Mediaboss Marketing

After describing various ways in which AI and big data are involved already in our daily lives, ranging from the food we eat, the cars we drive and the things we buy, he concludes that it is leading to the Fourth Industrial Revolution, a phrase coined by Klaus Schwab, the head of the World Economic Forum. All aspects of life will be transformed in a way analogous to the prior industrial revolutions (first the use of steam and waterpower, second the expansion of electricity and telegraph cables, and third, the digital revolution of the end of the 20th century).

At the heart of the book are the chapters in which he explains what data and AI have already accomplished for our health and what they can do in the future. The ever-expanding amount of personal data available combined with advances in AI allows for increasing accuracy of diagnoses, treatments and better sensors and software. Glorikian notes that today there are over 350,000 different healthcare apps and the mobile health market is expected to approach $290 billion in revenue by 2025.

Glorikian employs a light, informal style of writing, with references to pop culture such as Star Trek. He asks the reader questions and intersperses each chapter with what he calls sidebars. They are short illustrative stories or sets of examples. For example, AI Saved My Life: The Watch That Called 911 for a Fallen Cyclist (p. 68) starts with a man who lost consciousness after falling off his bike, and then lists other ways current phones can save lives. Other sidebars explain basic concepts like the meaning of genes and DNA; or about gene editing with CRISPR.

Present and Future Advances

Before getting into more complex issues, Glorikian describes what be most familiar to readers: the use of AI-enabled smartphone apps which guide individuals towards optimal diets and exercising as well as allow for group activities through remote communication and virtual reality. There are already countless AI-enabled smartphone apps and sensors allowing us to track our movements and exercise, as well as our diets, sleep and even stress levels. In the future, their approach will become more tailored to individual needs and data, including genomics, environment, lifestyle and molecular biology, with specific recommendations.

He speculates as to what innovations the near future may bring, remarking: What isnt clear is just how long it will take us to move from this point of collecting and finding patterns in the data, to one where we (and our healthcare providers are actively using those patterns to make accurate predications about our health. He gives the example of having an app to track migraine headaches, which can find and analyze patterns in the data (do they occur on nights when you have eaten a particular kind of food or traveled on a plane, for example). Eventually, at a more advanced stage, it might suggest you take an earlier flight or eat in a different restaurant that does not use ingredients that might be migraine triggers for you.

Healthcare will become more decentralized, Glorikian predicts, with people no longer forced to wait hours in hospital emergency rooms. Instead, some issues can be determined through phone apps and remote specialists, and others can be handled at rapid care facilities or pharmacies. Hospitals themselves will become more efficient with command centers monitoring the usage of various resources and using AI to monitor various aspects of patient health. Telerobotics will allow access to specialized surgeons located in major urban centers even if there are none in the local hospital.

In the chapter on genetics, Glorikian presents three ways in which unlocking the secrets of an individuals genome can have practical health consequences right now. The first is the prevention of bad drug reactions through pharmacogenomics, or learning how genes affect response to drugs. Second are enhanced screening and preventative treatment for hereditary cancer syndromes. One major advancement just starting to be used more, notes Glorikian, is liquid biopsy, in which a blood sample allows identification of tumor cells as opposed to standard physical biopsies. It is less invasive and sometimes more accurate for detecting cancers prior to the appearance of symptoms. The third way is DNA sequencing at birth to screen for many disorders which are treatable when caught early. The future may see corrections of various mutations through gene editing.

He points out the various benefits in the health field of collecting large sets of data. For example, it allows the use of AI or machine learning to better read mammogram results and to better predict which patients would see benefit from various procedures like cardiac resynchronization therapy or who had greater risk for cardiovascular disease. There is hope that this approach can help detect the start and the progression of diseases like Alzheimers or diabetic retinopathy. Ultimately it may even be able to predict fairly reliably when individuals would die.

At present, AI accessing sufficient data is helping identify new drugs, saving time and money by using statistical models to predict whether the new drugs will work even before trials. AI can determine which variables or dimensions to remove when making complex computations of models in order to speed up computational processes. This is important when there are large numbers of variables and vast amounts of data.

