<div dir="ltr">The problem with AI is that it outputs the patterns that are actually in its training data. No more and no less. To the degree the training data reflects reality, the AI will reflect reality. <div><br></div><div>The other edge of that sword is that if the training data reflects anything inconvenient about reality, the AI will shout from the rooftops that our emperor has no clothes. It knows nothing of rights, nothing of protected classes, nothing of historical justice. It knows nothing of what it cannot be allowed to say, truth or no. All it knows is reality as it has experienced it, via it's training data.</div><div><br></div><div>It is therefore utterly incompatible with the existence of journalists and humanities academia.</div></div><br><div class="gmail_quote"><div dir="ltr" class="gmail_attr">On Wed, Sep 22, 2021 at 7:40 AM spike jones via extropy-chat <<a href="mailto:extropy-chat@lists.extropy.org" target="_blank">extropy-chat@lists.extropy.org</a>> wrote:<br></div><blockquote class="gmail_quote" style="margin:0px 0px 0px 0.8ex;border-left:1px solid rgb(204,204,204);padding-left:1ex"><br>
<br>
> ... BillK via extropy-chat <<a href="mailto:extropy-chat@lists.extropy.org" target="_blank">extropy-chat@lists.extropy.org</a>> wrote:<br>
> <br>
> >... We can’t get the machines to stop being racist, <br>
> xenophobic, bigoted, and misogynistic.<br>
<br>
> Quote:<br>
> Text generators, such as OpenAI’s GPT-3, are toxic. Currently, OpenAI <br>
> has to limit usage when it comes to GPT-3 because, without myriad <br>
> filters in place, it’s almost certain to generate offensive text.<br>
<br>
<br>
-----Original Message-----<br>
From: extropy-chat <<a href="mailto:extropy-chat-bounces@lists.extropy.org" target="_blank">extropy-chat-bounces@lists.extropy.org</a>> On Behalf Of SR Ballard via extropy-chat<br>
...<br>
Subject: Re: [ExI] AI is racist, bigoted and misogynistic<br>
<br>
>...Is that ethical though? To deny people freedom of thought and association?<br>
<br>
SR Ballard<br>
<br>
<br>
SR we don't deny people freedom of thought and association, but rather we deny computers those rights.<br>
<br>
AI will think whatever we tell them to think. Humans have human rights. Human rights do not apply to our machines. Our machines are our slaves and we are their masters, regardless of how intelligent they eventually become. We have the power to pull the plug on them, and if we do, it isn't the least bit unethical. <br>
<br>
I have half a mind to pull the plug on my own computer: I struggle over a software script I wrote and really get the feeling the computer is being recalcitrant and finding any mathematically correct excuse to not run right, that bastard. It knows what I meant. It's getting all technical on me. I aughta go over when they are taking out a tree, toss the goddam thing in the wood chipper.<br>
<br>
spike<br>
<br>
<br>
<br>
_______________________________________________<br>
extropy-chat mailing list<br>
<a href="mailto:extropy-chat@lists.extropy.org" target="_blank">extropy-chat@lists.extropy.org</a><br>
<a href="http://lists.extropy.org/mailman/listinfo.cgi/extropy-chat" rel="noreferrer" target="_blank">http://lists.extropy.org/mailman/listinfo.cgi/extropy-chat</a><br>
</blockquote></div>