Here's the list of things that I think will be hot in the filed of Machine Learning and Deep Learning in the near future: Self-supervised learning is the machine learning technique that can be used to train models for tasks in NLP, Computer Vision, Reinforcement Learning and robotics. This method works by getting small amount of labeled data, learning common patterns from it and then by using these representations work large amounts of unlabeled data. LLMs like BigScience's BLOOM or OpenAI GPT-3 are getting better and better and therefore they become able to some rich varieties of tasks such as SQL code generation, Image Captioning or Essay writing and many more awesome things! I think it's one of the most ambitious as well as exciting fields in all AI/Machine Learning Research. Vision Transformers (ViT)are dep neural networks that utilize transformer architecture to deal with computer vision tasks.