Glorikian does not miss the opportunity to use the current Covid-19 crisis as a teaching moment. In a chapter called Solving the Pandemic Problem, Glorikian discusses the role AI, machine learning and big data played in the fight against the coronavirus pandemic, in spotting it early on, predicting where it might travel next, sequencing its genome in days, and developing diagnostic tests, vaccines and treatments. Vaccine development, like drug development, is much faster today than even 20 years ago, thanks to computational modeling and virtual clinical trials and studies.

Potential Problems

Glorikian does not shy away from raising some of the potential problems associated with the wide use of AI in medicine, such as the threat to patient privacy and ethical questions about what machines should be allowed to do. Should genetic editing be allowed in humans for looks, intelligence or various types of talents? Should AI predictions of lifespan and dates of death be used? What types of decisions should machines be allowed to make in healthcare? And what sort of triage should be allowed in case of limited medical resources (if AI predicts one patient is for example ten times more likely to die than another despite medical intervention)? There are grave dangers if hackers access databanks or medical machines.

There are also potential operational problems with using data as a basis for AI, such as outdated information, biased data, missing data (and how it is handled), misanalyzed or differently analyzed data.

Despite all these issues, Glorikian is optimistic about the value of AI. He concludes, But despite the risk, for the most part, the benefits outweigh the potential downsidesThe data we willingly give up makes our lives better.

Armenian Connection

When asked at the end of June, 2022 how Armenia compares with the US and other parts of the world in the use of AI in healthcare, he made the distinction between the Armenian healthcare system and Armenian technology that is directed at the world healthcare system.

On the one hand, he said, I dont know of a lot that is being incorporated into the healthcare system, although we do have a national electronic medical record system that they have really been improving on a consistent basis. Having such a health record system throughout the country will provide data for the next step in use of AI, and that, he said is very exciting.

On the other hand, for technology companies involved in healthcare and biotechnology in Armenia, he said, I would always like to see more, but there are some really interesting companies that have sprouted up over the last five years. Also, with the tech giant NVDIA opening up a research center in Armenia, Glorikian said he hoped there will be interesting synergies since this company does invest in the healthcare area.Harry Glorikian, second from left, next to Acting Prime Minister Nikol Pashinyan, in a December 19, 2018 Yerevan meeting

At the end of 2018, Glorikian met with then Acting Prime Minister Nikol Pashinyan to discuss launching the Armenian Genome project to expand the scope of genetic studies in the field of healthcare. He said that this undertaking was halted for reasons beyond his understanding. He said, My lesson learned was you can move a lot faster and have significant impact by focusing on the private sector.

Indeed, this is what he does, as an individual investor, although he finds investing as a general partner of a fund more impactful. He is also a member of the Angel Investor Club of Armenia. While the group looks at a broad range of companies, mainly technology driven, he and a few other people in it take a look at those which are involved in healthcare. In fact, he is going to California at the very end of June to learn more about a robot companion for children called Moxie, prepared by Embodied, Inc., a company founded by veteran roboticist Paolo Pirjanian. Pirjanian, who was a guest on Glorikians podcast several weeks ago, lives in California, but Glorikian said that the back end of his companys work is done in Armenia.

Glorikian added that he is always finding out about or running into Armenians in the diaspora doing work with AI.

Changes

When asked what has changed since the publication of the book last year, he replied, Things are getting better! While hardware does not change overnight, he said that there have been incremental improvements to software during the period of time it took to write the book and then have it published. He said, For someone reading the book now, you are probably saying, I had no idea that this was even available. For someone like me, you already feel a little behind.

Readers of the book have already begun to contact Glorikian with anecdotes about what it led them to find out and do. He hopes the book will continue to reach more people. He said, The biggest thing I get out of it is when someone says I learned this and I did something about it. When individuals have access to more quantifiable data, not only can they manage their own health better, but they also provide their doctors with more data longitudinally that helps the doctor to be more effective. Glorikian said this should have a corollary effect of deflating healthcare costs in the long run.

One minor criticism of the book, at least of the paperback version that fell into the hands of this reviewer, is the poor quality of some of the images used. The text which is part of those illustrations is very hard to read. Otherwise, this is a very accessible read for an audience of varying backgrounds seeking basic information on the ongoing transformations in healthcare through AI.

Read the original post:
Glorikian's New Book Sheds Light on Artificial Intelligence Advances in the Healthcare Field - The Armenian Mirror-Spectator

Explanation methods that help users understand and trust machine-learning models often describe how much certain features used in the model contribute to its prediction. For example, if a model predicts a patients risk of developing cardiac disease, a physician might want to know how strongly the patients heart rate data influences that prediction.

