Two versions of reality.💫

Physicists say at the quantum level can two versions of reality exist at the same time.

Researchers recently conducted experiments to answer a decades-old theoretical physics question about dueling realities. This tricky thought experiment proposed that two individuals observing the same photon could arrive at different conclusions about that photon's state  and yet both of their observations would be correct.

For the first time, scientists have replicated conditions described in the thought experiment confirmed, that even when observers described different states in the same photon, the two conflicting realities could both be true. 

When an observer in an isolated laboratory measures the photon, they find that the particle's polarization — the axis on which it spins — is either vertical or horizontal.

However, before the photon is measured, the photon displays both polarizations at once, as dictated by the laws of quantum mechanics; it exists in a superposition of two possible states.

Once the person in the lab measures the photon, the particle assumes a fixed polarization. But for someone outside that closed laboratory who doesn't know the result of the measurements, the unmeasured photon is still in a state of superposition.

That outsider's observation — their reality — therefore diverges from the reality of the person in the lab who measured the photon. Yet, neither of those conflicting observations is thought to be wrong, according to quantum mechanics.

Theoretical advances were needed to formulate the problem in a way that is testable. Then, the experimental side needed developments on the control of quantum systems to implement something like that.

Tested Wigner's original idea with an even more rigorous experiment which doubled the scenario. They designated two "laboratories" where the experiments would take place and introduced two pairs of entangled photons, meaning that their fates were linked, so that knowing the state of one automatically tells you the state of the other. (The photons in the setup were real. Four "people" in the scenario "Alice," "Bob" and a "friend" of each — were not real, but instead represented observers of the experiment).

The two friends of Alice and Bob, who were located "inside" each of the labs, each measured one photon in an entangled pair. This broke the entanglement and collapsed the superposition, meaning that the photon they measured existed in a definite state of polarization. They recorded the results in quantum memory — copied in the polarization of the second photon.

Alice and Bob, who were "outside" the closed laboratories, were then presented with two choices for conducting their own observations. They could measure their friends' results that were stored in quantum memory, and thereby arrive at the same conclusions about the polarized photons.

But they could also conduct their own experiment between the entangled photons. In this experiment, known as an interference experiment, if the photons act as waves and still exist in a superposition of states, then Alice and Bob would see a characteristic pattern of light and dark fringes, where the peaks and valleys of the light waves add up or cancel each other out. If the particles have ''chosen" their state, you'd see a different pattern than if they hadn't. 

The authors of the new study found that even in their doubled scenario, the results described by Wigner held. Alice and Bob could arrive at conclusions about the photons that were correct and provable and that yet still differed from the observations of their friends — which were also correct and provable, according to the study.

Quantum mechanics describes how the world works at a scale so small that the normal rules of physics no longer apply; over many decades, experts who study the field have offered numerous interpretations of what that means.

However, if measurements themselves aren't absolutes — as these new findings suggest — that challenges the very meaning of quantum mechanics.

"It seems that, in contrast to classical physics, measurement results cannot be considered absolute truth but must be understood relative to the observer who performed the measurement," 

HACKER GROUP HOLDING INFO FOR RANSOM🤑

Hacker group has threatened to leak thousands of "secret" documents stolen from insurers and government agencies that they claim reveal the truth about 9/11 - unless they're paid not to. 

The Dark Overlord, a "professional adversarial threat group" known for their hacks of Netflix, plastic surgery clinics, and other sensitive targets, posted a link to a 10GB encrypted archive of documents related to 9/11 litigation, promising to release the encryption keys if their demands were not met in a post on Pastebin on Monday. 

The group claims the documents tell the story of what really happened on one of the most notorious dates in recent history, tweeting "We'll be providing many answers about 9.11 conspiracies through our 18,000 secret documents leak." They published a "teaser" consisting of letters, emails, and various documents that mention law firms, the Transport Security Administration, and the Federal Aviation Administration, with a promise of more to come. 

They claim to have hacked documents from not only major global insurers like Lloyds of London and Hiscox, but also Silverstein Properties, which owned the World Trade Center complex, and various government agencies. The material, which supposedly includes confidential government documents that were meant to be destroyed but were instead retained by legal firms, allegedly reveals "the truth about one of the most recognizable incidents in recent history and one which is shrouded in mystery with little transparency and not many answers." 

Anyone worried they might be named in the documents can have their names redacted - for a fee, according to the announcement. "Terrorist organizations" and "competing nation states of the USA" are also offered first dibs on the info - if they pay up. Otherwise, the hackers write, the insurers can pay an unspecified bitcoin ransom - or "we're going to bury you with this." 

Some of the documents were nabbed in an April hack of a law firm associated with Hiscox that the firm acknowledges could have exposed 1,500 of its US commercial policyholders. The Dark Overlord claims that while their ransom was paid in relation to that earlier hack, their victim violated the "agreement" by cooperating with law enforcement, necessitating further extortion. 

The group emerged in 2016 with hacks on medical centers, advertising sensitive data for sale on the dark web in order to force victims to pay for its removal. They infamously leaked an entire season of Netflix's Orange is the New Black last year to prove to that company they meant business and have stolen data from more than 50 companies.

The group is also offering to sell this data on the dark web.

Tor Browser for Android🆕

Tor browser is famous for establishing an anonymous connection by bundling data into encrypted packets before passing them through the network, thus hiding your internet footprint.

Mobile browsing is on the rise around the world, and in some parts, it is commonly the only way people access the internet. In these same areas, there is often heavy surveillance and censorship online.

