On Tue, Oct 6, 2020, 9:24 PM James Bowery <[email protected]> wrote:
> HA! > > https://www.infoq.com/news/2020/10/training-exceeds-gpt3/ > I find it interesting that we needed a new, harder NLP benchmark, Superglue, because systems were already scoring above human level in GLUE. The appendix of the Superglue paper has some example problems. They would be hard for a lot of humans. So I guess AGI is solved. Or will we keep raising the bar? If a computer can do it, then it's not intelligence. ------------------------------------------ Artificial General Intelligence List: AGI Permalink: https://agi.topicbox.com/groups/agi/T6761a13445e5864b-Mc482a2014e4f769c53f505aa Delivery options: https://agi.topicbox.com/groups/agi/subscription
