Record🤔

uri:

"at://did:plc:aeug3w4mdhnc2wv73hmrh36v/app.bsky.feed.post/3kswv6mqwb422"

cid:

"bafyreifrjzogr6mhxyic6wuupahf5mirlzp6yddqlwyygwdfofwl3b4brq"

value:

tags:

text:

"Hey folks, let's have a little chat about construct validity, the concept that if you're going to use a psychological test, there should be good reason to believe that a) the thing it's meant to be testing is real and b) results on the test reflect something about that thing.

>>"

$type:

"app.bsky.feed.post"

embed:

$type:

"app.bsky.embed.external"

external:

uri:

"https://spectrum.ieee.org/theory-of-mind-ai"

thumb:

View blob content

$type:

"blob"

ref:

$link:

bafkreiekdvheyhverst2a3ewr4ku36v6ygzalvsjyrslvze2r2ofsdjonq

mimeType:

"image/jpeg"

size:

112432

title:

"AI Outperforms Humans in Theory of Mind Tests"

description:

"A new study compared the performance of several large language models, including GPT-4 and Llama 2, with that of human participants on "theory of mind" psychology tests. Theory of mind is the ability ..."

langs:

• "en"

facets:

createdAt:

"2024-05-20T18:22:36.715Z"