<html xmlns:v="urn:schemas-microsoft-com:vml" xmlns:o="urn:schemas-microsoft-com:office:office" xmlns:w="urn:schemas-microsoft-com:office:word" xmlns:m="http://schemas.microsoft.com/office/2004/12/omml" xmlns="http://www.w3.org/TR/REC-html40"><head><meta http-equiv=Content-Type content="text/html; charset=utf-8"><meta name=Generator content="Microsoft Word 15 (filtered medium)"><style><!--
/* Font Definitions */
@font-face
{font-family:"Cambria Math";
panose-1:2 4 5 3 5 4 6 3 2 4;}
@font-face
{font-family:Calibri;
panose-1:2 15 5 2 2 2 4 3 2 4;}
@font-face
{font-family:"Comic Sans MS";
panose-1:3 15 7 2 3 3 2 2 2 4;}
/* Style Definitions */
p.MsoNormal, li.MsoNormal, div.MsoNormal
{margin:0in;
font-size:11.0pt;
font-family:"Calibri",sans-serif;}
a:link, span.MsoHyperlink
{mso-style-priority:99;
color:blue;
text-decoration:underline;}
span.EmailStyle18
{mso-style-type:personal-reply;
font-family:"Calibri",sans-serif;
color:windowtext;}
.MsoChpDefault
{mso-style-type:export-only;
font-family:"Calibri",sans-serif;}
@page WordSection1
{size:8.5in 11.0in;
margin:1.0in 1.0in 1.0in 1.0in;}
div.WordSection1
{page:WordSection1;}
--></style><!--[if gte mso 9]><xml>
<o:shapedefaults v:ext="edit" spidmax="1026" />
</xml><![endif]--><!--[if gte mso 9]><xml>
<o:shapelayout v:ext="edit">
<o:idmap v:ext="edit" data="1" />
</o:shapelayout></xml><![endif]--></head><body lang=EN-US link=blue vlink=purple style='word-wrap:break-word'><div class=WordSection1><p class=MsoNormal><o:p> </o:p></p><p class=MsoNormal><o:p> </o:p></p><div style='border:none;border-top:solid #E1E1E1 1.0pt;padding:3.0pt 0in 0in 0in'><p class=MsoNormal><b>…</b>> <b>On Behalf Of </b>William Flynn Wallace via extropy-chat<br><b>Subject:</b> Re: [ExI] chatbots and marketing: was RE: India and the periodic table<o:p></o:p></p></div><p class=MsoNormal><o:p> </o:p></p><div><div><p class=MsoNormal><span style='font-size:18.0pt;font-family:"Comic Sans MS"'>>…<span style='color:black'>Spike we have thumb drives that can hold a trilobite. Wait, no. A terabyte. No, wait. That would be a piece of a monster. Anyway, quite a huge amount of memory, and we have memory problems? bill </span>w<o:p></o:p></span></p><p class=MsoNormal><span style='font-size:18.0pt;font-family:"Comic Sans MS"'><o:p> </o:p></span></p><p class=MsoNormal><span style='font-size:18.0pt;font-family:"Comic Sans MS"'><o:p> </o:p></span></p><p class=MsoNormal><span style='font-size:18.0pt;font-family:"Comic Sans MS"'>Billw, it isn’t memory, it is the processing needed to do the training once the trillion tokens are available on that trilobite on that memory stick. ChatGPT doesn’t just look up the answer from a huge text base, it creates probability matrices based on tokens, then looking at all the tokens in the prompt, then calculating what is the most likely next token, stacking them end to end, making sentences and paragraphs.<o:p></o:p></span></p><p class=MsoNormal><span style='font-size:18.0pt;font-family:"Comic Sans MS"'><o:p> </o:p></span></p><p class=MsoNormal><span style='font-size:18.0pt;font-family:"Comic Sans MS"'>Ever since I started learning about the GPT process, I have been seeing parallels (and some important differences) in how that algorithm works when compared and contrasted with a critical breakthrough in classical controls called the Kalman filter. The Kalman filter studies a database and creates a probability matrix of correlation coefficients, then figures out a probabilistic prediction.<o:p></o:p></span></p><p class=MsoNormal><span style='font-size:18.0pt;font-family:"Comic Sans MS"'><o:p> </o:p></span></p><p class=MsoNormal><span style='font-size:18.0pt;font-family:"Comic Sans MS"'>Large language models look at a huge text base, then do their statistical magic to predict the next token.<o:p></o:p></span></p><p class=MsoNormal><span style='font-size:18.0pt;font-family:"Comic Sans MS"'><o:p> </o:p></span></p><p class=MsoNormal><span style='font-size:18.0pt;font-family:"Comic Sans MS"'>As I understand it, the use of tokens (to represent words) makes the resultant LLM work in any language. Cool!<o:p></o:p></span></p><p class=MsoNormal><span style='font-size:18.0pt;font-family:"Comic Sans MS"'><o:p> </o:p></span></p><p class=MsoNormal><span style='font-size:18.0pt;font-family:"Comic Sans MS"'>So it isn’t so much a memory limit (agreed we have plenty of that) to a processor limit, which is why GPUs, super fast Graphical Processor Units become so critically important, which is why NVIDIA stock went thru the roof recently and its market cap is very close to a trillion dollars and might crash thru that 1E12 at any time…<o:p></o:p></span></p><p class=MsoNormal><span style='font-size:18.0pt;font-family:"Comic Sans MS"'><o:p> </o:p></span></p><p class=MsoNormal><span style='font-size:18.0pt;font-family:"Comic Sans MS"'><a href="https://companiesmarketcap.com/nvidia/marketcap/">https://companiesmarketcap.com/nvidia/marketcap/</a><o:p></o:p></span></p><p class=MsoNormal><span style='font-size:18.0pt;font-family:"Comic Sans MS"'><o:p> </o:p></span></p><p class=MsoNormal><span style='font-size:18.0pt;font-family:"Comic Sans MS"'>…because Nvidia makes the best GPUs and GPUs are necessary to make chatbots and Taiwan manufactures Nvidia’s GPUs and why China will likely grab Taiwan at its first opportunity: Taiwan makes the hardware which the world needs to make good AIs, and he who has the most and best AIs owns the planet.<o:p></o:p></span></p><p class=MsoNormal><span style='font-size:18.0pt;font-family:"Comic Sans MS"'><o:p> </o:p></span></p><p class=MsoNormal><span style='font-size:18.0pt;font-family:"Comic Sans MS"'>Any questions?<o:p></o:p></span></p><p class=MsoNormal><span style='font-size:18.0pt;font-family:"Comic Sans MS"'><o:p> </o:p></span></p><p class=MsoNormal><span style='font-size:18.0pt;font-family:"Comic Sans MS"'>spike<o:p></o:p></span></p><p class=MsoNormal><span style='font-size:18.0pt;font-family:"Comic Sans MS"'><o:p> </o:p></span></p><p class=MsoNormal><span style='font-size:18.0pt;font-family:"Comic Sans MS"'><o:p> </o:p></span></p><p class=MsoNormal><span style='font-size:18.0pt;font-family:"Comic Sans MS"'><o:p> </o:p></span></p><p class=MsoNormal><span style='font-size:18.0pt;font-family:"Comic Sans MS"'><o:p> </o:p></span></p><p class=MsoNormal><span style='font-size:18.0pt;font-family:"Comic Sans MS"'><o:p> </o:p></span></p></div></div></div></body></html>