R&D is critical in the world of healthcare, but it is also tricky; progress is slow, and further progress is hampered by the way different researchers and drug developers work in silos, data is kept secret, locked away from the rest of the world. Can AI in healthcare come to the rescue? Vas Narasimhan, CEO at Novartis, recently warned of a problem finding new data. In an interview with Bloomberg, he said that this lack of data has in part caused his initial enthusiasm for AI to turn more cautious. This may yet prove to be the single biggest hurdle in applying AI in healthcare to help find cures to new diseases, extend life, and improve the quality of life.

In a particularly pungent case of victim-blaming, "hard-line Republicans and conservative commentators are mounting a whispering campaign against Jamal Khashoggi that is designed to protect President Trump from criticism of his handling of the dissident journalist's alleged murder by operatives of Saudi Arabia -- and support Trump's continued aversion to a forceful response to the oil-rich desert kingdom," The Washington Post reports, citing four GOP officials involved in the discussions. The campaign includes "a cadre of conservative House Republicans allied with Trump" who in recent days have been "privately exchanging articles from right-wing outlets that fuel suspicion of Khashoggi," a Post columnist and Saudi government critic, the Post says. Still, the murmurs have begun to "flare into public view" as conservative media organizations and personalities -- Rush Limbaugh, Front Page, Donald Trump Jr., and a sanitized version on Fox News, to name a few -- "have amplified the claims, which are aimed in part at protecting Trump as he works to preserve the U.S.-Saudi relationship and avoid confronting the Saudis on human rights." The main lines of attack -- pushed by pro-Saudi accounts on Twitter -- focus on and distort Khashoggi's association with the Muslim Brotherhood in his young and interactions as a journalist with late Al Qaeda leader Osama bin Laden in the 1980s and '90s. "The GOP officials declined to share the names of the lawmakers and others who are circulating information critical of Khashoggi," the Post explains, "because they said doing so would risk exposing them as sources."

President Trump's threat of "severe" repercussions if the Saudis are found to have killed journalist Jamal Khashoggi were met with more threats from the Saudi side, including from a media figure close to the royal family who warned of high oil prices and a shift of alliances.

The size of the AI sector in India is difficult to determine, given that a lot of AI applications are in intermediary phases of production. Globally, one popular means of measuring the size of AI sectors is by adding up private sector investment in AI start-ups. According to one estimate, total AI funding worldwide has increased from $862 million in 2012 to $6.4 billion in 2017.1 The Indian AI sector, too, has seen growth in this period, with a total of $150 million invested in more than 400 companies over the past five years.2 Most of these investments have come in the last two years, when investment nearly doubled from $44 million in 2016 to $77 million in 2017.3 In India, too, the government is spearheading investments in AI and other emerging technologies. In the latest budget, the government set aside $480 million for investment into emerging technologies including AI. This commitment could help put India on the map, as this outlay compares favorably to those of Australia, Canada, and the European Union.5

Concept cars look beautiful and futuristic, but why do manufacturers spend millions developing them if they're never going to make it into production? Take, for example, the DS X e-tense, a "2035 dream car" produced by the French luxury brand DS. Half open-topped sports car, half luxury saloon, its outlandish styling looks as though it has come straight from the pages of a superhero magazine. It is designed to show what the company thinks a hugely powerful, all-electric self-driving machine might actually be like. It bears little relation to anything the brand currently produces, but that is hardly the point.

The biggest problem, or opportunity, for the current and many successive governments, is, and would be this - how to provide gainful employment to the millions of Indians entering the labour market every month? That one question has several corollaries to it. What role would automation play in all of this? What kind of jobs would be the first victims of automation? Is it wrong to expect manufacturing sector to provide jobs at a large scale now?

The country's premier think tank Niti Aayog and Microsoft India have signed a partnership to deploy artificial intelligence based solutions across various sectors such as agriculture and healthcare. According to the agreement, Microsoft India will support Niti Aayog by combining the cloud, AI, research and its vertical expertise for new initiatives and solutions across several core areas including agriculture and healthcare and the environment. Microsoft will also accelerate the use of AI for the development and adoption of local language computing, in addition to building capacity for AI among the workforce through education, the company said in a statement. Amitabh Kant, CEO - NITI Aayog said that for our country, the power of Artificial Intelligence needs to be brought to bear in sectors like healthcare, education, environment and agriculture, which are important for the inclusive development of India. "Simultaneously, we should utilize the power of AI to build understanding of the different regional languages prevalent in India."

It is designed to plug in to Metacorps, an experimental web app for annotating metaphor. As Metacorps users annotate metaphors, the system will use user annotations as training data. When the system is confident, it will suggest an identification and an annotation. Once approved by the user, this becomes more training data. This naturally allows for transfer learning, where the system can, with some known degree of reliability, classify one class of metaphor after only being trained on another class of metaphor. For example, in our metaphorical violence project, metaphors may be classified by the network they were observed on, the grammatical subject or object of the violence metaphor, or the violent word used (hit, attack, beat, etc.).

Recently, due to the increasing popularity of social media, the necessity for extracting information from informal text types, such as microblog texts, has gained significant attention. In this study, we focused on the Named Entity Recognition (NER) problem on informal text types for Turkish. We utilized a semi-supervised learning approach based on neural networks. We applied a fast unsupervised method for learning continuous representations of words in vector space. We made use of these obtained word embeddings, together with language independent features that are engineered to work better on informal text types, for generating a Turkish NER system on microblog texts. We evaluated our Turkish NER system on Twitter messages and achieved better F-score performances than the published results of previously proposed NER systems on Turkish tweets. Since we did not employ any language dependent features, we believe that our method can be easily adapted to microblog texts in other morphologically rich languages.

Neural approximate computing gains enormous energy-efficiency at the cost of tolerable quality-loss. A neural approximator can map the input data to output while a classifier determines whether the input data are safe to approximate with quality guarantee. However, existing works cannot maximize the invocation of the approximator, resulting in limited speedup and energy saving. By exploring the mapping space of those target functions, in this paper, we observe a nonuniform distribution of the approximation error incurred by the same approximator. We thus propose a novel approximate computing architecture with a Multiclass-Classifier and Multiple Approximators (MCMA). These approximators have identical network topologies and thus can share the same hardware resource in a neural processing unit(NPU) clip. In the runtime, MCMA can swap in the invoked approximator by merely shipping the synapse weights from the on-chip memory to the buffers near MAC within a cycle. We also propose efficient co-training methods for such MCMA architecture. Experimental results show a more substantial invocation of MCMA as well as the gain of energy-efficiency.