“Brandon Foodi states that current AI models are not as proficient with multimodal tasks because it has not been a primary focus for AI labs and is more difficult to implement with Reinforcement Learning.”