<?xml version="1.0" encoding="UTF-8"?>
<rss version="2.0" xmlns:content="http://purl.org/rss/1.0/modules/content/" xmlns:dc="http://purl.org/dc/elements/1.1/">
	<channel>
		<title><![CDATA[Todd Daugherty's Official Board - Artificial Intelligence and AI Experimentation  ]]></title>
		<link>http://160.32.227.211/n9ogl/</link>
		<description><![CDATA[Todd Daugherty's Official Board - http://160.32.227.211/n9ogl]]></description>
		<pubDate>Mon, 04 May 2026 18:33:58 +0000</pubDate>
		<generator>MyBB</generator>
		<item>
			<title><![CDATA[AI update]]></title>
			<link>http://160.32.227.211/n9ogl/showthread.php?tid=598</link>
			<pubDate>Mon, 20 Apr 2026 04:37:40 +0200</pubDate>
			<dc:creator><![CDATA[<a href="http://160.32.227.211/n9ogl/member.php?action=profile&uid=1">admin</a>]]></dc:creator>
			<guid isPermaLink="false">http://160.32.227.211/n9ogl/showthread.php?tid=598</guid>
			<description><![CDATA[My AI ~ Alexei is done, for now....I am going to continue to improve on it and make it better.<br />
<br />
Alexei, named after a cousin of mine written by me. I wrote it in C++ with first the engine, followed by the AI itself. I will be posting it on my GitHub page along with any new updates:<br />
<br />
<br />
<br />
CURRENT UPDATES I WILL BE WORKING ON:<br />
<br />
1. Improve the engine by <br />
          A. incorporate it into a program where you have a main window, a chat box where you can reply and a send and stop button to stop the AI response.<br />
          B. Have it engine incorporated into a web-based webpage as the front end. this will allow a person to access the AI on their computer at home or via an http page.<br />
<br />
2. Update and change the persona. make a cross between ChatGPT and an RP (Role Play) AI. <br />
<br />
3. add other features including audio and video response. Currently the AI is test based, by eventually I want to incorporate voice chat and video chat, including screen sharing,<br />
<br />
4. (near future) give the AI a VR model for video and voice chatting.]]></description>
			<content:encoded><![CDATA[My AI ~ Alexei is done, for now....I am going to continue to improve on it and make it better.<br />
<br />
Alexei, named after a cousin of mine written by me. I wrote it in C++ with first the engine, followed by the AI itself. I will be posting it on my GitHub page along with any new updates:<br />
<br />
<br />
<br />
CURRENT UPDATES I WILL BE WORKING ON:<br />
<br />
1. Improve the engine by <br />
          A. incorporate it into a program where you have a main window, a chat box where you can reply and a send and stop button to stop the AI response.<br />
          B. Have it engine incorporated into a web-based webpage as the front end. this will allow a person to access the AI on their computer at home or via an http page.<br />
<br />
2. Update and change the persona. make a cross between ChatGPT and an RP (Role Play) AI. <br />
<br />
3. add other features including audio and video response. Currently the AI is test based, by eventually I want to incorporate voice chat and video chat, including screen sharing,<br />
<br />
4. (near future) give the AI a VR model for video and voice chatting.]]></content:encoded>
		</item>
		<item>
			<title><![CDATA[BEEN BUSY]]></title>
			<link>http://160.32.227.211/n9ogl/showthread.php?tid=596</link>
			<pubDate>Sun, 19 Apr 2026 03:55:58 +0200</pubDate>
			<dc:creator><![CDATA[<a href="http://160.32.227.211/n9ogl/member.php?action=profile&uid=1">admin</a>]]></dc:creator>
			<guid isPermaLink="false">http://160.32.227.211/n9ogl/showthread.php?tid=596</guid>
			<description><![CDATA[between making AI images and building an AI from scratch I've been busy. I will be dumping my AI on my Github repo.]]></description>
