The Role of AI in Enabling Social Behaviors and Interactions

Designing Safer Public Spaces

One of the very instinctive ideas of how AI could help us, is in performing tasks which may be essentially dangerous or harmful for humans. It shows a great potential for fraud detection, responding to natural disaster and traffic control.

Tangling together the technology and its potential for social use in the physical security sector, two MIT graduates developed an algorithm named Synapse. Built on a large dataset of labeled visuals, Synapse detects prohibited items, such as guns or knives, at the airport check-ins. It uses deep learning and computer vision systems to automate the security at the airports worldwide by integrating onto x-ray machines through a hardware add-on. This technology does not completely eliminate a job of an airport security worker, but rather acts as a handy assistant. Except from the airports, this technology is also promising to potentially expand in serving critical infrastructure and schools. Development and deployment of automated threat-detection software that seems to be on the forefront of the industry, potentially contributing to elimination of human error and daunting consequences.

Understanding Consumer Behavior

AI can detect underlying patterns in consumer shopping behavior from the products we purchase and the way in which we purchase. For instance, the increase of jewelry purchases before holidays can help companies in supply and demand logistics. And by knowing that pasta is often bought with a tomato sauce, they can promote a bundle or put these objects next to each other in a store to prompt shopping decisions. Additionally, AI can make more complex calculations such as car tire production based on the number of new cars being sold and lifetime of tires based on the weather conditions during the last year. However, machine learning is applied in retail not only for demand planning, but also in moderating User Experience in-store. For instance, monitoring human flow and density and managing the number of check-outs to be open for processing, or even running in-store ads to navigate consumers to certain departments. For internet-based products, AI may be used for tailoring market propositions based on the content that was previously consumed by users or, in some cases, even generated by them. For big companies, AI is a part of social listening in analyzing textual and visual data to recognize the context of brand usage with NLP. In the era of customer-centricity, spotting your brand where no words are used and no hashtags are tagged is an area for improvement.

Whichever is the task, in order to execute these models successfully, AI needs large datasets labeled by humans, who translate visual inputs into a language readable by computers.

Often, these algorithms are used to better understand a target audience, for marketing research and nudging. However, in some cases, this may also have an additional behavioral influence or benefit for the buyers. A pharmaceutical company Walgreens leverages AI to track the spread of viruses in the USA. The company aggregates data from antiviral medication purchases from thousands of locations to track the spread of the flu on the country-wide scale and releases this information in a form of an interactive map. The analytics suggests Walgreens how to better stock the flu medicine and their map, even though is specific for their pharmacy only and cannot be generalized, may help customers get a grasp of how common the flu is in their area and, potentially, take precautionary measures.

Keeping Up Mental Health

Research suggests that positive internet-based communication may lead to decreased isolation, increased self-disclosure and higher self-esteem. Following the demand and gradually lowering stigma around mental health conversations, companies worldwide invest into and develop applications focusing exactly on this topic. Often presented as chatbots, journals or mood trackers, they enact like a first-step help for intense emotions and anxiety or assist in observation and documentation of thoughts and moods, while simultaneously providing privacy and anonymity. Naturally, mental health chatbots and therapy apps are still in need of extensive research. They aren't equipped to diagnose or prescribe medicine, however they can be a so-called "first line of defense" before seeking professional treatment or a pocket tool for mental self-exploration. As a possibility for further developments, researchers suggest a novel combination of emotion recognition and entered data to assess mental health of the users, which can be done with many readily-available datasets.

Promoting creativity and community engagement

Interactive technologies in public spaces can support human interaction, enable learning and intellectual engagement. Specifically, this has been majorly common in museums and galleries. In recent years, projects at the intersection between technology and art have been more and more popular but only a few of them can claim to have been socially engaging too. One of such art projects facilitating social behavior and playing on human psychology is called "The Shape of Story". It is a story circle, in which individuals collectively create a spoken-word narrative line-by-line. Built on rows upon rows of labeled audio data, AI in this project uses NLP to parse and semiotically visualize characters and objects into their representative symbols, creating a map of the narrative in real-time. These collaborations between people and between people and machines are a unique experience of "co-creative, multi-modal dialogue between humans and computers". From a stand of psychology, collaborative creative expression is a fantastic tool for community-building, development of closer social relationships and imaginative exchange.

We believe that AI, machine learning and deep learning are setting new milestones for socially-oriented and mindful technology. Whichever project you are planning on making — a good data is always a foundation of a high-performing AI. Here at Label Your Data, we strive to be a part of global changes in automation. Therefore, rooted in our extensive experience in the IT industry, we offer high quality data annotation, delivered fast and securely for projects worldwide. We assign you a dedicated team of trained and competent data labelers who work for you and understand your specific product and business needs. Additionally, we offer flexible and customizable software, which bends to accommodate your specific research and quality standards.