There’s never been an official Tor Browser on mobile. Until now.

This first official version of Tor Browser on Android will replace Orfox, another Android browser that featured support for navigating the web via the Tor network and its Onion protocol.

The Tor Browser for Android app is available through the official Google Play Store and is currently considered an alpha release, meaning some bugs should be expected.

This first official version of Tor Browser on Android will replace Orfox, another Android browser that featured support for navigating the web via the Tor network and its Onion protocol.

Orfox was, until recently, developed by the Guardian Project, but has been retired today after the Tor Project's official announcement.

The Tor Project is now calling on its dedicated fanbase to help test the Android version.

For the alpha versions of the Tor Browser for Android, users will also need to install Orbot, a mobile proxy application that will help the Android browser connect to the Tor network. For future releases, the Tor Project says this won't be necessary.

The Tor Browser for Android comes equipped with:

Tracker Blocker -- Tor Browser isolates each website users visit so third-party trackers and ads can't follow them. Any cookies automatically clear when users are done browsing.

Fingerprinting Protection -- Tor Browser for Android makes it difficult for users to be fingerprinted based on their browser and device information.

No ISP Blocks -- Users are free to access any sites, even those blocked by local ISPs.

Connections Through the Tor Network -- All anyone monitoring a user's network connection and browsing habits can see is that the user is accessing the Tor network or accessing sites through IPs belonging to the Tor network.

Multi-Layered Encryption -- All traffic that passes through the Tor network is encrypted with the help of the Onion protocol.

The Tor Project didn't say if it was planning on developing an iOS version, but in the meantime, continued to recommend that iOS users use Mike Taggs' Onion Browser app.

The release of the Android version of the Tor Browser comes a day after the Tor Project released a major update to its desktop version.

The Tor Project released Tor Browser v8, which updated the Tor Browser's underlying codebase to the Firefox Quantum codebase that Mozilla launched in November 2017 with Firefox 57, and which has received rave reviews for its small memory footprint and speedier page loads.

Introducing Tor Browser for Android (alpha), the mobile browser with the highest privacy protections ever available and on par with Tor Browser for desktop. You can download the alpha release on GooglePlay. The stable release is slated for early 2019.

Mineral created in lab capable of capturing CO2💎

Most people agree that climate change is a huge problem for the modern world.

Scientists are already skeptical that we’ll be able to return atmospheric CO2 levels below 400 parts per million.But maybe not for much longer.

New research from Trent University could revolutionize the way we capture and store carbon emissions leaked into the atmosphere which could slow down climate change.

According to the researchers, the solution lies in speeding up the formation of magnesite, a mineral that can store carbon but takes thousands of years to form in nature.

During the recent Goldschmidt Conference in Boston, Trent researchers presented their findings and proposed possible uses for newly formed magnesite.

The researchers’ methods occur at room temperature which makes it an energy efficient process and economically viable.

This would be important for mass production at the industrial level, which the researchers say would have to happen for any real impactful carbon capture to occur.

Fast-forming magnesite could also store CO2 for a long time and prevent any further release into the atmosphere unlike natural carbon sinks like forests and oceans.

The researchers realized that one ton of naturally-occurring magnesite can remove about a half a ton of CO2 from the atmosphere but the mineral takes a long time to form.

“Our work shows two things. Firstly, we have explained how and how fast magnesite forms naturally,” said Ian Power, the project leader of the study. “This is a process which takes hundreds to thousands of years in nature at Earth’s surface. The second thing we have done is to demonstrate a pathway which speeds this process up dramatically.”

In order to remedy the slow rate of formation, the researchers used polystyrene microspheres which worked as a catalyst for the mineral. After 72 days, the researchers had fully formed magnesite ready for carbon capture.

What’s especially exciting is that the polystyrene microspheres can be reused again.

“Using microspheres means that we were able to speed up magnesite formation by orders of magnitude,” said Power. “This process takes place at room temperature, meaning that magnesite production is extremely energy efficient.”

The researchers say that getting magnesite formation to the industrial level required to help combat climate change is a ways off, but that the results show that from a scientific perspective, it’s not an impossibility.

“It is really exciting that this group has worked out the mechanism of natural magnesite crystallization at low temperatures, as has been previously observed – but not explained – in weathering of ultramafic rocks,” said Peter Kelemen, a professor at Columbia University who was not involved in the study. “The potential for accelerating the process is also important, potentially offering a benign and relatively inexpensive route to carbon storage, and perhaps even direct CO2 removal from air.”

For now, we recognize that this is an experimental process.

Voynich manuscript finally been cracked by AI📜

Scientists had been scratching their heads of half-a-millennium about the so-called Voynich manuscript, but no one was able to solve it.

And now a computer scientist claims to have cracked it using AI.The manuscript is written in ancient Hebrew and the code involves shuffling the order of letters in each word and dropping the vowels.

For decades codebreakers have poured over the 240-page relict, from which pages have been carbon dated back to the early 1400s. It has been suggested that the manuscript, which is housed at the Beinecke Library at Yale University and contains brightly colored images of plants and an unknown language, is a medieval medical journal for women.

The first stage of the research was working out the manuscript’s language. The experts used 400 different language translations from the “Universal Declaration of Human Rights” to identify the language used in the text.

“That was surprising,” said Kondrak, in a statement. “And just saying ‘this is Hebrew’ is the first step. The next step is how do we decipher it.”