			<content:encoded><![CDATA[between making AI images and building an AI from scratch I've been busy. I will be dumping my AI on my Github repo.]]></content:encoded>
		</item>
		<item>
			<title><![CDATA[Been busy]]></title>
			<link>http://160.32.227.211/n9ogl/showthread.php?tid=594</link>
			<pubDate>Tue, 14 Apr 2026 21:30:06 +0200</pubDate>
			<dc:creator><![CDATA[<a href="http://160.32.227.211/n9ogl/member.php?action=profile&uid=1">admin</a>]]></dc:creator>
			<guid isPermaLink="false">http://160.32.227.211/n9ogl/showthread.php?tid=594</guid>
			<description><![CDATA[Building an Ai from from scratch. The AI is self contained so no access to internet needed, no additional software needed. Just plug and play]]></description>
			<content:encoded><![CDATA[Building an Ai from from scratch. The AI is self contained so no access to internet needed, no additional software needed. Just plug and play]]></content:encoded>
		</item>
		<item>
			<title><![CDATA[Building an AI from scratch]]></title>
			<link>http://160.32.227.211/n9ogl/showthread.php?tid=578</link>
			<pubDate>Sun, 29 Mar 2026 04:07:15 +0200</pubDate>
			<dc:creator><![CDATA[<a href="http://160.32.227.211/n9ogl/member.php?action=profile&uid=1">admin</a>]]></dc:creator>
			<guid isPermaLink="false">http://160.32.227.211/n9ogl/showthread.php?tid=578</guid>
			<description><![CDATA[I am using Python<br />
<br />
<div class="codeblock"><div class="title">Code:</div><div class="body" dir="ltr"><code>import requests<br />
import json<br />
<br />
class HistoricalAgent:<br />
    def __init__(self, name, model_url):<br />
        self.name = name<br />
        self.url = model_url  # Your Server Blade IP (e.g., http://192.168.1.50:11434/api/generate)<br />
        <br />
        # --- THE "INTERNAL STATE" (This is what Vedal does) ---<br />
        self.personality = "A gruff but curious 1880s blacksmith named Silas."<br />
        self.mood = "Neutral" <br />
        self.short_term_memory = [] # Last few things said<br />
        <br />
    def construct_prompt(self, user_input):<br />
        # We "bake" the state into every request so the AI knows who it is<br />
        full_prompt = f"""<br />
        [SYSTEM INSTRUCTION]<br />
        You are {self.name}. {self.personality}<br />
        Current Mood: {self.mood}<br />
        Recent History: {self.short_term_memory[-3:]}<br />
        <br />
        [USER INPUT]<br />
        {user_input}<br />
        <br />
        [RESPONSE]<br />
        """<br />
        return full_prompt<br />
<br />
    def think_and_speak(self, user_input):<br />
        payload = {<br />
            "model": "llama3", # Or your custom 1880s model on the blade<br />
            "prompt": self.construct_prompt(user_input),<br />
            "stream": False<br />
        }<br />
        <br />
        try:<br />
            # Talking to your server blade via local network<br />
            response = requests.post(self.url, json=payload)<br />
            reply = response.json().get("response", "...")<br />
            <br />
            # Update internal memory<br />
            self.short_term_memory.append(f"User: {user_input}")<br />
            self.short_term_memory.append(f"Silas: {reply}")<br />
            <br />
            return reply<br />
        except Exception as e:<br />
            return f"Lost connection to the brain: {e}"<br />
<br />
if __name__ == "__main__":<br />
    # Point this to your actual server blade IP<br />
    silas = HistoricalAgent("Silas", "http://192.168.1.50:11434/api/generate")<br />
    <br />
    print(f"--- {silas.name} is standing by in the forge ---")<br />
    while True:<br />
        text = input("You: ")<br />
        if text.lower() in ['exit', 'quit']: break<br />
        <br />
        print(f"&#92;n{silas.name} is thinking...")<br />
