While many American AI researchers make claims that large language models might be sentient enough to be moral patients, Google's DeepMind lab is a lot more realistic in this regard.
Senior staff scientist Alexander Lerchner has emphatically asserted that disembodied systems can never have sentience.
Great paper: https://philpapers.org/rec/LERTAF
Matthew Sheffield
npub1q8…49e3t
2026-07-18 21:22:32 UTC
Author Public Key
npub1q8ncj6m4utxz750500zal6qn5lg48xjlm49anafmd7pmv2rlztfsg49e3tSeen on
wss://relay.ditto.pubPublished at
2026-07-18 21:22:32 UTCEvent JSON
{
"id": "a3870e955885e7e0735c4aecba75b403c236226de0149d9f9f2a4f989e3516fa",
"pubkey": "01e7896b75e2cc2f51f47bc5dfe813a7d1539a5fdd4bd9f53b6f83b6287f12d3",
"created_at": 1784409752,
"kind": 1,
"tags": [
[
"proxy",
"https://mastodon.social/users/mattsheffield/statuses/116943077565724515",
"activitypub"
],
[
"client",
"Mostr",
"31990:6be38f8c63df7dbf84db7ec4a6e6fbbd8d19dca3b980efad18585c46f04b26f9:mostr",
"wss://relay.ditto.pub"
]
],
"content": "While many American AI researchers make claims that large language models might be sentient enough to be moral patients, Google's DeepMind lab is a lot more realistic in this regard.\n\nSenior staff scientist Alexander Lerchner has emphatically asserted that disembodied systems can never have sentience. \n\nGreat paper: https://philpapers.org/rec/LERTAF",
"sig": "cd2a09b6a89e1f4ef9a7278c90b3c747967e8458ca50c43ee8699f8b8cfd6d72c11a9ee8eef72c88225df1e0956892f25e7df042607675a5e6720abaa358aabe"
}