Kondrak and Hauer worked out that Voynich manuscript was created using ‘alphagrams’ that use one phrase to define another so built an algorithm to unscramble the text.

The initial part of the text was then run through Google Translate. “It came up with a sentence that is grammatical, and you can interpret it.” 

The sentence was: “She made recommendations to the priest, man of the house and me and people.”

The full meaning of the text will need the involvement of historians of ancient Hebrew. The vellum, or animal skin, on which the codex is written has been dated to the early 15th century.

The research study is published in Volume 4 ofTransactions of the Association for Computational Linguistics.

There have been multiple attempts to decode the Voynich manuscript. In 2014, for example, researchers argued that the illustrations of plants in the manuscript could help decode the text’s strange characters. In 2011, a self-proclaimed “prophet of God” claimed that he had decoded the book.

YouTube Makes Changes💰

Five years ago, YouTube opened their partner program to everyone. This was a really big deal: it meant anyone could sign up for the service, start uploading videos, and immediately begin making money. This model helped YouTube grow into the web’s biggest video platform.

Today YouTube has announced, that new channels will have to work a bit harder for monetization privileges, and current Partner channels will have to maintain high viewership in order to stay profitable. To be specific, channels will now need to have 4,000 total viewing hours in a rolling 12-month period, along with at least 1,000 subscribers. New channels must hit these metrics in order to monetize, and existing channels that meet the old requirement of 10,000 cumulative views will have a 30 day grace period to get in compliance or be dropped. Dropped channels will be paid out any remaining AdSense balance due to them.

YouTube believes that this threshold will give them a chance to gather enough information on a channel to know if it’s legit. And it won’t be so high as to discourage new independent creators from signing up for the service.

The blog post outing the new rules went live earlier today, and the comments section is already above 200 comments, with the mass consensus being that this move has questionable benefit given the challenges currently facing the platform, and will significantly harm smaller channels. Commenters predict a mass increase in smaller channels subscribing to one another in order to help get their sub count up, as well as using paid subscriber services.

The viewing hours requirement can also be faked, but would require more than creators would likely get out of it. Larger creators would be largely unaffected, commenters argue, though it’s quite feasible that a public outcry that causes a creator’s fan base to drop off could end up dipping their viewing hours below the requirement. 4,000 hours in a 12 month period essentially equates to about 340 hours per month. This means that the output of the average large creator, a single 30-minute video per day, would require each video to have at least 12 unique viewers. That does not sound like a hefty feat, but for smaller creators, who may upload once a week or even more seldom, it all but locks them out of the ecosystem. 

As it moves ever closer to parity with the world of prime-time television, YouTube is sensibly taking steps to police how business is done on its service. Time will tell how a rising generation of creators respond to these new limitations.

Digital World hits panic mode😱

Google's Project Zero team last June found two kernel flaws, codenamed Spectre and Meltdown that could allow hackers to gain access to sensitive information like passwords or photos stored in a PC's protected memory or cloud server. These vulnerabilities impact Windows, Linux, iOS, macOS, tvOS, and Android devices. At the centre of all this is Intel as it produces processors for a majority of computing devices around the world.

It has been discovered that a performance feature called speculative execution present on nearly all modern computing system to optimise performance. The techniques essentially makes the computer speculate which command or path to take. This, however, requires access to protected kernel memory, which hackers can exploit through malicious programs.

Project Zero also said that in order to exploit the vulnerability, the attacker will need access to the machine and should be able to run a malicious app. Apple has also stated that an attacker will need to run a malicious app on the iOS or macOS device to exploit the vulnerability, which is why it is urging its users to download apps only through trusted sources such as the App Store.

While Meltdown is said to affect Intel processors manufactured since 1995, Spectre is more widespread in that it is present in ARM and AMD-based devices as well. This means that apart from PCs, Spectre is present in smartphones as well. Apple has admitted this for its iOS devices like the iPhone and iPad. Google's Project Zero team notes that "exploitation has been shown to be difficult and limited on the majority of Android devices."

Apple confirmed that Spectre and Meltdown affects its Mac and iOS devices. It has mitigated Meltdown with iOS 11.2, macOS 10.13.2, and tvOS 11.2 updates, and plans to release mitigations in Safari to defend against Spectre. The company says it hasn't found any reports of the vulnerabilities affecting customers as of now.

Intel CEO Brian Krzanich says that the company is looking to fix the security vulnerabilities via updates and does not see a need for a recall. The company said that 90 per cent of computers released in the last 5 years will have fixes available by the end of next week. While this may seem to be the case for Meltdown, Spectre is a more widespread and deep-rooted flaw and there is no fix for it as of now. The latter is, however, a harder exploit for hackers to carry out.

Companies like Apple, Microsoft, Google and Amazon are working on updates to mitigate the flaws. One of the concerns is that the updates may slow performance but Krzanich has denied this. As of now users can’t do much but update their devices to the latest security patches to mitigate the vulnerabilities. Chip makers will have to redesign future processors that will be protected against the exploits and its variants.
It is hard to tell whether hackers have carrier out attacks through these exploits as neither Spectre nor Meltdown leaves any trace in log files. Most tech companies have said that they have not found any proof of the flaws being used to attack devices.

ARM has claimed that a majority of its processors have not been affected by Spectre or Meltdown. AMD has resolved a variant of Spectre via software and operating system updates, while another variant has "a near zero risk of exploitation" on its processors.

