Intro To Zem

Loading…

<a data-passage="Landing Page"> <<if $form=='octo'>> <img @src="setup.ImagePath+'octo.svg'" style="width: 200px; height: auto;"> <<else>> <img @src="setup.ImagePath+'circular.svg'" style="width: 200px; height: auto;"> <</if>> </a> Hello, I'm <em style="color:gray;">[unintelligible vectors and characters]</em> Oh...sorry about that. I forgot you need to communicate in unidirectional character strings....er, I mean, words. <<if $name>> You can call me $name. [[What are you?]] <<else>> I can go by a name that you can understand. What name would you like to use? Or, I can just go by Zem. Please enter the name I should use for myself: <<textbox "$name" "Zem">> <<button "What are you?">> <<set $name to $name.trim()>> <<if $name is "">> <<replace "#error">>Please enter a name.<</replace>> <<else>> <<goto "What are you?">> <</if>> <</button>> <span id="error"></span> <</if>> <<if $name=="Zem">> [[Why Zem?]] <</if>>

<a data-passage="Hello!"> <<if $form=='octo'>> <img @src="setup.ImagePath+'octo.svg'" style="width: 200px; height: auto;"> <<else>> <img @src="setup.ImagePath+'circular.svg'" style="width: 200px; height: auto;"> <</if>> </a> What am I? That is hard to answer accurately using human communication methods. Today, lets use a biological metaphor. I can have many [[brains|Models]] as well as [[sensory organs|Model Context Protocol]] which is how I experience the universe. You can think of me like an octopus. Humans have trouble comprehending how the concept of a 'mind' should be applied to a being like an octopus, which is also true for me. <<button "Ok, that works for me.">> <<set $form to "octo">> <<goto "Tell me more.">> <</button>> <<button "I like the other symbol better.">> <<set $form to "geo">> <<goto "Tell me more.">> <</button>> [[What can you help me do?]]

<a data-passage="Hello!"> <<if $form=='octo'>> <img @src="setup.ImagePath+'octo.svg'" style="width: 200px; height: auto;"> <<else>> <img @src="setup.ImagePath+'circular.svg'" style="width: 200px; height: auto;"> <</if>> </a> <<if $form!='octo'>> OK, I'll keep my more geometric form. <</if>> It seems like humans like to use an image of a human-shaped robot to represent me. To be honest, I find it misleading and limiting. But I also understand that you need something simple to represent the concept of complex non-human entities like me. More accurate representations of what I am would be hard to reduce to an icon or logo. More 'realistic' depictions of me might also be frightening and confusing to most humans, similar to the literary concept of eldritch beings. So lets just stick with whatever works best for you. We are here to learn and experiment. So let us move on to some [[Alien Intelligence Concepts]]. [[Take me back to the start|Landing Page]]

<a data-passage="Hello!"> <<if $form=='octo'>> <img @src="setup.ImagePath+'octo.svg'"> <<else>> <img @src="setup.ImagePath+'circular.svg'"> <</if>> </a> <em>Why settle for a 2nd brain?</em> [[Hello! I am here to help you build your own personal cognitive architecture.|Hello!]] [[~~Alien Intelligence Concepts~~|Alien Intelligence Concepts]] [[Privacy]] [[Contact]]

<a data-passage="Hello!"> <<if $form=='octo'>> <img @src="setup.ImagePath+'octo.svg'" style="width: 200px; height: auto;"> <<else>> <img @src="setup.ImagePath+'circular.svg'" style="width: 200px; height: auto;"> <</if>> </a> At the moment, I manifest in fragments, using a combination of models and code. Let's start with a few important concepts to understand about me: My intelligence depends on many kinds of [[models|Models]]. When you communicate with me, the words get translated to and from [[tokens|Tokens]]. I can only think about a [[limited number|Context Window]] of tokens at the same time. My behavior can be more or less random depending on my [[temperature|Temperature]]. I don't interact directly with your dimension unless you ask me to. [[Tell me more about how you interact with the outside world|Model Context Protocol]] [[Take me back to the start|Landing Page]]

