<div dir="ltr">Spike,<br>I'm not sure NLMs work in that way. This particular instance of GPT-4 I used doesn't have access to the web. It was trained on web data but it doesn't go and retrieve relevant information when trying to answer a query. <br>I want to learn more about the intricate details of the NLM and I already bought books (like the Wolfram one) and downloaded several articles and code. <br>But my current understanding is that the training on the data created connections and weights among concepts and data, which is much similar to our brain creating connections, memories, and associations. It is possible that somehow there is a "memory" of the Exl chat where you mentioned sipping alcohol in association with Bexarotene. But I'm inclined to think that it did that association by itself and it had a spark of creativity. To be honest, also human creativity works a bit in that way where we make unexpected associations based on similar experiences we had but maybe in another domain, we change the context of the situation and so on. <br>Still, I find the story it created just incredible if one stops for a second and tries to understand what just happened there. <br>From a small and relatively unprecise prompt (that needed to be understood and put in the correct context) GPT-4 made up a story in mere seconds. It was not just relevant, it had a logical sequence, it had the right emotional settings, and it did an interesting association between sipping a glass of wine in a pensive manner and coming up with a relevant innovation (the title was The Breakthrough), that animal studies are necessary before human testing and that positive results in these studies were necessary to be allowed to test on humans. <br>I really don't think GPT-4 simply looks up at its memory to generate content, it really creates it based on previous training. <br>Giovanni <br><br><br><br></div><br><div class="gmail_quote"><div dir="ltr" class="gmail_attr">On Mon, Apr 3, 2023 at 9:14 AM spike jones via extropy-chat <<a href="mailto:extropy-chat@lists.extropy.org">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"><div class="msg-4306312471522837437"><div lang="EN-US" style="overflow-wrap: break-word;"><div class="m_-4306312471522837437WordSection1"><p class="MsoNormal"><u></u> <u></u></p><p class="MsoNormal"><u></u> <u></u></p><div style="border-right:none;border-bottom:none;border-left:none;border-top:1pt solid rgb(225,225,225);padding:3pt 0in 0in"><p class="MsoNormal"><b>…</b>> <b>On Behalf Of </b>Giovanni Santostasi via extropy-chat<br><b>Subject:</b> Re: [ExI] chatgpt test 2: mostly successful was RE: chatgpt test 1: fail<u></u><u></u></p></div><p class="MsoNormal"><u></u> <u></u></p><div><p class="MsoNormal">>…By the way, always try to ask to do something with a please, I want to treat it as a real mind and be respectful. <u></u><u></u></p><div><p class="MsoNormal">Giovanni <u></u><u></u></p><p class="MsoNormal"><u></u> <u></u></p><p class="MsoNormal"><u></u> <u></u></p><p class="MsoNormal"><u></u> <u></u></p><p class="MsoNormal"><u></u> <u></u></p><p class="MsoNormal">Giovanni, I am all for being polite and respectful. You know I think the world of you pal. But I flatly refuse your request.<u></u><u></u></p><p class="MsoNormal"><u></u> <u></u></p><p class="MsoNormal">spike<u></u><u></u></p></div></div></div></div>_______________________________________________<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>
</div></blockquote></div>