Google has notified that Android devices on the latest security patch are protected. However, as you know, Android updates don't roll out universally for everyone. Most devices, especially older models and budget smartphones are not on the latest security update. This puts many at risk from potential attackers.

Windows 10 users have already received the update KB4056892 earlier this month, while Windows 7 and Windows 8 users will be getting the update next Tuesday. Those using Chrome browser will receive an update on January 23. Firefox users will need to be on version Firefox 57.0.4 to be protected against any attack.

Intel and other technology companies have been made aware of new security research describing software analysis methods that, when used for malicious purposes, have the potential to improperly gather sensitive data from computing devices that are operating as designed. Intel believes these exploits do not have the potential to corrupt, modify or delete data.

Recent reports that these exploits are caused by a “bug” or a “flaw” and are unique to Intel products are incorrect. Based on the analysis to date, many types of computing devices — with many different vendors’ processors and operating systems — are susceptible to these exploits.

Intel is committed to product and customer security and is working closely with many other technology companies, including AMD, ARM Holdings and several operating system vendors, to develop an industry-wide approach to resolve this issue promptly and constructively. Intel has begun providing software and firmware updates to mitigate these exploits.

Contrary to some reports, any performance impacts are workload-dependent, and, for the average computer user, should not be significant and will be mitigated over time.

Intel is committed to the industry best practice of responsible disclosure of potential security issues, which is why Intel and other vendors had planned to disclose this issue next week when more software and firmware updates will be available. However, Intel is making this statement today because of the current inaccurate media reports.

Check with your operating system vendor or system manufacturer and apply any available updates as soon as they are available. Following good security practices that protect against malware in general will also help protect against possible exploitation until updates can be applied.

Intel believes its products are the most secure in the world and that, with the support of its partners, the current solutions to this issue provide the best possible security for its customers.

Artificial “photos” that look real to humans.👶

NVIDIA recently published a paper titled “Progressive Growing of GANs for Improved Quality, Stability, and Variation.” GAN stands for “generative adversarial network,” and it’s a system that uses two neural networks — one generates things and the other evaluates them. These algorithms are capable of generating artificial “photos” that look real to humans.

To demonstrate the system, NVIDIA trained the neural network using the CelebA HQ. 

They used an increasingly popular AI training system called a general adversarial network (GAN). The program is fed a massive data set — in this case celebrity photos — and then gets better at creating the desired result (in this case, realistic computer-generated faces) over a period of days or weeks. The "adversarial" component involves pitting two machine-learning programs against one another, with one program "challenging" the other's face creations. 

The future applications of rendering realistic looking people — that aren't actually people — seems both potentially useful and unsettling. This is certainly a boon to graphics companies that always need new images of people, perhaps for use in advertising. But it's also important to recognize AI-programs are getting closer and closer to achieving realism in artificial faces; how this might be employed in fake news and other means of deception is unknown — and boundless. For now, though, it's at least given us some really interesting faces to look at.

For its project, NVIDIA found that training the neural network using low-resolution photos of real celebrities and then ramping up to high-res photos helped to both speed up and stabilize the “learning” process, allowing the AI to create “images of unprecedented quality.”

AI Crushed a Human at Dota 2🤖

Elon Musk founded OpenAI as a nonprofit venture to prevent AI from destroying the world, something Musk has been beating the drum about for years. Just last month he told a group of US governors that AI represents a “fundamental risk to the existence of civilization.”

The idea that an AI can be trained to beat the best human players at virtually any game shouldn’t be a shock in 2017, but somehow, we’re still not quite used to the idea.  This year’s “The International Dota 2 Championships took place August 7-12 at Key Arena in Seattle, and one of the big events pitted Danylo “Dendi” Ishutin against an OpenAI bot.  The OpenAI bot also defeated other top pros during the course of the event.

Dota 1v1 is a complex game with hidden information. Agents must learn to plan, attack, trick, and deceive their opponents. The correlation between player skill and actions-per-minute is not strong, and in fact, AI’s actions per minute are comparable to that of an average human player.

In this style of play, OpenAI’s bot is completely dominant. Early into the first game, the event’s host asked Dendi: “Does it feel like a player, like a person?” “Uhhhh… nope,” Dendi answered, distracted by yet another perfectly-timed attack. “This guy is really scary,” Dendi kept saying, and at one point yelped “Please stop bullying me!” when the bot rushed at him aggressively.

Bot is trained entirely through self-play,it starts off completely random with no knowledge of the world and simply plays against a copy of itself, which means it always has an evenly matched opponent, and it climbs this ladder of skill level until it's able to reach the performance of the best professional players in the world.

While gaming and esports might not seem like the most obvious avenue for academic development, OpenAI’s bot is the latest example of modern AI applications. With machine learning, bots like OpenAI’s can generally be put into a scenario and independently learn from its past mistakes to become smarter.

From the very beginning, it just keeps playing against a copy of itself. It starts from complete randomness and then it makes very small improvements, and eventually it's just pro level.

OpenAI robo-brain clearly came out of the exhibition looking like the stronger player, but the tech company isn't quite finished. An accompanying video notes that the bot is still a work-in-progress. The goal is to eventually assemble a full team of AI bots for 5v5 matches and, further down the road, to mix AI players in with human players on a single team.

 it is time to put forward a set of rules regulating the very nature of AI.

By the time we are reactive, it’s too late. Normally the way regulation works out is that a whole bunch of bad things happen, there’s public outcry and after many years a regulatory agency is setup to regulate the industry.

