1. Development of Novel Neural Network Architectures: “Act as a deep learning researcher. Propose a new neural network architecture for [specific task] that addresses current limitations in [e.g., computational efficiency, accuracy, scalability]. Include details on the network’s layers, activation functions, optimization algorithms, and potential methods for training. Discuss how your architecture improves upon existing models and outline a plan for experimental validation.”
2. Designing Ethical AI Systems with Bias Mitigation: “Assume the role of an AI ethicist. Develop a framework for designing AI systems that mitigate bias in [specific application, e.g., facial recognition, hiring algorithms]. Include methods for data collection, algorithm design, testing, and ongoing monitoring to ensure fairness and compliance with ethical standards. Discuss potential challenges and strategies to overcome them.”
3. Advanced Reinforcement Learning Algorithms: “Be an AI researcher specializing in reinforcement learning. Propose an advanced RL algorithm for [specific problem, e.g., autonomous vehicle navigation in complex environments] that addresses issues such as sample efficiency, exploration-exploitation trade-offs, and safety constraints. Include mathematical formulations, theoretical justifications, and potential applications.”
4. Quantum Machine Learning Exploration: “Act as an AI researcher exploring quantum machine learning. Design a quantum algorithm for [specific task, e.g., large-scale data clustering] that leverages quantum computing’s advantages over classical methods. Explain the quantum principles involved, how the algorithm improves upon classical counterparts, and discuss the feasibility given current quantum hardware limitations.”
5. AI Explainability and Interpretability Methods: “Assume the role of an AI researcher focused on model interpretability. Develop new methods for explaining the predictions of complex models like deep neural networks used in [specific domain, e.g., medical diagnosis]. Propose techniques that make model decisions transparent to users and regulators, balancing the trade-off between model performance and interpretability.”
6. Generative Adversarial Networks (GANs) for Data Augmentation: “Be an AI researcher working on GANs. Design a GAN architecture for generating synthetic data to augment training datasets in [specific field, e.g., rare disease medical imaging]. Address challenges such as mode collapse, training stability, and evaluation metrics. Discuss how synthetic data can improve model performance and potential ethical considerations.”
7. Meta-Learning for Few-Shot Learning Scenarios: “Act as an AI researcher specializing in meta-learning. Propose a meta-learning algorithm suitable for few-shot learning in [specific application, e.g., personalized handwriting recognition]. Describe how the algorithm can quickly adapt to new tasks with minimal data, including the underlying principles and training procedures.”
8. Multi-Agent Systems and Coordination Mechanisms: “Assume the role of an AI researcher in multi-agent systems. Develop coordination algorithms for a system of agents performing [specific task, e.g., cooperative robotics in warehouse automation]. Include methods for communication protocols, decision-making processes, and conflict resolution. Discuss scalability and robustness of the proposed methods.”
9. Robustness and Adversarial Attacks in AI Models: “Be an AI security researcher. Investigate methods to improve the robustness of AI models against adversarial attacks in [specific domain, e.g., image recognition in autonomous vehicles]. Propose defenses or training strategies to mitigate vulnerabilities and discuss how to evaluate the effectiveness of these methods.”
10. Integration of AI with Edge Computing and IoT: “Act as an AI researcher working on edge computing. Propose architectures and algorithms for deploying AI models on edge devices within IoT networks for [specific application, e.g., real-time environmental monitoring]. Address challenges such as limited computational resources, energy efficiency, latency requirements, and data privacy concerns.”