[A] popular type of generative AI model can provide turn-by-turn driving directions in New York City with near-perfect accuracy — without having formed an accurate internal map of the city. Despite the model’s uncanny ability to navigate effectively, when the researchers closed some streets and added detours, its performance plummeted.
When they dug deeper, the researchers found that the New York maps the model implicitly generated had many nonexistent streets curving between the grid and connecting far away intersections.
This could have serious implications for generative AI models deployed in the real world, since a model that seems to be performing well in one context might break down if the task or environment slightly changes. —news.mit.edu
Post was last modified on 8 Nov 2024 3:14 pm
What have my students learned about creative nonfiction writing? During class they are collaborating on…
Two years after the release of ChatGPT, it may not be surprising that creative work…
I both like and hate that Canvas tracks the number of unmarked assignments that await…
The complex geometry on this wedge building took me all weekend. The interior walls still…
My older siblings say they remember our mother sitting them down to watch a new…
I played hooky to go see Wild Robot this afternoon, so I went back to…