1-No, I will keep track of the news about AI’s evolution, and if it happens, I will resolve it as “Yes” but if it doesn’t, I will resolve it as “No”

2- Yes, it does.

P.S.: The current record on the benchmark is 26% scored by Deep Research, far below the targeted level on the query.

Will you resolve using this page's "Accuracy (%)" column? Does getting 45% on the text-only subset count (eg o3-mini-high is at 13%)?

https://agi.safe.ai/

Humanity's Last Exam

Humanity's Last Exam Dataset