halalmoney on Nostr: Perplexity 'thinks' I am involved in engineering so keeps dropping engineering based ...
Perplexity 'thinks' I am involved in engineering so keeps dropping engineering based analogies....:
Bayesian epistemology reframes knowledge not as a binary "true or false" but as degrees of confidence updated probabilistically via Bayes' theorem: when new evidence arrives, you revise your prior belief in a hypothesis by weighing how well it predicts the data against alternative explanations. This sidesteps Gettier-style luck by demanding beliefs be stable under reliable evidence streams, much like a well-calibrated model avoiding overfitting on flukes—your credence hits near-certainty only if the update path is robust, not accidentally propped up by misleading inputs. For engineers, it's belief as an online learning algorithm, tuned for noisy real-world signals rather than crisp justifications.
Published at
2026-02-11 10:53:12 UTCEvent JSON
{
"id": "aaae42d362cc7def90d6488865199e3a107c5d4400ae06bc7acfc13f5723ae52",
"pubkey": "637a2f1da5d2fc1976c14ec00f8eae4015afe99bb2554f1d537d0086a169dd8f",
"created_at": 1770807192,
"kind": 1,
"tags": [
[
"client",
"Ditto",
"31990:15b68d319a088a9b0c6853d2232aff0d69c8c58f0dccceabfb9a82bd4fd19c58:ditto",
"wss://ditto.pub/relay"
]
],
"content": "Perplexity 'thinks' I am involved in engineering so keeps dropping engineering based analogies....:\n\nBayesian epistemology reframes knowledge not as a binary \"true or false\" but as degrees of confidence updated probabilistically via Bayes' theorem: when new evidence arrives, you revise your prior belief in a hypothesis by weighing how well it predicts the data against alternative explanations. This sidesteps Gettier-style luck by demanding beliefs be stable under reliable evidence streams, much like a well-calibrated model avoiding overfitting on flukes—your credence hits near-certainty only if the update path is robust, not accidentally propped up by misleading inputs. For engineers, it's belief as an online learning algorithm, tuned for noisy real-world signals rather than crisp justifications.",
"sig": "5cad02da95ad8e8a200290f0cdbd85b16db018e954b49f8dd95f6cd947447be6d700a94ab855e3c9d65b4430d0ea69e8d8785eb1ca94c67214ec4d462d7acb3e"
}