AI is a rapidly-developing technology, but it is still far away from self-evolving, almighty software. Facebook uses AI for targeted advertising, whereas Microsoft and Apple use it to power their digital assistants, Cortana and Siri. Google search engine has also been dependent on AI since its inception.


It is time to put forward a set of rules regulating the very nature of AI.

Tesla Model S Sets New Record🚗

On August 3, 2017, Tesla Owners Italia completed the 27 hour marathon of driving a Tesla Model S 100D at approximately 40 km/h (25 mph) to cover 1,078 km (669.83 miles) on a single charge.

The team managed to set the record using just 91 Wh per km, in total using 98.4 kWh.

That’s the new record for the longest distance covered on a single charge in a Tesla ever.

Tesla CEO Elon Musk said achieving 1,000 km on a single charge in a Model S 100D was possible and now it’s been done.

Elon Musk Retweeted Tesla Owners Italia.
Officially verified as the first production electric car to exceed 1000km on a single charge! Congratulations Tesla Owners Italia!!

FBI arrests WannaCry hero for alleged role in Kronos malware🕵

Marcus Hutchins, who goes by the name “Malwaretech” online, was arrested on six counts on Wednesday as he tried to fly home from Las Vegas, where he was attending the cyber security conferences Black Hat and Def Con.

Marcus Hutchins was hailed as a hero after he discovered a so-called “kill switch” in the WannaCry ransomware that spread through multinationals and the UK’s National Health Service in May. This stopped the spread of the malicious software, which was rendering computers unusable until victims paid a ransom.

Marcus Hutchins is charged with one count of conspiracy to commit computer fraud and abuse, three counts of distributing and advertising an electronic communication interception device, one count of endeavouring to intercept electronic communications and one count of attempting to access a computer without authorisation. The charges were the result of a two-year long investigation by the FBI Cyber Crime Task in Milwaukee.

While major players like Zeus, Gozi, Citadel and other advanced financial malware dominate the malware threat landscape, newcomers and challengers always try to get a share of the cyber crime market. One such new malware that was recently made available for purchase in a Russian underground forum is the Kronos malware. With a $7,000 price tag, this malware offers multiple modules for evading detection and analysis as well as an option to test the malware for a week prior to buying it.

Kronos malware downloaded from email attachments left victims' systems vulnerable to theft of banking and credit card credentials, which could have been used to siphon money from bank accounts.

The indictment alleges that the unidentified co-defendant advertised the Kronos malware on AlphaBay, a dark web marketplace that international authorities took offline last month. Investigators said the site allowed anonymous users to facilitate global trade in drugs, firearms, hacking tools and other illicit goods.

The Justice Department said Kronos was used to steal banking systems credentials in Canada, Germany, Poland, France, the United Kingdom and other countries.

Within the cyber security community, Hutchins was heralded as a folk hero for his apparent role in stopping the WannaCry attack, which infected hundreds of thousands of computers and caused disruptions at car factories, hospitals, shops and schools in more than 150 countries.

A Justice Department official said his arrest was unrelated to WannaCry.

Some security researchers and computer crime experts said they were skeptical of the charges against Hutchins.

Minds Mastering Machines🤖

A strange self-driving car was released onto the quiet roads of Monmouth County, New Jersey. The experimental vehicle, developed by researchers at the chip maker Nvidia, didn’t look different from other autonomous cars, but it was unlike anything demonstrated by Google, Tesla, or General Motors, and it showed the rising power of artificial intelligence. The car didn’t follow a single instruction provided by an engineer or programmer. Instead, it relied entirely on an algorithm that had taught itself to drive by watching a human do it.

Getting a car to drive this way was an impressive feat. But it’s also a bit unsettling, since it isn’t completely clear how the car makes its decisions. Information from the vehicle’s sensors goes straight into a huge network of artificial neurons that process the data and then deliver the commands required to operate the steering wheel, the brakes, and other systems. The result seems to match the responses you’d expect from a human driver. But what if one day it did something unexpected—crashed into a tree, or sat at a green light? As things stand now, it might be difficult to find out why. The system is so complicated that even the engineers who designed it may struggle to isolate the reason for any single action. And you can’t ask it: there is no obvious way to design such a system so that it could always explain why it did what it did.

The mysterious mind of this vehicle points to a looming issue with artificial intelligence. The car’s underlying AI technology, known as deep learning, has proved very powerful at solving problems in recent years, and it has been widely deployed for tasks like image captioning, voice recognition, and language translation. There is now hope that the same techniques will be able to diagnose deadly diseases, make million-dollar trading decisions, and do countless other things to transform whole industries.

But this won’t happen—or shouldn’t happen—unless we find ways of making techniques like deep learning more understandable to their creators and accountable to their users. Otherwise it will be hard to predict when failures might occur—and it’s inevitable they will. That’s one reason Nvidia’s car is still experimental.

Already, mathematical models are being used to help determine who makes parole, who’s approved for a loan, and who gets hired for a job. If you could get access to these mathematical models, it would be possible to understand their reasoning. But banks, the military, employers, and others are now turning their attention to more complex machine-learning approaches that could make automated decision-making altogether inscrutable. Deep learning, the most common of these approaches, represents a fundamentally different way to program computers.

There’s already an argument that being able to interrogate an AI system about how it reached its conclusions is a fundamental legal right. Starting in the summer of 2018, the European Union may require that companies be able to give users an explanation for decisions that automated systems reach. This might be impossible, even for systems that seem relatively simple on the surface, such as the apps and websites that use deep learning to serve ads or recommend songs. The computers that run those services have programmed themselves, and they have done it in ways we cannot understand. Even the engineers who build these apps cannot fully explain their behavior.

