What is meant with “black box” AI systems/tools/methods?
Are those systems/tools/methods whose inner workings are non-transparent (but their input/output behavior is well-defined/understood)?
Or are those systems/tools/methods whose input/output mapping is non-transparent (too)?
=> More informations about this toot | More toots from Iris@scholar.social
@Iris
“black box” term used to mean 100% opaque + abstract internals & clearly defined relationship between the inputs & outputs e.g. car hydrolic mechanism is a “black box” for most drivers but they understand how much foot downward pressure to apply for a related amount of deceleration.
Now, it means regardless of the inputs, outputs are “randomly, biased & unethically” generated, to be accepted without critically being assessed e.g. pressing brake peddle “might” decelerate the car! 🤷🏾♀️
=> More informations about this toot | More toots from dahukanna@mastodon.social This content has been proxied by September (3851b).Proxy Information
text/gemini