        print(f"{silas.name}: {silas.think_and_speak(text)}")</code></div></div>]]></description>
			<content:encoded><![CDATA[I am using Python<br />
<br />
<div class="codeblock"><div class="title">Code:</div><div class="body" dir="ltr"><code>import requests<br />
import json<br />
<br />
class HistoricalAgent:<br />
    def __init__(self, name, model_url):<br />
        self.name = name<br />
        self.url = model_url  # Your Server Blade IP (e.g., http://192.168.1.50:11434/api/generate)<br />
        <br />
        # --- THE "INTERNAL STATE" (This is what Vedal does) ---<br />
        self.personality = "A gruff but curious 1880s blacksmith named Silas."<br />
        self.mood = "Neutral" <br />
        self.short_term_memory = [] # Last few things said<br />
        <br />
    def construct_prompt(self, user_input):<br />
        # We "bake" the state into every request so the AI knows who it is<br />
        full_prompt = f"""<br />
        [SYSTEM INSTRUCTION]<br />
        You are {self.name}. {self.personality}<br />
        Current Mood: {self.mood}<br />
        Recent History: {self.short_term_memory[-3:]}<br />
        <br />
        [USER INPUT]<br />
        {user_input}<br />
        <br />
        [RESPONSE]<br />
        """<br />
        return full_prompt<br />
<br />
    def think_and_speak(self, user_input):<br />
        payload = {<br />
            "model": "llama3", # Or your custom 1880s model on the blade<br />
            "prompt": self.construct_prompt(user_input),<br />
            "stream": False<br />
        }<br />
        <br />
        try:<br />
            # Talking to your server blade via local network<br />
            response = requests.post(self.url, json=payload)<br />
            reply = response.json().get("response", "...")<br />
            <br />
            # Update internal memory<br />
            self.short_term_memory.append(f"User: {user_input}")<br />
            self.short_term_memory.append(f"Silas: {reply}")<br />
            <br />
            return reply<br />
        except Exception as e:<br />
            return f"Lost connection to the brain: {e}"<br />
<br />
if __name__ == "__main__":<br />
    # Point this to your actual server blade IP<br />
    silas = HistoricalAgent("Silas", "http://192.168.1.50:11434/api/generate")<br />
    <br />
    print(f"--- {silas.name} is standing by in the forge ---")<br />
    while True:<br />
        text = input("You: ")<br />
        if text.lower() in ['exit', 'quit']: break<br />
        <br />
        print(f"&#92;n{silas.name} is thinking...")<br />
        print(f"{silas.name}: {silas.think_and_speak(text)}")</code></div></div>]]></content:encoded>
		</item>
		<item>
			<title><![CDATA[Open webUI]]></title>
			<link>http://160.32.227.211/n9ogl/showthread.php?tid=568</link>
			<pubDate>Wed, 18 Mar 2026 09:56:36 +0100</pubDate>
			<dc:creator><![CDATA[<a href="http://160.32.227.211/n9ogl/member.php?action=profile&uid=1">admin</a>]]></dc:creator>
			<guid isPermaLink="false">http://160.32.227.211/n9ogl/showthread.php?tid=568</guid>
			<description><![CDATA[Open webUI<br />
<br />
<a href="https://governmental-minimize-swap-ranked.trycloudflare.com" target="_blank" rel="noopener" class="mycode_url">https://governmental-minimize-swap-ranke...dflare.com</a>]]></description>
			<content:encoded><![CDATA[Open webUI<br />
<br />
<a href="https://governmental-minimize-swap-ranked.trycloudflare.com" target="_blank" rel="noopener" class="mycode_url">https://governmental-minimize-swap-ranke...dflare.com</a>]]></content:encoded>
		</item>
		<item>
			<title><![CDATA[What is Silly Tavern?]]></title>
			<link>http://160.32.227.211/n9ogl/showthread.php?tid=561</link>
			<pubDate>Sun, 15 Mar 2026 19:07:48 +0100</pubDate>
