I ran this yesterday curious about the “extreme cognitive enhancement”, alongside PN and KB. There were some wild almost psychedelic moments after one of my walks running this where it felt like adjustments were occurring, but as far as tangibles…
I decided to put off my work on fixing one of the last major problems of my app until today, letting the advice from Bard percolate. This is one of those typically very difficult problems to resolve: a heap corruption issue, which I was not looking forward to debugging on a VR headset. I had run into heap corruption before and the solution I came up with at the time was sub-optimal in that it didn’t solve the root cause: a bug in the global allocator in my pinned version of Rust.
Out of the blue this morning when considering the problem I realized the trivially simple solution to the problem was to simply experiment with alternative global allocators. With a little lawyerly questioning of Bard as well as ruling out jemalloc
due to cross compilation issues, I was able to decide on mimalloc
, found an example pattern to replace the global allocator, and was then able to remove the heap corruption issue immediately, leaving a much more manageable logic error for me to focus on.
I don’t know how much of this is ASBR’s extreme cognitive enhancement, but compared to the work I put in previously to resolve heap corruption in the past, this is definitely an improvement, and occurred organically.