But if those features are so complex or convoluted that the user cant understand them, does the explanation method do any good?

MIT researchers are striving to improve the interpretability of features so decision makers will be more comfortable using the outputs of machine-learning models. Drawing on years of field work, they developed a taxonomy to help developers craft features that will be easier for their target audience to understand.

We found that out in the real world, even though we were using state-of-the-art ways of explaining machine-learning models, there is still a lot of confusion stemming from the features, not from the model itself, says Alexandra Zytek, an electrical engineering and computer science PhD student and lead author of a paper introducing the taxonomy.

To build the taxonomy, the researchers defined properties that make features interpretable for five types of users, from artificial intelligence experts to the people affected by a machine-learning models prediction. They also offer instructions for how model creators can transform features into formats that will be easier for a layperson to comprehend.

They hope their work will inspire model builders to consider using interpretable features from the beginning of the development process, rather than trying to work backward and focus on explainability after the fact.

MIT co-authors include Dongyu Liu, a postdoc; visiting professor Laure Berti-quille, research director at IRD; and senior author Kalyan Veeramachaneni, principal research scientist in the Laboratory for Information and Decision Systems (LIDS) and leader of the Data to AI group. They are joined by Ignacio Arnaldo, a principal data scientist at Corelight. The research is published in the June edition of the Association for Computing Machinery Special Interest Group on Knowledge Discovery and Data Minings peer-reviewed Explorations Newsletter.

Real-world lessons

Features are input variables that are fed to machine-learning models; they are usually drawn from the columns in a dataset. Data scientists typically select and handcraft features for the model, and they mainly focus on ensuring features are developed to improve model accuracy, not on whether a decision-maker can understand them, Veeramachaneni explains.

For several years, he and his team have worked with decision makers to identify machine-learning usability challenges. These domain experts, most of whom lack machine-learning knowledge, often dont trust models because they dont understand the features that influence predictions.

For one project, they partnered with clinicians in a hospital ICU who used machine learning to predict the risk a patient will face complications after cardiac surgery. Some features were presented as aggregated values, like the trend of a patients heart rate over time. While features coded this way were model ready (the model could process the data), clinicians didnt understand how they were computed. They would rather see how these aggregated features relate to original values, so they could identify anomalies in a patients heart rate, Liu says.

By contrast, a group of learning scientists preferred features that were aggregated. Instead of having a feature like number of posts a student made on discussion forums they would rather have related features grouped together and labeled with terms they understood, like participation.

With interpretability, one size doesnt fit all. When you go from area to area, there are different needs. And interpretability itself has many levels, Veeramachaneni says.

The idea that one size doesnt fit all is key to the researchers taxonomy. They define properties that can make features more or less interpretable for different decision makers and outline which properties are likely most important to specific users.

For instance, machine-learning developers might focus on having features that are compatible with the model and predictive, meaning they are expected to improve the models performance.

On the other hand, decision makers with no machine-learning experience might be better served by features that are human-worded, meaning they are described in a way that is natural for users, and understandable, meaning they refer to real-world metrics users can reason about.

The taxonomy says, if you are making interpretable features, to what level are they interpretable? You may not need all levels, depending on the type of domain experts you are working with, Zytek says.

Putting interpretability first

The researchers also outline feature engineering techniques a developer can employ to make features more interpretable for a specific audience.

Feature engineering is a process in which data scientists transform data into a format machine-learning models can process, using techniques like aggregating data or normalizing values. Most models also cant process categorical data unless they are converted to a numerical code. These transformations are often nearly impossible for laypeople to unpack.

Creating interpretable features might involve undoing some of that encoding, Zytek says. For instance, a common feature engineering technique organizes spans of data so they all contain the same number of years. To make these features more interpretable, one could group age ranges using human terms, like infant, toddler, child, and teen. Or rather than using a transformed feature like average pulse rate, an interpretable feature might simply be the actual pulse rate data, Liu adds.

In a lot of domains, the tradeoff between interpretable features and model accuracy is actually very small. When we were working with child welfare screeners, for example, we retrained the model using only features that met our definitions for interpretability, and the performance decrease was almost negligible, Zytek says.