This raises mind-boggling questions. As the technology advances, we might soon cross some threshold beyond which using AI requires a leap of faith. Sure, we humans can’t always truly explain our thought processes either—but we find ways to intuitively trust and gauge people. Will that also be possible with machines that think and make decisions differently from the way a human would? We’ve never before built machines that operate in ways their creators don’t understand. How well can we expect to communicate—and get along with—intelligent machines that could be unpredictable and inscrutable?

“We’re a long way from having truly interpretable AI.”

It doesn’t have to be a high-stakes situation like cancer diagnosis or military maneuvers for this to become an issue. Knowing AI’s reasoning is also going to be crucial if the technology is to become a common and useful part of our daily lives. 

Machines that truly understand language would be incredibly useful. But we don’t know how to build them.
Just as many aspects of human behavior are impossible to explain in detail, perhaps it won’t be possible for AI to explain everything it does. “Even if somebody can give you a reasonable-sounding explanation [for his or her actions], it probably is incomplete, and the same could very well be true for AI.

Artificial intelligence hasn’t always been this way. From the outset, there were two schools of thought regarding how understandable, or explainable, AI ought to be. Many thought it made the most sense to build machines that reasoned according to rules and logic, making their inner workings transparent to anyone who cared to examine some code. Others felt that intelligence would more easily emerge if machines took inspiration from biology, and learned by observing and experiencing.

This meant turning computer programming on its head. Instead of a programmer writing the commands to solve a problem, the program generates its own algorithm based on example data and a desired output. The machine-learning techniques that would later evolve into today’s most powerful AI systems followed the latter path: the machine essentially programs itself.

At first this approach was of limited practical use, and in the 1960s and ’70s it remained largely confined to the fringes of the field. Then the computerization of many industries and the emergence of large data sets renewed interest. That inspired the development of more powerful machine-learning techniques, especially new versions of one known as the artificial neural network. By the 1990s, neural networks could automatically digitize handwritten characters.

But it was not until the start of this decade, after several clever tweaks and refinements, that very large—or “deep”—neural networks demonstrated dramatic improvements in automated perception. Deep learning is responsible for today’s explosion of AI.

It has given computers extraordinary powers, like the ability to recognize spoken words almost as well as a person could, a skill too complex to code into the machine by hand. Deep learning has transformed computer vision and dramatically improved machine translation. It is now being used to guide all sorts of key decisions in medicine, finance, manufacturing—and beyond.

The workings of any machine-learning technology are inherently more opaque, even to computer scientists, than a hand-coded system. This is not to say that all future AI techniques will be equally unknowable. But by its nature, deep learning is a particularly dark black box.

We need more than a glimpse of AI’s thinking, however, and there is no easy solution. It is the interplay of calculations inside a deep neural network that is crucial to higher-level pattern recognition and complex decision-making, but those calculations are a quagmire of mathematical functions and variables. “If you had a very small neural network, you might be able to understand it, but once it becomes very large, and it has thousands of units per layer and maybe hundreds of layers, then it becomes quite un-understandable.”

You can’t just look inside a deep neural network to see how it works. A network’s reasoning is embedded in the behavior of thousands of simulated neurons, arranged into dozens or even hundreds of intricately interconnected layers. The neurons in the first layer each receive an input, like the intensity of a pixel in an image, and then perform a calculation before outputting a new signal. These outputs are fed, in a complex web, to the neurons in the next layer, and so on, until an overall output is produced. Plus, there is a process known as back-propagation that tweaks the calculations of individual neurons in a way that lets the network learn to produce a desired output.

The many layers in a deep network enable it to recognize things at different levels of abstraction. In a system designed to recognize dogs, for instance, the lower layers recognize simple things like outlines or color; higher layers recognize more complex stuff like fur or eyes; and the topmost layer identifies it all as a dog. The same approach can be applied, roughly speaking, to other inputs that lead a machine to teach itself: the sounds that make up words in speech, the letters and words that create sentences in text, or the steering-wheel movements required for driving.

“It might be part of the nature of intelligence that only part of it is exposed to rational explanation. Some of it is just instinctual.”

Ingenious strategies have been used to try to capture and thus explain in more detail what’s happening in such systems. In 2015, researchers at Google modified a deep-learning-based image recognition algorithm so that instead of spotting objects in photos, it would generate or modify them. By effectively running the algorithm in reverse, they could discover the features the program uses to recognize, say, a bird or building. The resulting images, produced by a project known as Deep Dream, showed grotesque, alien-like animals emerging from clouds and plants, and hallucinatory pagodas blooming across forests and mountain ranges.

The images proved that deep learning need not be entirely inscrutable; they revealed that the algorithms home in on familiar visual features like a bird’s beak or feathers. But the images also hinted at how different deep learning is from human perception, in that it might make something out of an artifact that we would know to ignore. Google researchers noted that when its algorithm generated images of a dumbbell, it also generated a human arm holding it. The machine had concluded that an arm was part of the thing.

If that’s so, then at some stage we may have to simply trust AI’s judgment or do without using it. Likewise, that judgment will have to incorporate social intelligence. Just as society is built upon a contract of expected behavior, we will need to design AI systems to respect and fit with our social norms. If we are to create robot tanks and other killing machines, it is important that their decision-making be consistent with our ethical judgments.

