website’s privacy policy and read it in around 30 seconds. It then comes up with a summary and a graphic flowchart that highlights how the online service handles a user’s data.

Digestible Chunks

The researchers trained Polisis using data sets from 115 privacy policies, analyzed and annotated by Fordham Law students, and 130,000 others taken from Google Play Store apps. What makes this AI appealing is how it’s able to transform a boring, wordy privacy policy into something that’s digestible. And if users are still having trouble with aspects of the agreement, the researchers have another service on their website — an AI-powered chatbot called PriBot that users can discuss privacy policies with.

Google's privacy policy, summarized by Polisis. Image credit: Polisis, screenshot
Google’s privacy policy, summarized by Polisis. Image credit: Polisis, screenshot

Despite internet users claiming to have read a privacy policy, only about 20 percent actually do. Even then, they don’t necessarily understand the legal language embroidered with verbosity in every policy. That’s why Polisis could be so useful — helping internet users protect their privacy similarly to how a lawyer chatbot helped resolve driver’s parking ticket issues.

A bit of a fine print here, though. Even with a powerful tool like Polisis, the barest minimum effort is still required from the human user. In other words, a policy summary is not helpful if you don’t actually bother to read it. The AI can help, but human users have to help themselves first.

The post We Know You Don’t Really Read Privacy Policies. This AI Can Do It For You. appeared first on Futurism.

Futurism

Can a Brain Zap Really Boost Your Memory?

If this is the first you’re hearing about electrical stimulation of the brain, you’re probably imaging it as a plot device central to a Netflix Original Black Mirror / House of Cards crossover; some kind of torture method to get spies of the future to spill it.

Not only are these techniques not torturous, they’re not even particularly futuristic. Deep Brain Stimulation (DBS) has been used to treat neurological conditions like Parkinson’s and epilepsy for decades, and it’s now been explored as a way of keeping symptoms of dementia at bay. Transcranial Magnetic Stimulation (TMS), has existed since the 1980s as a treatment for major depression.

TMS uses magnetic pulses and has long been the less invasive of the two, but both techniques rely on the ability to target only certain areas of the brain. This is especially important in DBS, which uses electrodes implanted into the patient’s brain to target specific regions. As you might expect, implanting an electrode into someone’s brain so you can send an electrical current to it is not something neuroscientists do willy-nilly just to run experiments. Generally speaking, the research we have about these methods draws on the experiences of patients who already have the implants for treatment.

Image Credit: Creative Commons
Image Credit: Creative Commons

Two recent studies, one from the Mayo Clinic and the other from the University of Pennsylvania, looked at whether these therapies could have unrealized potential. Patients with degenerative neurological conditions can certainly have trouble with their memory, but could these therapies also be used in patients who don’t have a neurological disorder in need of treatment?

In order to study the effect of well-placed electrical zaps to the brain had on memory, researchers in the Mayo study asked groups of patients to try to remember a list of words as they zapped a few different regions of their brains. Of the 22 patients in the study, the four who had the lateral temporal cortex region of their brain electrically stimulated recalled more words than the others. This probably wasn’t a coincidence, because that’s the part of our brain that helps us process language.

Meanwhile, the researchers at the University of Pennsylvania were less concerned about which region got the electrical jolt, and more concerned about the timing of it. Their previous research had shown that zapping the brain at the wrong moment could actually have a negative effect on the patient’s ability to remember (oops). The Penn researchers also had a little help the second time around: a computer model that would help them get the timing just right by assessing how well a patient’s learning was going.

Based on the patient’s brain activity, the computer model could tell when they’d learned the words given to them in a memory test – and when they hadn’t. The electrical impulse was triggered whenever the model determined the patient hadn’t learned the word effectively.

The researchers may have been on to something when it came to not just well-placed, but well-timed, zaps: the study showed that they enhanced a patient’s learning and memory by up to 15 percent.

Both of these studies were limited in scope, though. Researchers elsewhere in neuroscience who have responded to the results are generally wary and point out that they don’t address one of the biggest qualms in the field: would a treatment like this work if the memory area of the brain was damaged?

For the time being, better learning through electrical brain implant is something relegated to the future. If you were hoping to use technology to enhance your memory, you’ll have to stick to those brain-training apps for now.

The post Can a Brain Zap Really Boost Your Memory? appeared first on Futurism.

Futurism