Re:"Erdos problem #728 was solved more or less autonomously ...
by ACS_Solver on Friday January 09, 2026 @08:44PM
(#46561868)
Surely AGI would be matching humans on most tasks. To me, surpassing humans on all cognitive tasks sounds like superintelligence, while AGI "only" need to perform most, but not necessarily all, cognitive tasks at the level of a human highly capable at that task.
[ Parent
| on HN ]
Re:"Erdos problem #728 was solved more or less autonomously ...
by fc417fc802 on Saturday January 10, 2026 @12:37AM
(#46563083)
Personally I could accept "most" provided that the failures were near misses as opposed to total face plants. I also wouldn't include "incompatible" tasks in the metric at all (but using that to game the metric can't be permitted either). For example the typical human only has so much working memory, so tasks which overwhelm that aren't "failed" so much as "incompatible". I'm not sure exactly what that looks like for ML but I expect the category will exist. A task that utilizes adversarial inputs might be an example of such.
Re:"Erdos problem #728 was solved more or less autonomously ...
by epolanski on Saturday January 10, 2026 @08:10AM
(#46565422)
Super intelligence is defined as outmatching the best humans in a field, but again, on all cognitive tasks, not just a subset.
AI can already beat humans in pretty much any game like Go or Chess or many videogames, but that doesn't make it general.