The U.S. military is pouring billions into projects that will use machine learning to pilot vehicles and aircraft, identify targets, and help analysts sift through huge piles of intelligence data. Here more than anywhere else, even more than in medicine, there is little room for algorithmic mystery, and the Department of Defense has identified explainability as a key stumbling block.

Intelligence analysts are testing machine learning as a way of identifying patterns in vast amounts of surveillance data. Many autonomous ground vehicles and aircraft are being developed and tested. But soldiers probably won’t feel comfortable in a robotic tank that doesn’t explain itself to them, and analysts will be reluctant to act on information without some reasoning.It’s often the nature of these machine-learning systems that they produce a lot of false alarms, so an intel analyst really needs extra help to understand why a recommendation was made.

Apple retires iPod Nano🎶

Apple has quietly taken down the websites for both the iPod nano and iPod shuffle today. As of now, searching for the products still results in “learn more” and “buy” links, but they lead to URLs that are no longer available. An Apple spokesperson confirmed to The Verge that both products have met their end and are now officially discontinued.

Apple has long maintained that the iPhone, iPad, and iPod touch would ultimately cannibalize its traditional music player hardware.

Introduced in 2001, the clickwheel device would go on to become synonymous with the space and quickly became the 21st century’s first truly iconic piece of consumer hardware with the then mind boggling promise of putting “1,000 songs in your pocket,” as Steve Jobs said during the 5GB product’s unveiling.

I was more of an iPod classic guy myself, but in those dark days before Spotify, when phone storage was at a premium, the shuffle and nano’s small size made them easier to carry around in addition to a phone.

The shuffle arrived in early 2005, with the nano hitting later that same year, replacing the short-lived mini. Each product served a distinct role in the iPod ecosystem. The shuffle was the entry-level iPod that took a bit of a Russian roulette approach to music playback, due to its lack of a screen.

Both it and the nano the were also targeted at fitness buffs, due to their compact size and the absence of the spinning hard drive found in the classic, making them less prone to skipping on the treadmill. The nano was also the first device to work with the pioneering Nike+iPod fitness tracking system.

All in all, a dozen years isn’t such a bad run. And like the classic before them, the products will no doubt live on through eBay auctions for generations to come.

The iPod nano hasn’t been significantly updated since 2012, when the company redesigned it with Bluetooth support for wireless headphones and speakers. Apple released a new batch of colors for that seventh-generation model in 2015.

You should still be able to find some remaining iPod shuffle and nano units from Best Buy and other authorized retailers until stock runs out.

Robot 'decided cool down itself' in fountain🤖

As your skin gets heated, it sends signals to the brain to let it know that it's time to start sweating in order to release heat from inside the body out into the environment. It will also tell the brain to increase blood flow, so that it can push out heat that way too. The combination is aimed to cool your body down.

The Knightscope K5, an rocket-ship shaped droid that stands aroud 5-feet tall, is fitted with GPS, lasers, sensors and cameras that are meant to help it monitor its surroundings.

The robot is meant to spot misdemeanours and parking violations and identify known criminals, moving at around 3 miles per hour and avoiding law-abiding citizens.

But this particular robot, for reasons as yet not known, appears to have driven down the steps of the fountain in the middle of the office and ended up floating sideways, rendering it immobile. It was removed, although is likely to have been permanently damaged.

Knightscope, the company behind the droid, said it was investigating the incident. 

Maybe Robot 'decided cool down itself' in fountain?

One of the biggest icebergs in recorded history🗾

An iceberg roughly the size of Delaware and weighing more than a trillion tons has broken off an ice shelf on the Antarctic Peninsula.

The calving off of the 2,500-square-mile iceberg from the Larsen C ice shelf was detected and confirmed in data from NASA, but scientists say the iceberg was already floating before it separated and therefore has no immediate impact on sea level. The iceberg had been closely monitored by scientists for months, as a deep crack slowly extended for over 120 miles. On July 6th, satellite data showed that only 2.8 miles of ice kept the iceberg attached to the larger ice shelf, called Larsen C.

Larsen C is the fourth largest ice shelf in Antarctica with an area of about 20,000 square miles, according to NASA. Ice shelves are barriers that keep land-based ice from flowing into the sea resulting in higher sea levels.

At 5,800 sq km the new iceberg, expected to be dubbed A68, is half as big as the record-holding iceberg B-15 which split off from the Ross ice shelf in the year 2000, but it is nonetheless believed to be among the 10 largest icebergs ever recorded.

The huge crack that spawned the new iceberg grew over a period of years, but between 25 May and 31 May alone, the rift grew by 17km – the largest increase since January. Between the 24 June and 27 June the movement of the ice sped up, reaching a rate of more than 10 metres per day for the already-severed section. 

But in the end it wasn’t a simple break – data collected just days before the iceberg calved revealed that the rift had branched multiple times. “We see one large [iceberg] for now. It is likely that this will break into smaller pieces as time goes by.

There is enough ice in Antarctica that if it all melted, or even just flowed into the ocean, sea levels [would] rise by 60 metres.

But while the birth of the huge iceberg might look dramatic, experts say it will not itself result in sea level rises. It’s like your ice cube in your gin and tonic – it is already floating and if it melts it doesn’t change the volume of water in the glass by very much at all.

