abobla@lemm.ee to

Science@beehaw.orgEnglish · 10 months ago

Even the best AI models studied can be fooled by nonsense sentences, showing that “their computations are missing something about the way humans process language.”

zuckermaninstitute.columbia.edu

1

6

Even the best AI models studied can be fooled by nonsense sentences, showing that “their computations are missing something about the way humans process language.”

zuckermaninstitute.columbia.edu

abobla@lemm.ee to

Science@beehaw.orgEnglish · 10 months ago

1

Verbal Nonsense Reveals Limitations of AI Chatbots

zuckermaninstitute.columbia.edu

You must log in or register to comment.

Chat

AnalogyAddict@beehaw.org
link
fedilink
arrow-up
1·
10 months ago
Well, yes. AI models don’t extract meaning. They parrot statistically likely responses based on words used. They had to research that?

Science@beehaw.org

science@beehaw.org

You are not logged in. However you can subscribe from another Fediverse account, for example Lemmy or Mastodon. To do this, paste the following into the search field of your instance: !science@beehaw.org

Studies, research findings, and interesting tidbits from the ever-expanding scientific world.

Subcommunities on Beehaw:

Space

Be sure to also check out these other Fediverse science communities:

This community’s icon was made by Aaron Schneider, under the CC-BY-NC-SA 4.0 license.

Visibility: Public

This community can be federated to other instances and be posted/commented in by their users.

3 users / day
194 users / week
614 users / month
2.51K users / 6 months
3 local subscribers
12.8K subscribers
928 Posts
4.52K Comments
Modlog