nprofile1qy2hwumn8ghj7un9d3shjtnddaehgu3wwp6kyqpqgpaqvffj4tmy7nu0ygznk6nfkf66ckec0p4vyj24mt4f2s8fvulsxyhujc (nprofile…hujc) Has anyone checked the repeatability of AI systems - for a given input, do you get the same output?
If not, how does anyone know whether the system is behaving correctly?
What do the acceptance tests for AI software look like? If you don't have acceptance tests or can't demonstrate the software meets its functional requirements, how do you know the software works?
arclight
npub1qs…ty42h
2025-01-15 15:13:48 UTC
in reply to nevent1q…hj4h
Author Public Key
npub1qs2xm72kzfjvv4prrl5xnnnwmnrxk5tk689kusnwa20hsmu6532qfty42hSeen on
wss://relay.ditto.pubPublished at
2025-01-15 15:13:48 UTCEvent JSON
{
"id": "23fc70775421bb661626c1bfe40de8f3c53350e76b1a7daa5e3fa3afce35f1e2",
"pubkey": "04146df9561264c654231fe869ce6edcc66b5176d1cb6e426eea9f786f9aa454",
"created_at": 1736954028,
"kind": 1,
"tags": [
[
"p",
"407a062532aaf64f4f8f22053b6a69b275ac5b38786ac24955daea9540e9673f",
"wss://relay.mostr.pub"
],
[
"p",
"0ec11b8e4b69cd09a5a9659cdc1c49d356e9c953a3aa097ca1361cd48f8297e3",
"wss://relay.mostr.pub"
],
[
"e",
"46fba5c15ec83bfc2c321fa2c2046c2c6e655a87f3861fd7782b26ae9471ba62",
"wss://relay.mostr.pub",
"reply"
],
[
"proxy",
"https://oldbytes.space/users/arclight/statuses/113833019200840916",
"activitypub"
]
],
"content": "nostr:nprofile1qy2hwumn8ghj7un9d3shjtnddaehgu3wwp6kyqpqgpaqvffj4tmy7nu0ygznk6nfkf66ckec0p4vyj24mt4f2s8fvulsxyhujc Has anyone checked the repeatability of AI systems - for a given input, do you get the same output?\n\nIf not, how does anyone know whether the system is behaving correctly?\n\nWhat do the acceptance tests for AI software look like? If you don't have acceptance tests or can't demonstrate the software meets its functional requirements, how do you know the software works?",
"sig": "75b3a82357441fe4ab35709f5b50d3e194d07ef728d3d1d5f172c106ba0361d7aacc6905130b5aaaac8dd438203bf6cd6ccedb132ca86798bd537b495d789ec4"
}