I received an "AI summary" of a Google Meeting. It wasn't very good.
The voice recognition part works, but the summary was just a collection of some of the sentences that were spoken and hard fail in trying to organize those sentences in a way that was useful as a summary.
For example, the "next steps/action items" section referenced items that had already been done.
If everyone stopped trying to use LLMs for things they're not designed for maybe people wouldn't hate "AI" so much?
Grant Gulovsen /
npub13w…aa9sf
2024-05-31 16:32:18
Author Public Key
npub13w95rfs97u8pv7gyww7kqexvplwehs067jgrmrktpmtty7t33qvspaa9sfPublished at
2024-05-31 16:32:18Event JSON
{
"id": "82b1f4338bac2451ef7e227b6ec9017df55c3c81d39b750247f9debca2b7adc5",
"pubkey": "8b8b41a605f70e16790473bd6064cc0fdd9bc1faf4903d8ecb0ed6b279718819",
"created_at": 1717165938,
"kind": 1,
"tags": [
[
"proxy",
"https://mastodon.social/users/gulovsen/statuses/112536186974340146",
"activitypub"
]
],
"content": "I received an \"AI summary\" of a Google Meeting. It wasn't very good.\n\nThe voice recognition part works, but the summary was just a collection of some of the sentences that were spoken and hard fail in trying to organize those sentences in a way that was useful as a summary.\n\nFor example, the \"next steps/action items\" section referenced items that had already been done.\n\nIf everyone stopped trying to use LLMs for things they're not designed for maybe people wouldn't hate \"AI\" so much?",
"sig": "e6dfea0a1813ae4d7a072b20b7b8832ec13278f0ac26fe63bdfe639d3bf8db98e7c39eeab55a9254f0603e2495ebad64d08b3fb11bdf5409649b027f42f66fdc"
}