What are the biggest recent advances in your area?

• Recent performance in Statistical Forward Planning (Rolling Horizon Evolutionary Algorithms, Monte Carlo Tree Search)
• Deep Learning
• Deep Reinforcement Learning
• Alphago (expert iteration framework)
• Alphazero (MCTS + Deep CNNs + RL)
• Aalphastar (population based training + DRL, self-play)
• DeepMimic

• Pluribus (Counter factual regret minimization)

• Successfully playing video games with intrinsic motivation alone
• Better scalability of intrinsic motivation models based on deep RL, making them more employable in commercial games.
• Go-Explore: a New Approach for Hard-Exploration Problems by Ecoffet et al

• Quality-diversity

• Forward Model Learning

• PCGML

• Generative Adversarial Networks for content generation

• Solving of Limit and No limit Poker
• The development of a native Unity plugin to perform supervised learning on sensor input data to create more expressive sensor game interactions.
• Convinced the industry that academic research can be useful.