AI: We Build How We Fail
The Flawed Reflection
When I first began studying the failures of AI systems, I approached them as I would any technical problem, looking for bugs to fix, parameters to adjust, or data to clean. But over time, I noticed that the ways that these systems failed weren't random. They were eerily, uncomfortably familiar.
Consider the phenomenon of AI "hallucination", when a large …
Keep reading with a 7-day free trial
Subscribe to AI Sherpa to keep reading this post and get 7 days of free access to the full post archives.