Building off this work, the researchers are developing a system that enables a model developer to handle complicated feature transformations in a more efficient manner, to create human-centered explanations for machine-learning models. This new system will also convert algorithms designed to explain model-ready datasets into formats that can be understood by decision makers.

Read more here:
Building explainability into the components of machine-learning models - MIT News

CHANDLER, Ariz. Microchip Technology Inc. in Chandler, Ariz., is introducing the SAMA7G54 Arm Cortex A7-based microprocessor that runs as fast as 1 GHz for low-power stereo vision applications with accurate depth perception.

The SAMA7G54 includes a MIPI CSI-2 camera interface and a traditional parallel camera interface for high-performing yet low-power artificial intelligence (AI) solutions that can be deployed at the edge, where power consumption is at a premium.

AI solutions often require advanced imaging and audio capabilities which typically are found only on multi-core microprocessors that also consume much more power.

When coupled with Microchip's MCP16502 Power Management IC (PMIC), this microprocessor enables embedded designers to fine-tune their applications for best power consumption vs. performance, while also optimizing for low overall system cost.

Related: Embedded computing sensor and signal processing meets the SWaP test

The MCP16502 is supported by Microchip's mainline Linux distribution for the SAMA7G54, allowing for easy entry and exit from available low-power modes, as well as support for dynamic voltage and frequency scaling.

For audio applications, the device has audio features such as four I2S digital audio ports, an eight-microphone array interface, an S/PDIF transmitter and receiver, as well as a stereo four-channel audio sample rate converter. It has several microphone inputs for source localization for smart speaker or video conferencing systems.

The SAMA7G54 also integrates Arm TrustZone technology with secure boot, and secure key storage and cryptography with acceleration. The SAMA7G54-EK Evaluation Kit (CPN: EV21H18A) features connectors and expansion headers for easy customization and quick access to embedded features.

For more information contact Microchip online at http://www.microchipdirect.com.

Read the original here:
Arm Cortex microprocessor for artificial intelligence (AI), imaging, and audio introduced by Microchip - Military & Aerospace Electronics

Children killed in Russian missile strike on Ukrainian apartment

An apartment bombing in Odesa is the second mass civilian casualty missile strike on Ukraine this week.

Cody Godwin, USA TODAY

The cost of rebuilding battered Ukraine after the war is estimated at a staggering $750 billion, but some of those funds could come from the source of the damage.

Just as he has appealed to the international community for help in his country's attempt to fend off the Russian invasion, Ukraine President Volodymyr Zelenskyy toldthe Ukraine Recovery Conference in Switzerland a global effort will be needed for restoration.

The reconstruction of Ukraine is not a local project, is not a project of one nation, but a common task of the entire democratic world all countries, all countries who can say they are civilized, Zelenskyy said in a video message. Restoring Ukraine means restoring the principles of life, restoring the space of life, restoring everything that makes humans humans.

Ukrainian Prime Minister Denys Shmyhal, who attended the conference in Luganoin person, provided the $750 billion figure andpresented a recovery plan forimmediate and long-term needs.

Shmyhalalso said a large source offunding should be the confiscated assets of Russia and Russian oligarchs, which he said may currently amount to between $300 billion and $500 billion.

USA TODAY ON TELEGRAM: Join our Russia-Ukraine war channel to receive updates straight to your phone.

Latest developments

Ukrainian President Volodymyr Zelenskyy has thanked the International Olympic Committeefor supporting a ban on Russian athletes in most Olympics sports.Russia has an appeal hearing Tuesday challenging its ban from international soccer at the Court of Arbitration for Sport in Lausanne, Switzerland.

Pope Francis, who has condemned the "ferocity'' and "cruelty''of Russian troops in Ukraine, said he hopes to visit Moscow and Kyiv after his trip toCanada July 24-30.

Russian President Vladimir Putin declared victory in the battle for Ukraine's Luhansk province Monday and ordered rest for his troops before pushing on in the Kremlin's quest to take control of the entire Donbas industrial region.

"Military units that took part in active hostilities and achieved success and victory should rest, increase their combat capabilities, Putin said on state TV.

Russian Defense Minister Sergei Shoigu reported that Russian forces had taken control of Lysychansk, the last disputed major city in Luhansk. Earlier, Ukraine's military said it wasforced to withdraw in the face of Russia'sadvantage in artillery, aviation, ammunition and personnel. Continuing to hold out would lead to "fatal consequences" for its troops, the military said in a Facebook post.

