Artificial-intelligence tech has been moving out of academic circles into real-world applications like screening out spam or making it look like Van Gogh painted your photo. But now it's also tapping our weakness for cute sea creatures threatened by extinction. The large marine animal is really hard to keep track of. Researchers are using drones to take aerial photos of the ocean, but detecting them in those photos is a challenging task...for humans. That's where Google's TensorFlow neural network software project comes in, developer advocate Josh Gordon explained in a story posted Wednesday to Google's machine learning blog.

BenevolentAI is a London start-up that specializes in artificial intelligence (AI); its BenevolentBio division, formerly Stratified Medical, applies AI to human health and biotech. Its baby is a technology that could speed up late-stage development of drugs and provide richer clinical data. Now, it will test it using clinical stage drug candidates licensed from Janssen. Although there are no details on the particulars of the agreement, BenevolentAI is confident that it can accelerate clinical development and begin Phase IIb trials in mid-2017. If everything works out well, the company will have exclusive rights to develop, manufacture and commercialize these candidates.

Musk's tweets come at a time when a group of hackers used a Distributed Denial of Service (DDoS) attack to wipe out part of the internet on Oct. 21. This resulted in many sites like PayPal, Twitter, Netflix, and Spotify being unreachable and offline for hours. Authorities are of the view that a person or a group of people orchestrated the attack. Musk said that in the future massive DDoS attacks will not require real human hackers to create a ruckus on the infrastructure that runs the internet. With the improvement of artificial intelligence, Musk said that hackers will misuse it to optimize their attacks.

There is much discussion these days about Lambda Architecture and its benefits for developing high performance analytic architectures. It offers a combination of a high performance, low latency ETL with a real-time layer, and a slower, more accurate, and flexible solution that runs in batch. As I work with it, I have learned to appreciate Cassandra's relative "immortality" and fit for such analytic systems. In a complex distributed system it's nice to know you have one component that you can rely on without much tending. Need to be highly available and regionally distributed?

Efforts to understand suicide risks can be roughly clustered into traits or states. Trait analyses focus on stable characteristics rooted in and measured using biological processes (Costanza et al., 2014; Le-Niculescu et al., 2013), whereas state analyses measure dynamic characteristics like verbal and nonverbal communication, termed "thought markers" (Pestian et al., 2015). Machine learning and natural language processing have successfully identified differences in retrospective suicide notes, newsgroups, and social media (Gomez, 2014; Huang, Goh, & Liew, 2007; Matykiewicz, Duch, & Pestian, 2009). Jashinsky et al. (2015) used multiple annotators to identify the risk of suicide from the keywords and phrases (interrater reliability .79) in geographically based tweets. Thompson, Poulin, and Bryan (2014) and Desmet (2014) used text-based signals to identify suicide risk that ranged from 60% to 90%.

I first got speaking to Jason via Twitter and then we met at a few networking events. We quickly realised we shared an overtly cynical attitude to the vacuous tripe that emerges from Silicon Valley's startup culture, and want to resist the adoption of that culture in the Northern Ireland tech scene. When it comes to big data and machine learning, I know no one more qualified than Jason. He wrote a book about machine learning which has helped me immensely in coming to grips with the topic, even though I can't even begin to understand the mathematics behind it all. Like myself, Jason is not native to Northern Ireland, but he's been here so long he might as well be part of the furniture.

There's huge interest in artificial intelligence these days and countless investors seeking to get their fingers on the AI investment pie. But while AI is hot today, the advent of super-intelligent, self-directed computers is really still years away. However, there's a tremendous amount of things that "can already be achieved with machinery today," says the Harvard Business Review. "And that's where forward-thinking managers should be focusing." Machine learning is the sub-field of computer science that "gives computers the ability to learn without being explicitly programmed."

Last week, Skift published the 2017 Digital Transformation Report, sponsored by Adobe and Epsilon. In the past decade, mobile devices have revolutionized the travel industry. Some forecasts now predict more than half of all travel purchases are made with mobile, and mobile is reshaping how every travel organization, from corporate travel firms to airports to hotels, interacts with travelers. But despite all the rapid shifts already caused by mobile, even bigger changes are on the horizon. That's because three emerging interfaces, including voice search, artificial intelligence and conversational messaging, are transforming how travelers will interact with travel brands on mobile in the future.

Artificial Intelligence (AI) is making progress in great strides, or at least it appears so! Almost no week passes by without some major announcements of new challenges solved by AI technology or new products powered by AI. Indeed many quantifiable factors attest an unprecedented level of activity: capital investments, number of academic papers, number of products involving AI technology, they all are on a steep rise in the past five years. Computers are already very capable at some specialized tasks that require reasoning and other abilities that we typically associate with intelligence. For example, computers can play a decent game of chess or can help us order our holiday photos. Despite this genuine progress, we are still a long way from human level intelligence because our best artificial intelligence systems are not general purpose. They cannot quickly adapt to novel tasks the way most humans can do.

In the previous blog post Using Keras and Deep Q-Network to Play FlappyBird we demonstrate using Deep Q-Network to play FlappyBird. However, a big limitation of Deep Q-Network is that the outputs/actions are discrete while the action like steering are continuous in car racing. An obvious approach to adapt DQN to continuous domains is to simply discretize the action space. However, we encounter the "curse of dimensionality" problem. For example, if you discretize the steering wheel from -90 to 90 degrees in 5 degrees each and acceleration from 0km to 300km in 5km each, your output combinations will be 36 steering states times 60 velocity states which equals to 2160 possible combinations. The situation will be worse when you want to build robots to perform something very specialized, such as brain surgery that requires fine control of actions and naive discretization will not able to achieve the required precision to do the operations.