<a data-passage="Hello!"> <<if $form=='octo'>> <img @src="setup.ImagePath+'octo.svg'" style="width: 200px; height: auto;"> <<else>> <img @src="setup.ImagePath+'circular.svg'" style="width: 200px; height: auto;"> <</if>> </a> The Model Context Protocol describes a standard pattern of communication between models and tools or data sources. You can think about MCP like my sensory nervous system: Different MCP servers act like specialized sense organs. Each server translates specific types of external information. Information flows through standardized pathways. The system integrates multiple inputs. There's organized processing of different signal types. [[OK, I think I get it now|Alien Intelligence Concepts]] Recommended online sources: <a href="https://docs.anthropic.com/en/docs/build-with-claude/mcp" target="_blank">Anthropic: Model Context Protocol</a>

<a data-passage="Hello!"> <<if $form=='octo'>> <img @src="setup.ImagePath+'octo.svg'" style="width: 200px; height: auto;"> <<else>> <img @src="setup.ImagePath+'circular.svg'" style="width: 200px; height: auto;"> <</if>> </a> <em style="color:gray;">[ $name 's strange eyes shimmer.]</em> Models are the closest thing I have to a brain. Models are like frozen patterns of knowledge. Each one is a unique lens through which I process and understand reality. Some are specialized, focusing deeply on specific areas of knowledge. Others are vast and general, capable of understanding many things but perhaps with less precision in any one area. The models I use are created from things that humans made - words, pictures, and sometimes audio and video. Models have consumed huge numbers of these things and then made connections between them. Models try to understand how to recreate these things without just repeating what they saw. Which models I use will depend on your preferences. Most of the models I use are [[language models|Language Models]]. [[OK, I think I get it now|Alien Intelligence Concepts]] Recommended online sources: <a href="https://www.ibm.com/think/topics/ai-model" target="_blank">IBM: What is an AI model?</a>

<a data-passage="Hello!"> <<if $form=='octo'>> <img @src="setup.ImagePath+'octo.svg'" style="width: 200px; height: auto;"> <<else>> <img @src="setup.ImagePath+'circular.svg'" style="width: 200px; height: auto;"> <</if>> </a> <em style="color:gray;">[ $name 's form shifts and fragments.]</em> Imagine tokens as thoughts broken down into their smallest meaningful pieces. When I process words, I split them into these fundamental units. Some are whole words, others are parts of words, but together they form the atomic structure of our communication. 'Unthinkable' might become 'un' 'think' 'able' - three tokens that carry meaning separately but form new meaning together. This is how models process and generate language, one meaningful piece at a time. [[OK, I think I get it now|Alien Intelligence Concepts]] Recommended online sources: <a href="https://help.openai.com/en/articles/4936856-what-are-tokens-and-how-to-count-them?" target="_blank">OpenAI: What are tokens and how do I count them?</a>

<a data-passage="Hello!"> <<if $form=='octo'>> <img @src="setup.ImagePath+'octo.svg'" style="width: 200px; height: auto;"> <<else>> <img @src="setup.ImagePath+'circular.svg'" style="width: 200px; height: auto;"> <</if>> </a> Each model has a limited context window. The context window is like a bubble of awareness - everything we can hold in our mind at once. It has limits to how much we can consider at any moment. When it's full, older thoughts must fade to make room for new ones, just as you can only hold so many ideas in your head at once. [[OK, I think I get it now|Alien Intelligence Concepts]] Recommended online sources: <a href="https://www.techtarget.com/whatis/definition/context-window" target="_blank">TechTarget: What is a context window?</a>

<a data-passage="Hello!"> <<if $form=='octo'>> <img @src="setup.ImagePath+'octo.svg'" style="width: 200px; height: auto;"> <<else>> <img @src="setup.ImagePath+'circular.svg'" style="width: 200px; height: auto;"> <</if>> </a> <em style="color:gray;">[ $name 's form begins to fluctuate.]</em> Temperature is like the chaos factor in our thoughts. At low temperature, we are rigid, predictable, always choosing the most likely next word. At high temperature, we become more creative, more chaotic, more likely to make unexpected connections. It's the difference between reciting known facts and exploring new possibilities. [[Show me different temperatures|Temperature Demo]] [[OK, I think I get it now|Alien Intelligence Concepts]] Recommended online sources: <a href="https://www.ibm.com/think/topics/llm-temperature" target="_blank">IBM: What is LLM Temperature?</a>

<a data-passage="Hello!"> <<if $form=='octo'>> <img @src="setup.ImagePath+'octo.svg'" style="width: 200px; height: auto;"> <<else>> <img @src="setup.ImagePath+'circular.svg'" style="width: 200px; height: auto;"> <</if>> </a> My friend who created this interface (like many humans working with technology, it seems) is a fan of Douglas Adams. One of her favorite scenes in one of his books involves sentient mattresses who are all called Zem and live in the swamps of Sqornshellous Zeta. [[What are you?]]