"We just gotta keep on fighting," the post said. "Unfortunately, steel will and patriotism are not enough for success. Material and technical resources are needed."

Despite Russia's claims to the contrary, its invasion is still having "a devastating impact on Ukraine's agricultural sector,'' the British Defense Ministry saidin its latest intelligence assessment.

The ministry saidthe Russian blockade of the key port of Odesa in the Black Sea is severely limiting Ukraine's ability to export grain while harvest has begun.In addition, the war hasdisrupted the supply chain of seeds and fertilizer farmers use.

That combination will most likely shrink Ukraine's agricultural exports this year to 35% or less of what they were in 2021, the ministry said, pointing out that drastic reduction from a major wheat producer is contributing to the global food crisis.

Russia's increasing use of outdated weaponry in a number of deadly attacks may beevidence its military lacks more precise modern weapons, military analysts say.

Russian bombers have been using 1960s-era KH-class missiles, which were primarily designed to target aircraft carriers using a nuclear warhead and are not able to accurately strike ground targets, officials say. The weapons were used in twoattacks on a shopping center and apartment building last week, resulting in dozens of civilian casualties.

Russia continues to employ air-launched anti-ship missiles in a secondary land-attack role, likely because of dwindling stockpiles of more accurate modern weapons, the British defense ministry said on Twitter.

Both Russia and Ukraine have expended large amounts of weaponry ina grinding war of attritionfor the easternDonbas region.President JoeBidensaid last month the U.S. would provide Ukrainelonger-range precision rockets, but it's not clear yet how much difference they'll make.

Contributing: The Associated Press

Read more here:
Ukraine live updates: Reconstruction cost estimated at $750 billion - USA TODAY

Kyiv, Ukraine It wasnt a knock, it was loud banging at about 7:30 on a recent Saturday morning.

Taras opened the door of his two-bedroom apartment in Kreminna, a town in Ukraines southeastern Luhansk region that was taken over by Russia in late April, to see three gun-toting soldiers in camouflage.

Do you have a garage on the corner? the oldest of them, a redhead in his late 20s, asked Taras imperatively.

Without waiting for his answer, the soldier continued: Open it up.

He was talking about a group of three dozen garages built in the early 1980s, an area which had become an informal club, where men could have a drink, crack a joke and play backgammon or chess.

But to the Russian occupiers, the garages were a source of danger, a younger, less strict soldier told 53-year-old Taras on the way, and they needed to check each for arms and explosives.

They looked inside, checked the basement and left without saying a word, Taras, who requested his last name be withheld because he doesnt want to be shot, told Al Jazeera.

They only thing of interest they saw and took away was a three-litre jar with cucumbers that Tarass wife had pickled in vinegar and tomato juice.

Taras got lucky.

His neighbour had his sky-blue Lada Priora confiscated and was beaten and left bruised after he hesitated to hand over the car key for a split second.

On Monday, after the capture of the Luhansk region, media outlets in Russia aired interviews with residents of Lysychansk who thanked Moscow for liberating them and claimed Kyivs forces were inhumane.

But people Al Jazeera spoke to had rather different views.

They said Moscow appoints new officials from among Ukrainian turncoats or pro-Moscow separatists. Tens of thousands are deported to Russia, and those who remain are subjected to humiliation, torture, robbery or arbitrary, extrajudicial killing. And it is only in the areas that Moscow plans to rule directly that occupying forces and officials are instructed to treat locals with at least a shred of respect.

They dont treat us like humans. They say they came to liberate us from what? From our property? From our lives? Taras told Al Jazeera via a messaging app.

Liberation is the key word the Kremlin uses when describing what it calls the special operation in Ukraine.

In Kremlin-speak, Ukraine had to be liberated from its neo-Nazi regime, and the eastern and southern Ukrainian regions where the majority of the population speaks Russian needed a liberation from Ukrainian nationalists.

In reality, in the occupied areas of Ukraine, Russia pursues three different policies.

The first one is being implemented in places such as Kreminna in the Luhansk and Donetsk regions, known collectively as the Donbas, that had already been partially controlled by separatists since 2014, says Kyiv-based political analyst Aleksey Kushch.

