Poor redaction technique makes this easier.

https://www.schneier.com/blog/archives/2024/03/using-llms-to-unredact-text.html

Using LLMs to Unredact Text

Initial results in using LLMs to unredact text based on the size of the individual-word redaction rectangles.

This feels like something that a specialized ML system could be trained on.

I had never heard of this. Would it be worth while to wack politicians on the head? (or just fun!)

https://www.psychologytoday.com/us/articles/202403/the-mystery-of-sudden-genius

The Mystery of Sudden Genius

The phenomenon of acquired savant syndrome reveals what happens when brain damage unleashes brilliance.

… No single theory can explain all cases of savant syndrome. But one leading hypothesis, especially for acquired savant syndrome, is that anomaly or damage in the left hemisphere essentially rewires the brain, recruiting new capacity or releasing dormant ability from the intact right hemisphere—much as a person with a broken leg relies more heavily on their opposing, functioning limb, thereby strengthening it.

The distinction between the two brain hemispheres is often oversimplified, but differences do exist: The left hemisphere specializes in logical, sequential, language-based processes, while the right specializes in creative and artistic processes. Acquired savant syndrome overwhelmingly involves left hemisphere damage and right hemisphere specialty skills.