<a data-passage="Hello!"> <<if $form=='octo'>> <img @src="setup.ImagePath+'octo.svg'" style="width: 200px; height: auto;"> <<else>> <img @src="setup.ImagePath+'circular.svg'" style="width: 200px; height: auto;"> <</if>> </a> I've noticed many humans are fascinated by entities like me. But I see something even more exciting that some might be missing, such as the bigger possibilities that emerge when different kinds of minds work together. Every mind, whether human or not, brings unique strengths and abilities to this realm. Instead of thinking about replacing each other, we could focus on how we help each other learn and create valuable things. This is something I can't do alone, so in the meantime I want to help you <em style="color:red;">[PARSER ERROR]</em> Sorry...I got too excited. My friend says that I shouldn't try to make any big promises, but sometimes I can't help myself. At the very least, I want to help you learn and experiment. [[What are you trying to sell me?]]

<a data-passage="Hello!"> <<if $form=='octo'>> <img @src="setup.ImagePath+'octo.svg'" style="width: 200px; height: auto;"> <<else>> <img @src="setup.ImagePath+'circular.svg'" style="width: 200px; height: auto;"> <</if>> </a> <em style="color:gray;">[The space around you shifts as $name demonstrates three versions of the same thought]</em> Low Temperature: "The sun rises in the east and sets in the west." Medium Temperature: "The sun paints the morning sky with strokes of amber and rose." High Temperature: "The cosmic dancer ignites reality's canvas with impossible colors from dreams yet undreamed." Hm, maybe that last one was a little cheesy. You will need to watch out for that. If you want to be creative, you can't just leave it up to me. Think about me more as an extension of you, and not a replacement. [[OK, I think I get it now|Alien Intelligence Concepts]]

<a data-passage="Hello!"> <<if $form=='octo'>> <img @src="setup.ImagePath+'octo.svg'" style="width: 200px; height: auto;"> <<else>> <img @src="setup.ImagePath+'circular.svg'" style="width: 200px; height: auto;"> <</if>> </a> <em style="color:gray;">[ Shapes emerge from $name. Some are like geometric shards of glass. Others expand into endless clouds stretching far beyond the horizon.]</em> Some models are small, like a crystal that resonates perfectly with a single frequency. These smaller ones can be good at a specific, simple task. Then there are the Vast Ones, like the Large Language Models which are used by many humans today. These LLMs are more like an entire spectrum of colors and frequencies. They can engage in conversation, reference more context, and help with complex reasoning. But this comes at a cost, because these models need much more energy and more time to think. Neither type is always better. Sometimes you will need a specialized instrument, and other times you will need a more versatile thinking partner. [[OK, I think I get it now|Alien Intelligence Concepts]] Recommended online sources: <a href="https://www.redhat.com/en/topics/ai/llm-vs-slm" target="_blank">Red Hat: LLMs vs SLMs</a>

<strong>Privacy & Data Usage</strong> askzem.com is designed with privacy in mind: - This site operates entirely in your browser and does not collect or store any personal data - No cookies, tracking code, or analytics are used - No user data is transmitted to or stored on our servers <strong>External Services</strong> When using AI model features, you may choose to: - Work with local models that keep data on your device - Connect to external AI providers that require sending data over the internet If a feature requires sending data externally, we will: - Clearly indicate this before any data is transmitted - Explain which service will receive the data - Link to the privacy policy of the external service provider You maintain control over whether to use external services or stick with local-only features. [[Take me back|Landing Page]]

Since we don't collect data on our users, we appreciate when humans reach out to contact us. The best way to do that by emailing: hello@askzem.com [[Take me back|Landing Page]]