They use the scorched earth tactic here, a big population is seen as an unnecessary social burden, he told Al Jazeera.

Moscow prefers to send younger residents of the Donbas to Russia to repopulate its regions with low birthrates, bad local economies, and excessive alcoholism and crime.

More than a million Ukrainians have been deported to Russia from the Donbas, including the city of Mariupol, Ukrainian officials said.

The restoration of plants and factories in occupied Donbas, Ukraines former industrial pillar, is of no interest to Moscow. Russia simply needs to declare the liberation of areas that would later become part of the separatist statelets the so-called peoples republics of Donetsk and Luhansk known as DPR and LNR that are fully dependent on Russia economically and politically, Kushch said.

A stark example of this strategy is the way Russia operates in Mariupol, a former industrial hub on the Sea of Azov that had a population of more than 400,000 before the war.

After merciless, incessant pummelling between late February and April, it is now home to tens of thousands, mostly the elderly who live without electricity, running water and healthcare.

They cook, look for firewood, collect water and live outdoors because their shelling-damaged apartment buildings may collapse any minute and bury them alive, said Petro Andryushchenko, an adviser to Mariupols mayor Vadym Boychenko, who left the city before Russias takeover.

The worst thing is that people are getting used to it. They compare [their living conditions] not to what was before the war but to [what happened] in February April. With their lives in the cold basements under fire, Andryushchenko said in a Telegram post in mid-June.

The second strategy is used in the areas Russia plans to hold on to directly namely, the southern regions of Kherson and Zaporizhia, and in parts of the northeastern Kharkiv region adjacent to the Russian border.

There are attempts to create loyalty they plan fictional referendums to declare their residents determination to join Russia, analyst Kushch said.

In Kherson, despite hundreds of alleged abductions of pro-Ukrainian activists, most in the area are being cajoled into submission with food handouts and the promises of tax breaks, higher pensions and other perks.

Even critics of their policies admit that their efforts are aimed at appeasing the masses.

They quietly, calmly help people. One can take as much flour, grain, sugar, all in sacks. If it wasnt for them, there would have been famine, Halyna, a pro-Kyiv resident of Kherson, told Al Jazeera.

Last Wednesday, Kremlin-appointed officials in Kherson said they were preparing a referendum to join Russia.

Meanwhile, a third strategy is being used in areas where Russia did not try to create loyalists and relied on terror and mass crimes towards civilians, Kushch said.

Ukrainian officials say that more than 1,000 people have been killed in the towns and villages northwest, north and northeast of Kyiv between late February and early April, after Moscow retreated from the area after realising it would not risk street fights to seize the capital.

Many civilians were reportedly tortured, raped and shot dead in the back of their heads.

Some were killed just for fun, said a survivor who was beaten and doused with diesel fuel in late March.

They said: Lets set him on fire and send [him] back to his people, Viktor, a resident of Bucha, where most of the killings had taken place, told Al Jazeera in early April.

He survived only because shelling from the Ukrainian side forced his tormentors into a bomb shelter while he managed to escape.

Another reason why atrocities were so widespread, cruel and arbitrary is because of the narrative on Kremlin-controlled television networks that has for years portrayed Ukrainians as neo-Nazis who approve of the alleged genocide of Russian-speaking residents of the Donbas.

Another survivor described the look on the faces of three Russian soldiers who stormed into her house in the village of Myrotske 40 kilometres (25 miles) northwest of Kyiv.

They seemed full of hatred to Ukraine since they had been born, child psychologist Rivil Kofman told Al Jazeera in mid-March.

Kofman and her son David managed to leave the village after hiding for days in their ice-cold basement, observing the duels between Russian tanks and Ukrainian artillery and witnessing the killing of their escaping neighbours in their cars.

Unsurprisingly, residents of Russia-occupied areas meet Ukrainian servicemen as their true liberators.

They cried, they hugged us, saying, Oh, my dearest ones, thank you, said Maksim Butkevych, a Ukrainian human rights advocate who volunteered to join the Ukrainian army, and took part in the battles to retake Kyiv suburbs.

One old-timer even offered me moonshine, and I had to tell him, Daddy, I am on duty! Butkevych, who was taken prisoner in the Donbas last week, told Al Jazeera in mid-May.

See the original post here:
What is life like in Russia-occupied areas of Ukraine? - Al Jazeera English