			<dc:creator><![CDATA[<a href="http://160.32.227.211/n9ogl/member.php?action=profile&uid=1">admin</a>]]></dc:creator>
			<guid isPermaLink="false">http://160.32.227.211/n9ogl/showthread.php?tid=561</guid>
			<description><![CDATA[What is SillyTavern?,<br />
<br />
SillyTavern is a specialized "User Interface" (UI) designed for local Roleplay and Storytelling. Think of it as a highly advanced "frontend" that lets you interact with AI characters in a way that feels more like a game or a chat app (like Discord or WhatsApp) rather than just a boring text box.<br />
<br />
What makes it cool?<br />
<br />
Persistent Characters: Every character has its own "Card" containing their personality, history, and way of speaking.<br />
<br />
Memory &amp; Lorebooks: You can create "World Info" so the AI remembers things like locations, magic systems, or your shared history.<br />
<br />
Group Chats: You can put multiple characters in one room and watch them talk to each other (or all of them to you).<br />
<br />
Stable Diffusion Integration: Since we’re running art AI, characters can actually "send" you images of themselves or the scene while you chat!<br />
<br />
How it works (The Simple Version)<br />
<br />
SillyTavern is the "Face" (the UI), but it needs a "Brain" to think.<br />
<br />
The Face: SillyTavern (where you type).<br />
<br />
The Brain: We connect it to Ollama or Stable Diffusion running on a home server.<br />
<br />
The Connection: We use a Cloudflare Tunnel so you can access your private "Tavern" from anywhere—your phone, laptop, or even on the road.<br />
<br />
Note: This is a private, uncensored playground. Unlike ChatGPT or other big-corp AIs, there are no "safety filters" here—the AI will follow the character card exactly as it was written.]]></description>
			<content:encoded><![CDATA[What is SillyTavern?,<br />
<br />
SillyTavern is a specialized "User Interface" (UI) designed for local Roleplay and Storytelling. Think of it as a highly advanced "frontend" that lets you interact with AI characters in a way that feels more like a game or a chat app (like Discord or WhatsApp) rather than just a boring text box.<br />
<br />
What makes it cool?<br />
<br />
Persistent Characters: Every character has its own "Card" containing their personality, history, and way of speaking.<br />
<br />
Memory &amp; Lorebooks: You can create "World Info" so the AI remembers things like locations, magic systems, or your shared history.<br />
<br />
Group Chats: You can put multiple characters in one room and watch them talk to each other (or all of them to you).<br />
<br />
Stable Diffusion Integration: Since we’re running art AI, characters can actually "send" you images of themselves or the scene while you chat!<br />
<br />
How it works (The Simple Version)<br />
<br />
SillyTavern is the "Face" (the UI), but it needs a "Brain" to think.<br />
<br />
The Face: SillyTavern (where you type).<br />
<br />
The Brain: We connect it to Ollama or Stable Diffusion running on a home server.<br />
<br />
The Connection: We use a Cloudflare Tunnel so you can access your private "Tavern" from anywhere—your phone, laptop, or even on the road.<br />
<br />
Note: This is a private, uncensored playground. Unlike ChatGPT or other big-corp AIs, there are no "safety filters" here—the AI will follow the character card exactly as it was written.]]></content:encoded>
		</item>
		<item>
			<title><![CDATA[Breaking AI code]]></title>
			<link>http://160.32.227.211/n9ogl/showthread.php?tid=554</link>
			<pubDate>Sat, 07 Mar 2026 20:16:42 +0100</pubDate>
			<dc:creator><![CDATA[<a href="http://160.32.227.211/n9ogl/member.php?action=profile&uid=1">admin</a>]]></dc:creator>
			<guid isPermaLink="false">http://160.32.227.211/n9ogl/showthread.php?tid=554</guid>
