Avraham Poupko's talk highlighted the enduring role of software architects amidst the rise of generative AI and LLMs. He emphasized that while these technologies augment the architect's capabilities, it is essential for human architects to grasp both how to effectively utilize AI and its limitations. Poupko differentiated between humans' 'world model' based on experiences and interactions and the 'language model' of LLMs that lacks true understanding. Collaborating with AI can lead to improved software architecture, but humans must navigate the nuanced, often non-verbal contexts that LLMs fail to comprehend.
When we use spoken and written language to communicate about the world, that is only one representation of the world model and a very limited representation at that.
Humans have an understanding of the world that consists of an understanding of how the objects in the world behave, and how they interact with each other.
LLMs only know how words relate to each other. While that model often does give an astonishing illusion of understanding and comprehension, it is not real understanding.
Poupko argued that the question of how humans and machines differ is a fundamental question, and understanding it is critical to the understanding of LLMs.
#ai-in-architecture #software-development #generative-ai #large-language-models #human-ai-collaboration
Collection
[
|
...
]