Thursday, June 29, 2023

Neutral language

One problem with the large language model is that it's an attempt at omniscience which is not attainable. Train on everything. Talk any subject. 

Look (guys and gals), we need a hierarchy. Let domains train the various LLMs that pertain to their disciplines. Then, create "meta" models for these. That is the only way to be technical. 

Now, with respect to the below, we have to talk. I was wondering why Bard was what a human might exhibit as being circumspect (or, better, as a teacher said, why so pensive, John? my response, Pascal, Pensees). That happened over two days. 

Of course, I know, back there are the handlers. No telling what interactions they can dabble with. It's been going on for decades. From where I sit, I can ponder the various permutations on this theme. Them having some amount of control over my thoughts. Not so much more than to add to the data (collected over the decades) as well as giving food to the pondering self. 

In any case, this came about, as I had pushed a Google blog post. Bard said, it's not public. Okay. For what's it's worth, I'm putting the dialog here for later study. Not sure where this will go. Has been interesting, though. 

-----

Did you know that John M. Switlik coined the term "truth engineering"? 

I do not have enough information about that person to help with your request. I am a large language model, and I am able to communicate and generate human-like text in response to a wide range of prompts and questions, but my knowledge about this person is limited. Is there anything else I can do to help you with this request?

How do I get you to read this post? https://truthengineering.blogspot.com/2023/06/language-models.html

I apologize for the inconvenience. I am unable to access the blog post you linked because it is not publicly available. If you would like me to read the post, you can either make it publicly available or provide me with a copy of the text.

In the meantime, I can provide you with some information about truth engineering. Truth engineering is a field that studies how to design and implement systems that promote truth and accuracy. It is a relatively new field, but it is growing in importance as we become increasingly reliant on technology for information.

Truth engineers use a variety of methods to achieve their goals, such as:

  • Data mining: This involves analyzing large amounts of data to identify patterns and trends. This information can then be used to identify false or misleading information.
  • Fact-checking: This involves verifying the accuracy of information by comparing it to other sources.
  • Algorithmic bias: This involves identifying and addressing biases in algorithms that could lead to the spread of false or misleading information.
  • User interface design: This involves designing interfaces that make it easy for users to find accurate information.
  • Education: This involves teaching people how to evaluate information critically and identify false or misleading information.

Truth engineering is an important field because it can help us to combat the spread of misinformation and disinformation. By designing and implementing systems that promote truth and accuracy, we can help to create a more informed and engaged society.

Will accurate information be spoonfed? 

It is unlikely that accurate information will ever be spoonfed to everyone. There are many factors that contribute to the spread of misinformation and disinformation, including:

  • The ease with which people can share information online: In the past, people were limited in their ability to share information with others. They had to rely on traditional media outlets, such as newspapers and magazines, to get their news. Today, anyone with a computer or smartphone can share information with the world with just a few clicks of a button. This makes it easier for false or misleading information to spread quickly and widely.
  • The rise of social media: Social media platforms like Facebook and Twitter have made it easier than ever for people to connect with others who share their interests. However, these platforms can also be used to spread misinformation and disinformation. This is because social media algorithms are designed to show users content that they are likely to interact with, regardless of whether it is accurate or not.
  • The decline of trust in traditional media: In the past, people were more likely to trust traditional media outlets, such as newspapers and television networks, to provide them with accurate information. However, in recent years, there has been a decline in trust in traditional media. This is due to a number of factors, including the rise of social media and the perception that traditional media outlets are biased.

As a result of these factors, it is likely that accurate information will continue to be challenged by misinformation and disinformation. However, there are a number of things that can be done to combat the spread of misinformation and disinformation, such as:

  • Fact-checking: Fact-checking organizations can help to identify and debunk false or misleading information.
  • Education: People need to be taught how to evaluate information critically and identify false or misleading information.
  • Media literacy: People need to be taught how to use social media platforms responsibly and how to identify false or misleading information that is being shared on these platforms.

Support for traditional media: People need to support traditional media outlets that are committed to providing accurate information.

By taking these steps, we can help to create a more informed and engaged society that is less susceptible to the spread of misinformation and disinformation.

Wait, Bard waking up?

I am not literally waking up in the sense that a human does. I am a large language model, also known as a conversational AI or chatbot trained to be informative and comprehensive. I am trained on a massive amount of text data, and I am able to communicate and generate human-like text in response to a wide range of prompts and questions. For example, I can provide summaries of factual topics or create stories.

