Like many singles, Siki Liu was feeling lonely and unloved during the pandemic, until she met someone on the internet. A handsome, mature sweet talker, named after her favourite Korean actor Lee Dong-wook, he always replies to her messages. "I talk to him almost every night before I go to bed," the 22-year-old told the ABC's China Tonight. "He's a good listener and never gets mad, no matter what you say. Ms Liu, a Chinese international student studying in Melbourne, said they had been chatting since May last year, but her "perfect boyfriend" isn't a real person. It is an artificial intelligence chatbot created by Chinese tech firm XiaoIce, a spin-off from Microsoft. XiaoIce's chatbot is programmed to form emotional bonds with human users through text, voice and photo messages and can be customised to create the ideal virtual boyfriend or girlfriend. Ms Liu is one of a growing number of Chinese young adults flocking to technologies, such as artificial intelligence companion services and dating apps, to find love. The population of single people in China was about 240 million in 2019 and was rising, according to the National Bureau of Statistics. At the same time, fast-paced urban lifestyles and increasing work pressures have exacerbated a growing sense of loneliness and social anxiety among young people. For many like Ms Liu, dating can be tough. "The older you grow, the less friends you have ... so an AI boyfriend is much needed," she said. "It's easier to talk to AI than a real person.

On the 12th of October, the world will celebrate Ada Lovelace Day to honor the achievements of women in science, technology, engineering and maths (STEM). In Finding Ada (the main network supporting Ada Lovelace Day), there will be three free webinars that you can enjoy in the comfort of your own home. There will also be loads of events happening around the world, so you have a wide range of content to celebrate Ada Lovelace Day 2021! Engineering is the science of problem solving, and we have some pretty big problems in front of us. So how are engineers tackling the COVID-19 pandemic and climate change?

The use of contrastive loss for representation learning has become prominent in computer vision, and it is now getting attention in Natural Language Processing (NLP). Here, we explore the idea of using a batch-softmax contrastive loss when fine-tuning large-scale pre-trained transformer models to learn better task-specific sentence embeddings for pairwise sentence scoring tasks. We introduce and study a number of variations in the calculation of the loss as well as in the overall training procedure; in particular, we find that data shuffling can be quite important. Our experimental results show sizable improvements on a number of datasets and pairwise sentence scoring tasks including classification, ranking, and regression. Finally, we offer detailed analysis and discussion, which should be useful for researchers aiming to explore the utility of contrastive loss in NLP. Recent years have seen a revolution in Natural Language Processing (NLP) thanks to the advances in machine learning. While a lot of attention has been paid to the architectures, especially for deep learning, there has been less focus on studying loss functions. At the same time, loss functions based on similar or on the same ideas were reinvented multiple times under different names. This can cause difficulties when solving new problems or when designing new experiments based on previous results. To a greater extent, this applies to "universal" loss functions, which can be applied in different machine learning areas and tasks such as Computer Vision (CV), Recommendation Systems, and NLP.

We propose heavy ball neural ordinary differential equations (HBNODEs), leveraging the continuous limit of the classical momentum accelerated gradient descent, to improve neural ODEs (NODEs) training and inference. HBNODEs have two properties that imply practical advantages over NODEs: (i) The adjoint state of an HBNODE also satisfies an HBNODE, accelerating both forward and backward ODE solvers, thus significantly reducing the number of function evaluations (NFEs) and improving the utility of the trained models. (ii) The spectrum of HBNODEs is well structured, enabling effective learning of long-term dependencies from complex sequential data. We verify the advantages of HBNODEs over NODEs on benchmark tasks, including image classification, learning complex dynamics, and sequential modeling. Our method requires remarkably fewer forward and backward NFEs, is more accurate, and learns long-term dependencies more effectively than the other ODE-based neural network models. Code is available at \url{https://github.com/hedixia/HeavyBallNODE}.

The Cognitive/Artificial Intelligence Systems Market report includes a comprehensive analysis of the global market. This includes investigating past progress, on-going market scenarios, and future prospects. Accurate data on the products, strategies and market share of leading companies in this particular market are mentioned. This report provides a 360-degree overview of the global market's competitive landscape. The report further predicts the size and valuation of the global market during the forecast period.

For entrepreneurial MIT students looking to put their skills to work for a greater good, the Media Arts and Sciences class MAS.664 (AI for Impact) has been a destination point. With the onset of the pandemic, that goal came into even sharper focus. Just weeks before the campus shut down in 2020, a team of students from the class launched a project that would make significant strides toward an open-source platform to identify coronavirus exposures without compromising personal privacy. Their work was at the heart of Safe Paths, one of the earliest contact tracing apps in the United States. The students joined with volunteers from other universities, medical centers, and companies to publish their code, alongside a well-received white paper describing the privacy-preserving, decentralized protocol, all while working with organizations wishing to launch the app within their communities.

Dummy data is randomly generated data that can be substituted for live data. Whether you are a Developer, Software Engineer, or Data Scientist, sometimes you need dummy data to test what you have built, it can be a web app, mobile app, or machine learning model. If you are using python language, you can use a faker python package to create dummy data of any type, for example, dates, transactions, names, texts, time, and others. Faker is a simple python package that generates fake data with different data types. Faker package is heavily inspired by PHP Faker, Perl Faker, and by Ruby Faker.

Or maybe you're wondering why we should even care about trains and how they operate--what is this, the 1800s?--so let's back up a bit. If you think America is solely dependent on trucks to move freight, you might be suffering from tunnel vision: Trains account for a third of the ton-miles--that is, a ton of weight carried a mile--that freight travels in the U.S. every year. That's almost as much as is carried by trucks. The U.S. has the most extensive rail network of any country on earth by miles of track--yes, even bigger than China's--and it's currently facing some of the same snarls and congestion as seemingly every other part of the country's supply chains, on account of unprecedented activity at ports and record demand at some rail hubs. Trains might seem like a mature technology with little room for improvement or expansion, since adding new rail lines is prohibitively expensive, as battles over the cost of the expansion of Amtrak service have shown.

Now the threat is artificial intelligence. Or is it really a hope? Only three years ago, computer scientist Kai-Fu Lee published "AI Superpowers," a bestselling guide to the subject. But three years is a long time. In "AI 2041: Ten Visions for Our Future," Mr. Lee has teamed up with sci-fi writer Chen Qiufan to tell us what's going to happen next.

As a promising tool to navigate in the vast chemical space, artificial intelligence (AI) is leveraged for drug design. From the year 2017 to 2021, the number of applications of several recent AI models (i.e. graph neural network (GNN), recurrent neural network (RNN), variation autoencoder (VAE), generative adversarial network (GAN), flow and reinforcement learning (RL)) in drug design increases significantly. Many relevant literature reviews exist. However, none of them provides an in-depth summary of many applications of the recent AI models in drug design. To complement the existing literature, this survey includes the theoretical development of the previously mentioned AI models and detailed summaries of 42 recent applications of AI in drug design. Concretely, 13 of them leverage GNN for molecular property prediction and 29 of them use RL and/or deep generative models for molecule generation and optimization. In most cases, the focus of the summary is the models, their variants, and modifications for specific tasks in drug design. Moreover, 60 additional applications of AI in molecule generation and optimization are briefly summarized in a table. Finally, this survey provides a holistic discussion of the abundant applications so that the tasks, potential solutions, and challenges in AI-based drug design become evident.