AI Assistants: in the news: what have you missed?

The latest to succumb is Go's top-ranked player, Ke Jie, who lost 3-0 in a series hosted in China this week. The AI, developed by London-based DeepMind, which was acquired by Google for around $500 million in 2014, also overcome a team of five top players during a week of matches. AlphaGo first drew headlines last year when it beat former Go world champion Lee Sedol, and the China event took things to the next level with matches against 19-year-old Jie, and doubles with and against other top Go pros. Challengers defeated, AlphaGo has cast its last competitive stone, DeepMind CEO Demis Hassabis explained. This week's series of thrilling games with the world's best players, in the country where Go originated, has been the highest possible pinnacle for AlphaGo as a competitive program.

Kim pointed to convolutional neural networks, recurrent neural networks, and Long Short Term Memory (LSTM) networks, among others, each of which is designed to solve a specific problem, such as image recognition, speech or language translation. Norm Jouppi, a Google distinguished hardware engineer, unveiled details of the company's several-year effort, the Tensor Processor Unit (TPU), an ASIC that implements components of a neural network in silicon--as opposed to using raw silicon compute power and memory banks and software on top of that, which is something that Google also does. In discussing the Google TPU, Jouppi highlighted one way that teams of researchers and engineers around the world can benchmark their work and the performance of the hardware and software they are utilizing: ImageNet. Arnold Smeulders and Theo Gevers, the general chairs of ECCV 2016, told Semiconductor Engineering that many of the attendees of ECCV do work in the area of semiconductor technologies (as opposed to software that runs on silicon) that enable computer vision.

Facebook CEO Mark Zuckerberg, left, and Tesla and SpaceX CEO Elon Musk have started an online smackdown over the possible threat of artificial intelligence. Despite the public debate between Elon Musk and Mark Zuckerberg on whether or not artificial intelligence (AI) is good for humanity and on the heels of Facebook's news it had shut down its AI engine because chatbots had started talking to one another in their own language, AI is still in high-growth mode. Analyst firm Research and Markets predicts the global AI market to grow to $23.4 billion by 2025. In July 2017, UK startup, Graphcore, closed a $30 M investment round for Niklas Zennstrom's venture capital firm, Atomico. Zennstrom was one of the co-founders of Skype.

To explore what machine learning could mean in education, EdSurge convened a meetup this past week in San Francisco with Adam Blum (CEO of OpenEd), Armen Pischdotchian, (an academic technology mentor at IBM Watson), Kathy Benemann (CEO of EruditeAI), and Kirill Kireyev (founder of instaGrok and technology head at TextGenome and GYANT). As you shift from statistical evaluation models to deep machine learning [involving neural networks], what hasn't kept pace is "explainability." Now, say you have a neural network or some machine learning program that's better at predicting student outcomes. It's just another way to enable student learning and teacher practice.

Deep learning (DL) is a branch of machine learning that learns multiple levels of representations from data [1, 2]. DL has been successfully applied in pattern recognition and classification tasks such as image recognition and language processing. Recently, the application of DL to physics research is rapidly growing, such as in particle physics [3-7], nuclear physics [8], and condensed matter physics [9-14]. DL is shown to be very powerful in extracting pertinent features especially for complex nonlinear systems with high-order correlations that conventional techniques are unable to tackle. This suggests that it could be utilized to unveil hidden information from the highly implicit data of heavy-ion experiments.

Traditional Recurrent Neural Networks assume vectorized data as inputs. However many data from modern science and technology come in certain structures such as tensorial time series data. To apply the recurrent neural networks for this type of data, a vectorisation process is necessary, while such a vectorisation leads to the loss of the precise information of the spatial or longitudinal dimensions. In addition, such a vectorized data is not an optimum solution for learning the representation of the longitudinal data. In this paper, we propose a new variant of tensorial neural networks which directly take tensorial time series data as inputs. We call this new variant as Tensorial Recurrent Neural Network (TRNN). The proposed TRNN is based on tensor Tucker decomposition.

Using the previously developed concepts of semantic spacetime, I explore the interpretation of knowledge representations, and their structure, as a semantic system, within the framework of promise theory. By assigning interpretations to phenomena, from observers to observed, we may approach a simple description of knowledge-based functional systems, with direct practical utility. The focus is especially on the interpretation of concepts, associative knowledge, and context awareness. The inference seems to be that most if not all of these concepts emerge from purely semantic spacetime properties, which opens the possibility for a more generalized understanding of what constitutes a learning, or even `intelligent' system. Some key principles emerge for effective knowledge representation: 1) separation of spacetime scales, 2) the recurrence of four irreducible types of association, by which intent propagates: aggregation, causation, cooperation, and similarity, 3) the need for discrimination of identities (discrete), which is assisted by distinguishing timeline simultaneity from sequential events, and 4) the ability to learn (memory). It is at least plausible that emergent knowledge abstraction capabilities have their origin in basic spacetime structures. These notes present a unified view of mostly well-known results; they allow us to see information models, knowledge representations, machine learning, and semantic networking (transport and information base) in a common framework. The notion of `smart spaces' thus encompasses artificial systems as well as living systems, across many different scales, e.g. smart cities and organizations.

We introduce the task of Visual Dialog, which requires an AI agent to hold a meaningful dialog with humans in natural, conversational language about visual content. Specifically, given an image, a dialog history, and a question about the image, the agent has to ground the question in image, infer context from history, and answer the question accurately. Visual Dialog is disentangled enough from a specific downstream task so as to serve as a general test of machine intelligence, while being grounded in vision enough to allow objective evaluation of individual responses and benchmark progress. We develop a novel two-person chat data-collection protocol to curate a large-scale Visual Dialog dataset (VisDial). VisDial v0.9 has been released and contains 1 dialog with 10 question-answer pairs on ~120k images from COCO, with a total of ~1.2M dialog question-answer pairs. We introduce a family of neural encoder-decoder models for Visual Dialog with 3 encoders -- Late Fusion, Hierarchical Recurrent Encoder and Memory Network -- and 2 decoders (generative and discriminative), which outperform a number of sophisticated baselines. We propose a retrieval-based evaluation protocol for Visual Dialog where the AI agent is asked to sort a set of candidate answers and evaluated on metrics such as mean-reciprocal-rank of human response. We quantify gap between machine and human performance on the Visual Dialog task via human studies. Putting it all together, we demonstrate the first 'visual chatbot'! Our dataset, code, trained models and visual chatbot are available on https://visualdialog.org

Specifically, you can tell the robot what flavor you like, such as "chewy candy," "sweet chocolate" or "hard mint." The robot then processes your instructions via voice recognition and natural language processing, recommends a particular kind of candy and uses image recognition to recognize and select that recommendation. The entire demo is powered by deep-learning technology running on Cloud Machine Learning Engine (the fully-managed TensorFlow runtime from Google Cloud) and Cloud machine learning APIs. This demo is intended to serve as a microcosm of a real-world machine learning (ML) solution. For example, Kewpie, a major food manufacturer in Japan, used the same Google Cloud technology to build a successful Proof of Concept (PoC) for doing anomaly detection for diced potato in a factory.