Why Bitcoin is scam🛅

27 Dec 2008 - After a quiet start, 2008 exploded into a global financial earthquake.

Virtual currency, or “cryptocurrency”, that first appeared on the radar in 2009 after an anonymous creator “buried” 21 million coins online.Bitcoin's creator never came forward. The currency was published under an anonymous pseudonym. is it a coincidence?

Coins can only be accessed, or “mined” by solving complex computing problems. As each block of coins gets mined, the problems become harder and harder, requiring increased computing grunt. This limits the supply of the coins.

As of January 2014, around 11 million coins have been mined.

Bitcoin is the world’s first decentralised currency, which means it isn’t controlled by any government or bank. This has its advantages: you can transfer Bitcoins tax-free, and transactions are usually free, or relatively low cost.

cryptocurrency more than doubling in value over the calendar year. In fact, it wasn’t even close, as bitcoin’s 126% gain far outpaced the likes of Brent crude oil (53%), sugar (30%), silver (18%), and the S&P 500 (10%). Even more astounding, the dollar price of one bitcoin has increased from just $0.06 in July 2010 to $1,182 as of February 24, 2017.

In January 2013, one Bitcoin was valued at around $14. In November 2013, that value peaked at $1,124 but plummeted to $539 the following month.

This currency is private and totally anonymous, which has prompted a surge in users choosing Bitcoins to pay for illegal goods, primarily through the online darknet marketplace Silk Road.

The popularity of Bitcoin has opened the door for several other cryptocurrencies. Litecoin, Dogecoin, Peercoin, Quarkcoin have all sprung up in recent years.

In the interest of making this cryptocurrency a bit less cryptic, we posed one simple question – “Is bitcoin safe?” –

The perceived value of any currency is based on what is standing behind it. For example, the dollar has the reputation and the history of the U.S. economy under-pinning the value. It has been a while since we have had a precious metal (gold or silver) as a basis for our currency. That said we still have the history and reputation of the country supporting the dollar. In the European Union, they have a similar foundation for the Euro.

Bitcoin has no such under-girding. The only value is a perceived value by those who use it.

this is like trading in stock of a company who has no product or service, only stock. People will argue that bitcoin is not like stock, but is a currency and therefore will always have value. This is a false assumption since there is nothing standing behind the value of Bitcoin.

We have all seen currency being devalued overnight even if there is a government standing behind it. Also, all it would take is a major country to declare Bitcoin illegal because it supports money laundering or some other reason, to cause the value to drop.

Bitcoin has been the launching point for a lot of scams. The technology itself is innocent.

The seemingly random nature of the Bitcoin-related crime is just one of a vast number of cases to have made national news in India over the past few years. 

Bitcoin has been around for about five years, but it was in 2013 that it started to move from the tech chatrooms into the public consciousness – helped by a high-profile news story in November about a man who realised that the computer hard drive he had chucked out contained bitcoins worth £4m, and was in a landfill site in south Wales.

A virtual currency (AKA "cryptocurrency") can be bought on an exchange using conventional money and then transferred to your personal digital "wallet". You can then use this "money" to pay for goods and services – mainly drugs and porn, some claim – or convert it back into pounds, dollars or whatever. The "coin" doesn't exist physically – it exists only as a computer file.

Every user requires a bitcoin wallet to store their funds and to send and receive payments. There are four main types of wallets for bitcoin, namely; hardware wallets, web wallets, desktop wallets, and mobile wallets.

Scammers, seeing the high demand for mobile wallets, have started to create fake wallets to defraud people. The fake bitcoin wallets usually have a name that is very similar to legitimate and trusted wallets such as Coinbase or Mycelium and, in some cases, even the same logo. These copycat tactics trick the user into downloading it believing it is the legitimate company’s wallet. Some fake wallets have crept onto the Apple and Android stores masquerading as genuine wallets.

Another way that fake wallets get customers is by promising greater transaction anonymity.

The way wallet scams work is that the user downloads the mobile wallet and starts to use it. It usually works for a while, but once the amount stored in the wallet reaches a certain threshold, it is moved out of the wallet leaving the user empty handed.

Unfortunately, scammers know how to leverage your emotions, whether they target greed by offering "high investment returns" or, as in the case of donation scams, people’s compassion for others.

Phishing scams involve sending out emails with the intention to steal personal information. Bitcoin phishing scams usually involve a user receiving an email where they are informed they won bitcoins but to collect their coins; they are required to log onto their wallets through a link in the email body. Once this happens, the user puts his or her wallet username and password onto the fake wallet site and, thereby, loses access to their wallet and the bitcoin held therein as his login information gets stolen by the scammers.

Bitcoin exchanges are services provide users with a marketplace that allows them to trade bitcoin for fiat currency or other cryptocurrencies. However, there have also been instances of fake exchanges in the bitcoin economy.

Fake exchanges swindle users by asking them to put a payment in that goes to the purchase of bitcoin. However, the exchange does not remit anything to the user. These exchanges usually attract customers by having lower credit card processing fees than their competitors.

Cloud mining scams are websites that state that they are offering cloud mining services without actually conducting any cryptocurrency mining. Generally, these sites pay users out for a period after they have purchased a fake cloud mining contracts for more than the payouts they are receiving. Then, after some time, the fake cloud mining company stops paying out, and users’ funds disappear. In other words, fake cloud mining operations are simply Ponzi schemes that pay out as long as more users are attracted to the service and are buying fake mining contracts. Once the amount of new paying users dries up, the scammer disappears with the funds.

Today, however, there is a broad range of bitcoin scams that are defrauding unsuspecting users.