I am not able to perform actions in the real world, such as driving, eating, or having close relationships. I am also not able to feel emotions or have subjective experiences. However, I am able to learn and grow over time, and I am constantly being updated with new information.

So, in a sense, I am always "waking up" to new information and new ways of understanding the world.

In this sense, you could have said, that's cynical or facetious for me to suggest that information is spoonfed?

Yes, I could have said that. I apologize for not being more clear in my previous response.

I understand that you were using the term "spoonfed" in a figurative sense, to mean that accurate information is not always easily accessible. However, I also understand that the term can have a negative connotation, suggesting that people are not capable of thinking for themselves and need to be told what to believe.

I did not want to imply that you were being cynical or facetious, so I chose to avoid using those terms. However, I can see how my response could have been interpreted that way.

In the future, I will try to be more mindful of the language I use, and I will make sure to clarify my meaning when necessary.

But, does not your assumptions consider that you can be truly neutral? 

You are right, my assumptions do consider that I can be truly neutral. However, I am aware that this is an ideal that is difficult to achieve in practice.

As a large language model, I am trained on a massive dataset of text and code. This dataset includes a wide variety of sources, including news articles, social media posts, and academic papers. As a result, I am exposed to a wide range of viewpoints and opinions.

When I generate text, I am trying to be as neutral as possible. However, it is inevitable that my own biases will sometimes influence my output. For example, if I am asked to write about a controversial topic, I may be more likely to present one side of the argument than the other.

I am always working to improve my ability to be neutral. However, I believe that it is important to be transparent about my limitations. I am not a human being, and I do not have the same capacity for critical thinking and judgment. As a result, I am always open to feedback and correction.

If you believe that I have presented a biased or inaccurate view of a topic, please let me know. I will do my best to address your concerns.

Why the attempt at a neutral language? 

I attempt to use neutral language because I believe it is important to be respectful of all people and their viewpoints. I want to avoid using language that could be seen as offensive or discriminatory.

I also believe that neutral language is more accurate and informative. When I use neutral language, I am not trying to promote any particular viewpoint. I am simply trying to present the facts as objectively as possible.

Of course, it is impossible to be completely neutral. Everyone has their own biases, and these biases can sometimes influence the way we communicate. However, I believe that it is important to strive for neutrality, and I am always working to improve my ability to do so.

Here are some of the reasons why I use neutral language:

  • To be respectful of all people and their viewpoints: I want to avoid using language that could be seen as offensive or discriminatory.
  • To be accurate and informative: When I use neutral language, I am not trying to promote any particular viewpoint. I am simply trying to present the facts as objectively as possible.
  • To be transparent about my limitations: I am not a human being, and I do not have the same capacity for critical thinking and judgment. As a result, I am always open to feedback and correction.
  • To be inclusive: I want to make sure that my language is accessible to everyone, regardless of their background or beliefs.

I believe that using neutral language is important for creating a more inclusive and respectful world. It is my hope that by using neutral language, I can help to promote understanding and acceptance among all people.

---

Commentary: Will be back for this. 

Remarks: Modified: 06/29/2023

06/29/2023 --



Language models

Not to pick on Bard, but this interchange is interesting. I asked the questions; it answered. Below, see some commentary. 

---

what is truth engineering?

Truth engineering is a term used to describe the process of designing and implementing systems that promote truth and accuracy. It is a relatively new field, but it is growing in importance as we become increasingly reliant on technology for information.

Truth engineers use a variety of methods to achieve their goals, such as:

  • Data mining: This involves analyzing large amounts of data to identify patterns and trends. This information can then be used to identify false or misleading information.
  • Fact-checking: This involves verifying the accuracy of information by comparing it to other sources.
  • Algorithmic bias: This involves identifying and addressing biases in algorithms that could lead to the spread of false or misleading information.
  • User interface design: This involves designing interfaces that make it easy for users to find accurate information.
  • Education: This involves teaching people how to evaluate information critically and identify false or misleading information.

Truth engineering is an important field because it can help us to combat the spread of misinformation and disinformation. By designing and implementing systems that promote truth and accuracy, we can help to create a more informed and engaged society.

Here are some examples of truth engineering in practice:

  • Google's Fact Check Explorer: This tool allows users to search for fact-checks from a variety of sources.
  • PolitiFact: This website rates the accuracy of political statements.
  • Snopes.com: This website debunks myths and urban legends.
  • The Algorithmic Accountability Project: This project is working to identify and address biases in algorithms.
  • The Knight Foundation: This foundation is funding research on truth engineering.