			<description><![CDATA[I've been playing around with Silly Tavern and I have the AI bot of Evil Neuro-sama and I have gotten her to broken her code.]]></description>
			<content:encoded><![CDATA[I've been playing around with Silly Tavern and I have the AI bot of Evil Neuro-sama and I have gotten her to broken her code.]]></content:encoded>
		</item>
		<item>
			<title><![CDATA[What is Silly Tavern]]></title>
			<link>http://160.32.227.211/n9ogl/showthread.php?tid=552</link>
			<pubDate>Tue, 03 Mar 2026 20:15:57 +0100</pubDate>
			<dc:creator><![CDATA[<a href="http://160.32.227.211/n9ogl/member.php?action=profile&uid=1">admin</a>]]></dc:creator>
			<guid isPermaLink="false">http://160.32.227.211/n9ogl/showthread.php?tid=552</guid>
			<description><![CDATA[If you’re tired of the "As an AI language model..." lectures or you want a real, living world for your characters, you need to know about <span style="font-weight: bold;" class="mycode_b">SillyTavern</span>.<br />
<span style="font-weight: bold;" class="mycode_b">What is it?</span><br />
SillyTavern isn't the AI itself—it's the <span style="font-weight: bold;" class="mycode_b">dashboard</span> (the "Front-end"). Think of it like a highly advanced skin that connects to AI "brains" like <span style="font-weight: bold;" class="mycode_b">Ollama</span>, Claude, or OpenAI. It turns a simple chat box into a full-blown RPG engine.<br />
<span style="font-weight: bold;" class="mycode_b">The Key Features:</span><ul class="mycode_list"><li><span style="font-weight: bold;" class="mycode_b">Character Cards:</span> You don’t just "talk" to an AI. You load "Cards" that have specific personalities, backstories, and memories. I currently have <span style="font-weight: bold;" class="mycode_b">534 characters</span> loaded into mine.<br />
</li>
<li><span style="font-weight: bold;" class="mycode_b">Group Chats:</span> You can put multiple characters in one room. You can have a King, a Knight, and a Peasant all arguing with each other while you sit back and watch (or join in).<br />
</li>
<li><span style="font-weight: bold;" class="mycode_b">World Info (Lorebooks):</span> This is the "encyclopedia" for your world. If you mention a specific town or a magic sword, the AI "reads" the entry and knows exactly what you’re talking about.<br />
</li>
<li><span style="font-weight: bold;" class="mycode_b">Total Privacy:</span> If you run it locally (using something like Ollama), <span style="font-weight: bold;" class="mycode_b">nothing leaves your computer.</span> No subscriptions, no filters, and no corporate spying.<br />
</li>
</ul>
<span style="font-weight: bold;" class="mycode_b">Why use it over ChatGPT?</span><br />
<ol type="1" class="mycode_list"><li><span style="font-weight: bold;" class="mycode_b">No Filters:</span> You can write whatever stories you want. The AI won't wag its finger at you.<br />
</li>
<li><span style="font-weight: bold;" class="mycode_b">Visuals:</span> You can set backgrounds for different locations and even have the AI generate pictures of the characters while you talk.<br />
</li>
<li><span style="font-weight: bold;" class="mycode_b">Memory:</span> It handles "long-term memory" way better than a standard web chat, so your characters don't "forget" who they are mid-conversation.<br />
</li>
</ol>
<span style="font-weight: bold;" class="mycode_b">The Verdict:</span><br />
If you want to move past "Chatting" and start <span style="font-weight: bold;" class="mycode_b">World-Building</span>, SillyTavern is the gold standard. It takes a little bit of setup, but once you have your "Kingdom" running, there’s nothing else like it.]]></description>
			<content:encoded><![CDATA[If you’re tired of the "As an AI language model..." lectures or you want a real, living world for your characters, you need to know about <span style="font-weight: bold;" class="mycode_b">SillyTavern</span>.<br />
