FAQ

AI, or Artificial Intelligence, is a technology that enables machines to perform tasks that typically require human intelligence. It involves creating algorithms and systems that can analyze data, learn from it, and make decisions or solve problems.

An AI assistant is a program powered by machine learning that can respond to you, provide information, anticipate your needs, and perform tasks at your request. While these assistants are most commonly thought of in terms of smartphones and smart home speakers, they can exist in a range of devices and will become common in XR glasses, home appliances, connected cars, and more.

AI excels at tackling complex, multi-dimensional problems that may be challenging for traditional engineering approaches. This includes tasks such as image recognition, natural language processing, and autonomous decision-making.

One of the current challenges in AI relates to explainability. Most modern AI systems work very well but are basically black boxes from the standpoint of human users. Data bias is another major challenge: when an AI model is trained on poor or biased data, the model will not perform well or fairly.

A recent groundbreaking advancement in AI involves sequential decision-making, where AI systems learn autonomously by observing human actions, a significant departure from traditional training methods relying solely on human-driven or machine learning-based data processing. For example, an AI assistant might observe that you seem to go to the same place by visiting a set of shopping areas first and it may start to automatically learn your habits and propose the shortest path towards your desired destination.

PEER is an AI-based application for sequential decision-making scenarios that tries to assist end-users by suggesting optimal decisions. PEER can be used by different types of clients, whether a machine operator in a manufacturing facility or a person requiring road navigation.

PEER keeps end-users in the development loop, starting from design goals, PEER captures the sociotechnical user requirements in order to provide AI solutions in line with human values.

Traditional AI solutions follow an isolated development pipeline and the AI solution is evaluated at the end, leading to a decreased acceptance level and trustworthiness and a waste of time and engineers' efforts. On the other hand, PEER aims to increase user acceptance by keeping the end users in the development cycle and giving the end users the autonomy to accept or override the AI decisions after deployment.

Yes, PEER AI assistance makes its decisions by considering human social values and by interacting with end users. Through this interaction, it can narrow down the possible solutions to achieve the desired outcomes.

PEER embodies a holistic approach to AI integration with its accessible design, collaborative problem-solving, and human-centric metrics. Through accessible design, PEER demystifies AI capabilities and limitations for stakeholders at every stage of development, fostering transparency and facilitating informed decision-making. PEER promotes collaborative problem-solving by establishing bidirectional communication between AI and users, nurturing a dynamic feedback loop that not only enhances user engagement but also refines outcomes continuously. Moreover, PEER prioritizes human-centric metrics by developing qualitative and quantitative measures that go beyond traditional AI performance indicators. These metrics assess interactivity, acceptance, explainability, trustworthiness, and perceived fairness, ensuring that the AI aligns seamlessly with human values and needs, ultimately enhancing its effectiveness and societal integration.