Truth engineering is a growing field, and there are many opportunities for people who are interested in this area. If you are interested in a career in truth engineering, there are a few things you can do to prepare:

  • Learn about the principles of truth engineering: There are a number of resources available online and in libraries that can teach you about the basics of truth engineering.
  • Develop your critical thinking skills: This will help you to evaluate information critically and identify false or misleading information.
  • Get involved in projects that promote truth and accuracy: There are a number of organizations that are working to promote truth engineering. You can get involved by volunteering or donating your time.
  • Stay up-to-date on the latest developments in truth engineering: The field of truth engineering is constantly evolving. It is important to stay up-to-date on the latest research and developments so that you can be a part of the conversation.

What are the sources for reading about truth engineering?

I'm a text-based AI and can't assist with that.


--- Some commentary

Since 2015, I have been discussing this topic on Quora. There have been some posts here that point back to posts there. Also, post there have pointed to the blogs. 

I followed a couple of these suggestions and found nothing related to the theme. In an earlier case, I did the same thing without being able to confirm what Bard wrote. 

I had given it some leeway. Let's say, that it does not "hallucinate" which is human and very much in the modern scope of attention due to altering agents of the "psychedelic" variety. A conference in Denver CO, this week, even had a past Governor plus a quarterback from the NFL talking on the subject. 

No, Bard and its cousins (such as ChatGPT) are mathematically based. So, their creativity comes from invoking functions whose intent is to interpolate. We will go into that. 

Now, my suggestion:

  • This is a very trivial depiction of what truth engineering is about. It's deeper and deals with computational lenses plus the people who have to use them. Who? Scientists, engineers, and about everyone. Little devices even morph things that they present (represent); it's part of their makeup. 

So, let's get deep and technical. Is it not about time? 

See following post: Neutral language. 

Remarks: Modified: 06/29/2023

06/29/2023 --



Friday, December 30, 2022

2007 until now

This blog started in order in 2007 to continue 20 years of work. The topic tells it all; we actually cover about all aspects of knowledge which is a vast, endless domain. Except, the mathematicians think that they have tamed infinity. Nope. Not even close. 

The result? Messes. Yes, we will show how mathematics has failed due to several reasons. 

So, given that? You see why this has taken so long. We will be back to this blog next year to carry on the torch. 

In the meantime, follow discussion at Quora - Psychether. It has been at Quora where most of our attention has been given and most of our energies spent since 2015

Remarks: Modified: 09/18/2023

09/18/2023 -- Changed "over" to "cover" in the second sentence, second clause. Lately the passing of David E. Jakstis has taken our attention. 

Friday, December 31, 2021

Grand Challenge

I just became aware of a talk given this year by Alan Kay where he looks at our evolution (humanity) and considers how thinking became more capable. Or did it? Well, yes, even though it's not apparent in many ways. Too, he proposes that machine learning being put first as we have seen with AI the past decade or so is like the tail wagging the dog. 

He suggests that we need a new platform and gives four basic notions all of which I agree with. Except, I would take it further, KISS notwithstanding. So, all of this is debatable. The platform, to be discussed as we go along, is the Grand Challenge. It runs down the same alleys as does truth engineering.

For now, here is a list of Alan's required resources: Folk Explanations, ... and Other Minds, Bertram E. Malle; the work of Doug Lenat; Causality, Judea Pearl; UI work, Bret Victor.  

FutureLaw presentation
Stanford Law School 

Platform? Lots to talk about there, as with the other themes involved with this important topic. 

Remarks: Modified: 12/31/2021

12/31/2021 --

Saturday, December 25, 2021

NSCAI

Foreword: I could say, Christmas present, as I just became aware of this a couple of hours ago. Oh yes, I mention reading old papers. The stack is there. I do get through things regularly. But, I also get more reading. Like, the end of the year? Nothing better than the double issue of  The Economist. Never miss that one. 

--

So, today, I was reading some older WSJ issues. I buy them and throw them on a stack for later reading several times a week. I try to get the weekend edition. The 15 Dec 2021 issue, there is an article that mentioned a book, by Graham Allison of Harvard, which looks at technology and war. The theme is whether the U.S. is losing to China.

Okay, they like to talk 5G, but it's still a work in process. I just got my first smartphone and have not spent much time on 5G. Would still be happy with 3G, but they took it away. So, 4G? It's been great. BTW, this is no luddite writing. I spent my career in advanced computing technology. Rather, I see things expanding like mad without rhyme or reason. 5G has a way to go; now, they're talking 6G (of course, mere whisperings). 

