Message:

<p>[QUOTE=&quot;Cherd, post: 24489955, member: 24754&quot;]Considering the theoretical boundaries on what AI could become, human judgement wouldn&#039;t be a very high bar to achieve. So, it could easily become the case that AI is generally considered to be more &quot;Correct&quot; in every case (which would be valid). AI might develop into something with cognitive abilities that we are simply incapable of understanding (like an earthworm trying to understand what a human is thinking). In that case, we would have no grounds upon which to question its rational.</p><p><br /></p><p>But to your point, even if we took every precaution to retain full control over its motivations, I suppose &quot;trustworthiness&quot; would always be a question worth evaluating. If it&#039;s goals deviated even slightly from our interests, then that could be very bad for us.</p><p><br /></p><p>However, I can easily see a situation down the road where, for instance, the AI gives 2 people drastically different prison sentences for the same crime. Even though the reasoning may not be readily apparent to us, our explanation would become, &quot;Because the AI said so&quot;. This is both promising (assuming the AI&#039;s intentions align with our desired ends) and scary (because we don&#039;t know how we are being guided towards those ends).[/QUOTE]</p><p><br /></p>

Your name or email address:

Do you already have an account?

• No, create an account now.
• Yes, my password is:
• Forgot your password?

Stay logged in