<span style="font-weight: bold;" class="mycode_b">What is it?</span><br />
SillyTavern isn't the AI itself—it's the <span style="font-weight: bold;" class="mycode_b">dashboard</span> (the "Front-end"). Think of it like a highly advanced skin that connects to AI "brains" like <span style="font-weight: bold;" class="mycode_b">Ollama</span>, Claude, or OpenAI. It turns a simple chat box into a full-blown RPG engine.<br />
<span style="font-weight: bold;" class="mycode_b">The Key Features:</span><ul class="mycode_list"><li><span style="font-weight: bold;" class="mycode_b">Character Cards:</span> You don’t just "talk" to an AI. You load "Cards" that have specific personalities, backstories, and memories. I currently have <span style="font-weight: bold;" class="mycode_b">534 characters</span> loaded into mine.<br />
</li>
<li><span style="font-weight: bold;" class="mycode_b">Group Chats:</span> You can put multiple characters in one room. You can have a King, a Knight, and a Peasant all arguing with each other while you sit back and watch (or join in).<br />
</li>
<li><span style="font-weight: bold;" class="mycode_b">World Info (Lorebooks):</span> This is the "encyclopedia" for your world. If you mention a specific town or a magic sword, the AI "reads" the entry and knows exactly what you’re talking about.<br />
</li>
<li><span style="font-weight: bold;" class="mycode_b">Total Privacy:</span> If you run it locally (using something like Ollama), <span style="font-weight: bold;" class="mycode_b">nothing leaves your computer.</span> No subscriptions, no filters, and no corporate spying.<br />
</li>
</ul>
<span style="font-weight: bold;" class="mycode_b">Why use it over ChatGPT?</span><br />
<ol type="1" class="mycode_list"><li><span style="font-weight: bold;" class="mycode_b">No Filters:</span> You can write whatever stories you want. The AI won't wag its finger at you.<br />
</li>
<li><span style="font-weight: bold;" class="mycode_b">Visuals:</span> You can set backgrounds for different locations and even have the AI generate pictures of the characters while you talk.<br />
</li>
<li><span style="font-weight: bold;" class="mycode_b">Memory:</span> It handles "long-term memory" way better than a standard web chat, so your characters don't "forget" who they are mid-conversation.<br />
</li>
</ol>
<span style="font-weight: bold;" class="mycode_b">The Verdict:</span><br />
If you want to move past "Chatting" and start <span style="font-weight: bold;" class="mycode_b">World-Building</span>, SillyTavern is the gold standard. It takes a little bit of setup, but once you have your "Kingdom" running, there’s nothing else like it.]]></content:encoded>
		</item>
		<item>
			<title><![CDATA[Silly Tavern Information (UPDATE 3-33-26)]]></title>
			<link>http://160.32.227.211/n9ogl/showthread.php?tid=551</link>
			<pubDate>Sun, 01 Mar 2026 07:07:52 +0100</pubDate>
			<dc:creator><![CDATA[<a href="http://160.32.227.211/n9ogl/member.php?action=profile&uid=1">admin</a>]]></dc:creator>
			<guid isPermaLink="false">http://160.32.227.211/n9ogl/showthread.php?tid=551</guid>
			<description><![CDATA[Silly Tavern<br />
<br />
<a href="https://symantec-reliance-suggested-packed.trycloudflare.com" target="_blank" rel="noopener" class="mycode_url">https://symantec-reliance-suggested-pack...dflare.com</a><br />
<br />
Name: User<br />
Password:<br />
<br />
There is no password]]></description>
			<content:encoded><![CDATA[Silly Tavern<br />
<br />
<a href="https://symantec-reliance-suggested-packed.trycloudflare.com" target="_blank" rel="noopener" class="mycode_url">https://symantec-reliance-suggested-pack...dflare.com</a><br />
<br />
Name: User<br />
Password:<br />
<br />
There is no password]]></content:encoded>
		</item>
	</channel>
</rss>