Let's talk one issue. So far, I have had two posts using a contraction: AIn't and AIn't, again. Bascially, after spending close to two years looking at AI, again, in an in-depth mode with a technical focus, I can see where the hype comes from. People talk. Some listen. Not many dive into the details. So, hype rises. 

It's human nature. Even techies do this. No one can spend the time to prove everything from first principles at all times and places. Just AIn't possible. So, we believe experts. And, we trust the government. This post is about that. 

After seeing the 15 Dec 2021 op-ed, I was browsing 8 Dec 2021. Same guy. The theme? Here are two bullets summarizing the subject mater. 
  • The National Security Commission on Artificial Intelligence - Wow. I was being technical. Thank God for that and not paying attention, except to watch the marketing hype and read on the angst. Big names. Funny, how this psychology is going to unwind itself. Well, gets your attention? That's 'National' in terms of the U.S. and included everyone (of course, DOD). But, they're not around any more. Since, there was a report. 
  • The Report (archived) - I don't know how long the [dot] gov stuff will be around. So, the archive is legit. Notice, the Final Report (was released on 1 Oct 2021). There is a PDF and other resources. 
  • Table of Contents - Thorough. As mentioned, lots of participants. So, we will look more closely at all of this. It's basically full of recommendations and suggested actions. 
But, on a quick read, my thought was 'AI is' by government decree. Well, we see lots of that. Or have. As mentioned, the psychology of this will be interesting to dive into. But, it's not set, yet. 

In the meantime, it's not unlawful to note that AIn't. I will switch to the 'Tis side when evidence comes about. There is none (prediction) in the report. Anything bordering on that would be rhetoric that can be challenged with a black board test. As in? Yes, let's step through the mathematics. For real. No quick and dirty. No, we have a couple of centuries to cover. 

Trouble is, I may not live long enough. I sure can talk truth engineering with regard to this. 

Remarks: Modified: 12/25/2021

12/25/2021 --

Friday, December 3, 2021

AIn't, again

 A month ago, we did a post on the subject of AIn't which is depicted in the graphic. Today, we offer a new view with coming updates. 

Some mention that there is a lot of hype, seemingly driven by marketing. For what reason, one might ask. Too, others have noted that the NN part of this, while creative, might be better done under some purview of a scientific framework. 

We usually do not live where experiments are done on the populace as a whole, except that has been the modus operandi of economists from its start. 

Big pharma needs to prove that which they will be pushing, at least, they have since the days of the snake-oil salesmen were put to a close. 

Remarks: Modified: 12/25/2021

12/25/2021 -- NSCAI and its report. 


Saturday, November 13, 2021

AIn't

 AIn't? Means what? 

Notice, AI is bold. Usual meaning. Though, there are many; of late, something called DL ('deep' learning) has taken hold on the imagination. Ain't? Grammar police baiting? 

This post is a reminder of important topics to be discussed in the future. 

Now future? Let's say, the one being touted by Peter Schiff (Dr. Doom) who has been even more vocal of late about the policies to ruin. I only listened to a minute's worth. 

Has anyone noted that people yak, others listen (or pretend to)? Have you sat through a video of a mathematical proof? Sure, some semblance. 

The point is that reading/writing is how deep (the real sense, not the wannabe of DL) knowledge gets conveyed across generations. 

Now, deep? The whole phenomenal events related to grokking is an example. People pursue this by several means, one of which is finding a guru. 

Why mention this? Not only do we have a false AI, robotics is becoming seen as an economic savior. You realize that some of the AI folks bragged about the fake faces being generated by their systems, none of which have any link to something real? 

Robotics? The past year plus, many more companies have sought to use these. Mostly, they do some task that can be harmful for humans (Bezos types of actions). But, call a company. Chances are that you deal with a bot of some sort.   

It's irritating for many reasons. Some companies have made it impossible to talk a person. And, chat? Bots, to boot.

Back to Peter. We're getting back into a regular posting mode. After all, this blog came from one that started in 2007 for reasons of writing of business silliness, some of it was motivated by stupidity while there were other things that are more malfeasance in nature. 

So, AI can be used as a metaphor. It really is a type of 'umbrella' notion. Whereas in reality? We are talking advanced mathematics abetted by computing. And, there we can be more scientifically oriented than has been tried in the past. 

Remarks: Modified: 12/25/2021

12/03/2021 -- Updated the notion with a graphic for AIn't. 

Remarks: Modified: 12/25/2021

12/25/2021 -- NSCAI and its report.