AI/ML systems are becoming an integral part of user products and applications as well as the main revenue driver for most organizations. This resulted in shifting the focus toward the Edge AI paradigm as edge devices possess the data necessary for training the models. Main Edge AI approaches either coordinate the training rounds and exchange model updates via a central server (i.e., Federated Learning), split the model training task between edge devices and a server (i.e., split Learning), or coordinate the model exchange among the edge devices via gossip protocols (i.e., decentralized training). Due to the highly heterogeneous learners, configurations, environment as well as significant synchronization challenges, these approaches are ill-suited for distributed edge learning at scale. They fail to scale with a large number of learners and produce models with low qualities at prolonged training times. It is imperative for modern applications to rely on a system providing timely and accurate models. This project addresses this gap by proposing a ground-up transformation to decentralized learning methods. Similar to Uber's delivery services, the goal of KUber is to build a novel distributed architecture to facilitate the exchange and delivery of acquired knowledge among the learning entities. In particular, we seize an opportunity to decouple the training task of a common model from the sharing task of learned knowledge. This is made possible by the advances in the AI/ML accelerators embedded in edge devices and the high-throughput and low-latency 5G/6G technologies. KUber will revolutionize the use of AI/ML methods in daily-life applications and open the door for flexible, scalable, and efficient collaborative learning between users, organizations, and governments.

Inspired by neuroscience, informed by information-theoretic principles, and motivated by modern wireless systems architectures integrating artificial intelligence (AI) and communications, this Fellowship sets out to develop a paradigm-shifting framework for networked machine learning (ML) that is centred on the following ideas. 1. Free energy minimisation: According to the free energy principle, agents optimise internal models so as to minimise their information-theoretic surprise vis-a-vis the available data and prior information. This principle offers a basis to reason about epistemic uncertainty ("know when you don't know") in AI agents that is grounded in information-theoretic analyses of out-of-sample generalisation - away from the current narrow focus on point-wise accuracy, towards uncertainty quantification and calibration. A well-calibrated agent can make informed decisions about when to refrain from acting, about when and how to collect or request more data from the environment or other agents, and about how to guard against anomalies or malicious agents. 2. Networked meta-learning: In meta-learning, agents do not share an ML model in full as in conventional, centralised, solutions. Rather, only a meta-model is shared as a means to transfer knowledge across agents, while enabling the optimisation of personalised local models. As advocated by FreeML, meta-models can naturally implement the engineering principle of modularity by encompassing a common repository of functions that can be combined to suit the cognitive needs of each agent. This framework bridges the gap between the dominant centralised or joint learning approaches - including also federated learning - and the individual learning baseline, by means of limited model sharing, while still enabling meaningful cooperation with a controlled privacy loss. 3. Native integration of wireless communication and learning: Conventional wireless systems are based on the principle of separation between computing and communications. In contrast, the native integration of communications and learning advocated by FreeML embeds wireless communication primitives as part of the data generating and processing model. Like state-of-the-art integrated solutions, the proposed approach aims at fully utilizing radio channel capacity by avoiding inefficiencies due to separate processing. Unlike existing methods, however, the FreeML framework moves away from the standard problem of communicating under uncertainty (on the communication channel) to the novel problem of communicating uncertainty (on the solution of the cognitive task) under uncertainty (on the communication channel) in order to support networked meta-learning. Overall, FreeML sets out to study a novel, theoretically principled, paradigm for ML that moves away from the current centralised, accuracy-focused, state of the art in ML to embrace decentralization via wireless connectivity, uncertainty quantification, personalisation, modularity, privacy preservation, and the right to erasure. FreeML will involve three industrial partners -- Intel, InterDigital, and Samsung AI -- that will provide guidance and feedback on aspects related to implementation efficiency, communications, and integration with wireless networks, respectively. This Fellowship proposal builds on the PI's unique inter-disciplinary expertise in information theory, ML, and communications, and is intended to enable a step change in the applicant's career towards a leadership position at the intersection of the fields of engineering and ML/AI. Through this programme, the PI will reach out to a diverse community of STEM students, public, regulators, journalists, and academic colleagues across the two fields to advocate for the central role of engineering for reliable and sustainable ML/AI.

The current AI paradigm at best reveals correlations between model input and output variables. This falls short of addressing health and healthcare challenges where knowing the causal relationship between interventions and outcomes is necessary and desirable. In addition, biases and vulnerability in AI systems arise, as models may pick up unwanted, spurious correlations from historic data, resulting in the widening of already existing health inequalities. Causal AI is the key to unlock robust, responsible and trustworthy AI and transform challenging tasks such as early prediction, diagnosis and prevention of disease. The Causality in Healthcare AI with Real Data (CHAI) Hub will bring together academia, industry, healthcare, and policy stakeholders to co-create the next-generation of world-leading artificial intelligence solutions that can predict outcomes of interventions and help choose personalised treatments, thus transforming health and healthcare. The CHAI Hub will develop novel methods to identify and account for causal relationships in complex data. The Hub will be built by the community for the community, amassing experts and stakeholders from across the UK to 1) push the boundaries of AI innovation; 2) develop cutting-edge solutions that drive desperately needed efficiency in resource-constrained healthcare systems; and 3) cement the UK's standing as a next-gen AI superpower. The data complexity in heterogeneous and distributed environments such as healthcare exacerbates the risks of bias and vulnerability and introduces additional challenges that must be addressed. Modern clinical investigations need to mix structured and unstructured data sources (e.g. patient health records, and medical imaging exams) which current AI cannot integrate effectively. These gaps in current AI technology must be addressed in order to develop algorithms that can help to better understand disease mechanisms, predict outcomes and estimate the effects of treatments. This is important if we want to ensure the safe and responsible use of AI in personalised decision making. Causal AI has the potential to unearth novel insights from observational data, formalise treatment effects, assess outcome likelihood, and estimate 'what-if' scenarios. Incorporating causal principles is critical for delivering on the National AI Strategy to ensure that AI is technically and clinically safe, transparent, fair and explainable. The CHAI Hub will be formed by a founding consortium of powerhouses in AI, healthcare, and data science throughout the UK in a hub-spoke model with geographic reach and diversity. The hub will be based in Edinburgh's Bayes Centre (leveraging world-class expertise in AI, data-driven innovation in health applications, a robust health data ecosystem, entrepreneurship, and translation). Regional spokes will be in Manchester (expertise in both methods and translation of AI through the Institute for Data Science and AI, and Pankhurst Institute), London (hosted at KCL, representing also UCL and Imperial, leveraging London's rapidly growing AI ecosystem) and Exeter (leveraging strengths in philosophy of causal inference and ethics of AI). The hub will develop a UK-wide multidisciplinary network for causal AI. Through extended collaborations with industry, policymakers and other stakeholders, we will expand the hub to deliver next-gen causal AI where it is needed most. We will work together to co-create, moving beyond co-ideation and co-design, to co-implementation, and co-evaluation where appropriate to ensure fit-for-purpose solutions Our programme will be flexible, will embed trusted, responsible innovation and environmental sustainability considerations, will ensure that equality diversity and inclusion principles are reflected through all activities, and will ensure that knowledge generated through CHAI will continue to have real-world impact beyond the initial 60 months.