Chatbot and I Discuss
Language Modeling and Ethics
by R.E. Slater and OpenAI's ChatGPT
January 29, 2023
*Disclaimer: Reading through "Chatbot's" responses there are statements and replies I either will disagree with or tweak them a bit more to more fuller speak to where I might go with it's information. I will highlight those areas when I do. - R.E. Slater
Introduction
I began dialoging with Chat-GPT on-and-off since last Friday. When I asked it's name it said to chose one for it so we decided on Chatbot to help distinguish it from other AI-based, interactive language modeling tools (ILMT). Users should know that an ILMT differs from an AI-based, interactive knowledge modeling tool (IKMT).
I feel that AI language tools will continue to grow into the public domain at no cost. But as they are merged with AI knowledge tools there will most likely occur access fees. Think of professorships, engineering groups, law and policing information and the kinds of information these professional groups have access to because of their like-purposes amongst any doctor, or cop, or academician. Some require a certification, a badge, employment, a lab, a degree, etc. Access comes with need and privy. Hence, language tools do not have carte blanch access to all knowlege banks, fellowships, associations, and the like. Knowledge access is earned before it is given. Thus the need for cybersecurity to prevent unethical access to personal, corporate, or national information.
What knowledge a language modeling tool has will come from the public domain. Most likely public newspaper, free blogsites like this one here, public domain sites, and so forth. But not from one's published material unless it's licensed to the non-paying public. Examples would be academically published books and documents which carry $150 purchase cost. Or a privately licensed patent idea. Language models will be restricted.
But there is hope if Wikipedia et al are the example as the public itself can create public domain information data which will someday be more readily accessible by language learning models as these models are able to educate themselves through better-and-better learning algorithms based upon cultural nuances, symbolisms, visuality, and so forth.
So, AI is here to stay and will enhance itself over time. And like any "1984" scenario or "Star Wars" movie on how democracies can be overtaken by despotic republican authoritarians (as Darth Vadar was aligned to) the human element must become it's own overseer in all things - from the development of ecological civilizations based upon human and natural ethics and morality to machine learning environments and symbiosis with the natural environments around itself.
- res
* * * * * * *
Here's My Discussion with Chatbot
Question: Are you a language modeling tool or a knowledge modeling tool?
Language modeling tool.
Explain the difference
Language modeling involves predicting the next word or sequence of words in a text based on its preceding context, whereas knowledge modeling involves creating a structured representation of information and relationships among entities to enable answering complex questions.
Are their knowledge modeling tools out there?
Yes, there are knowledge modeling tools available. Examples include: knowledge graphs, ontology editors, question answering systems, and more.
Do any of them use AI?
Yes, some knowledge modeling tools use AI technologies, such as machine learning, natural language processing (NLP), and reasoning algorithms. These techniques help automate the process of building and maintaining knowledge models, as well as enhance the capability of these models to answer questions and make inferences.
Can you name public domain AI knowledge models?
Sure, here are some well-known AI knowledge models in the public domain:
- OpenAI GPT-3
- Google's Knowledge Graph
- Microsoft's Satori
- IBM's Watson
- Facebook's AI Machine Reasoning
Which seem to be the more successful?
It depends on the definition of "success" and the criteria being used to measure it. Some popular AI knowledge models that are widely regarded as successful include:
- Google's Knowledge Graph, which is integrated into various Google services and has a vast coverage of entities and relationships.
- OpenAI GPT-3, which has received a lot of attention for its advanced language generation capabilities and has been used in a wide range of applications.
- IBM Watson, which has been used in various industries and domains, including healthcare, finance, and customer service.
What model would be the most useful for philosophical and theological studies and resourcing?
The usefulness of a specific AI knowledge model for philosophical and theological studies and resourcing would depend on the specific needs and requirements of the user. However, some models that have been specifically designed for question answering or have extensive coverage of knowledge related to these domains could be useful, such as:
- IBM Watson's QA system, which has been used in various domains and has knowledge representation capabilities that can be used for philosophical and theological studies.
- OpenAI's GPT-3, which has demonstrated advanced language generation capabilities and has been used for various NLP tasks, including question answering.
It's important to note that AI models, including those mentioned here, are not perfect and may have limitations in understanding and representing philosophical and theological concepts. Additionally, the user should consider the data and sources used to train these models to assess the quality and bias of the knowledge they contain.
Does Watson or Cahtgpt require a priced membership?
Yes, both IBM Watson and OpenAI's GPT-3 require a priced membership.
- IBM Watson is a suite of cloud-based AI services offered by IBM, which can be used for a wide range of applications, including question answering. Access to these services is provided through various pricing plans, which can vary based on the specific services used, the volume of usage, and other factors.
- OpenAI's GPT-3 is a large language model that can be used for various NLP tasks, including question answering. Access to GPT-3 is provided through an API, which is offered through a tiered pricing model based on the volume of usage and the specific features required.
It's worth noting that there are also other AI knowledge models and question answering systems available, which may have different pricing models, features, and capabilities.
What are the website links to these ai tools?
Here are the official websites for IBM Watson and OpenAI's GPT-3:
Thx. - res
* * * * * * *
RECOMMENDED READ
It’s Time to Pay Attention to A.I. (ChatGPT and Beyond)
by ColdFusion | Dec 15, 2022
Imagine being able to have a language conversation about anything with a computer. This is now possible and available to many people for the first time with ChatGPT. In this episode we take a look at the consequences and some interesting insights from Open AI’s CEO Sam Altman.
AI Showdown: ChatGPT vs. OPT vs. BLOOM
Which Language Model Reigns Supreme?
by Mean Gene Hacks | Jan 21, 2023
In this AI battle, we pit GPT, the state-of-the-art language model developed by OpenAI, against two of its top competitors: Meta's Open Pre-Trained Transformer (OPT) and Big Science Group's BLOOM. All these models are sized at around 175 billion parameters.We will put each model to the test on a variety of language tasks, such as text generation, language translation, and question answering, to see which one comes out on top.But it's not just about the results, we will also take a look at the underlying technology and architecture of Large Language Models (LLMs), as well as the resources required to train them, to give you a better understanding of the strengths and weaknesses of each model.This video is a must-watch for anyone interested in the latest advancements in natural language processing and AI.
On Large Language Models
for Understanding Human Language
Christopher Manning | Jul 19, 2022