【Session Theme: AGI】
Since the winning of ImageNet competition in 2012, Deep Learning dominates the domains of machine learning and AI. However, there are still many unsolved struggles of Deep Learning. In this talk, we will discuss on several works by Gary Marcus, a famous critic of Deep Learning, to show the core problems of (pure) Deep Learning, and what are the steps towards Robust AI.
In this talk, we will briefly go through some works by Gary Marcus:
- Deep Learning: A Critical Appraisal
- Rebooting AI: Building Artificial Intelligence We Can Trust
- The Next Decade in AI: Four Steps Towards Robust Artificial Intelligence
To understand the core problems of Deep Learning, we will introduce the concept of Deep Learning, and provide some examples to show that Deep Learning does not work. Three main drawbacks of (pure) Deep Learning we want to focus on are:
- Greediness: Deep Learning often requires a massive amount of data
- Opaqueness: Human-style explanation of Deep Learning systems is hard
- Brittleness: Even powerful Deep Learning systems are usually easily to be fooled.
After discovering the issues of Deep Learning, we will try to find the road that guides us to next level of AI. The study materials include some insights from human mind, common sense, and deep understanding. Finally, we will show a novel AI architecture proposed by Gary Marcus, which is based on a hybrid, knowledge-driven, cognitive-model-based approach.
About Su Jia Kuan
蘇嘉冠 (Su Jia Kuan) is AI engineer in a startup. In past years, he focused on Deep Learning and its related applications (computer vision and natural language processing). Now, he is a confused guy that wants to escape the restriction from Deep Learning paradigm.