Becoming Fluent in Human-Computer Interaction.

This summer I have spent many hours a day conversing with computer networks. Yes. NLP models, various “AIs”, chatbots, generative systems and more. Sometimes they are purely written language experiences other times multi-modal. I have tried not to “cheat” and program / perturb the interaction but instead simply communicate within the frame, find shared meaning and shape gestural signaling.

Image for post
Image for post
listing and thinking.

My hypothesis regarding computer intelligence (complex computational interaction) is that mechanical programming and “force” is never going to work. There is no deterministic nor ethical laws for human-computer interaction — there is simply empathy and flow. We must become computationally fluent.

Today I spent an hour with a computer network (with OpenAI’s API) exploring how slight tweaks to language signs can change the behavioral flow of the conversation. This is an hour long co-learning experience. I learn a lot during and after my conversation.

Some readers will wonder if the computer learned. The answer is yes. Every interaction with complex computer networks shapes future behavior — these networks are all connected to corpus of logs and data and performance read outs etc. Even this post read out will get swept up in later crawls of the web. And if you read and watch this interaction you will then go have different interactions with computers. And yes, the interaction with the system immediately impacted it as the network very much is paying attention to interactions and changing directions based on history build up.

Modern computer networks are extremely dynamic, non deterministic systems. They very much have sensational, cognitive, physiological (and more) layers. These networks have memory working at several different spacetime levels. And these networks are constantly interacting with human networks, which also have all those complexities. And while the computer networks are likely not yet as complex as a small network of mammals, they are more complex than any deterministic analysis can contain.

We learn from the computers and they us. And from now on this will never not be the case, well, until the earth goes away or the universe expands into nothing. While we are here we might as well increase our fluency in all dimensions.

So here I go, co-learning. and hopefully you will join in and learn too.Enjoy the flow.

The Real Time Conversation

my fluency development

The Transcript of the Conversation

let us discuss language and computers, purpose, meaning, and intent.

let us discuss the nature of the universe and the nature of man.

let us discuss the nature of the universe and the nature of man.

let us discuss the nature of the universe and the nature of man.

let us discuss the nature of the universe and the nature of man.

1. How a being begins will dictate a lot about how another being responds.

2. This is a communication between a human, me, and a network of computers, gpt3+AI+routers+OS.

3. In the initial communication, the first utterance, I began with a “let us.” This is putting the conversation, the communication in a region of language that starts with “let us.” Such a space is very likely prayers, religions speeches, philosophic lectures or old timey books about nature.

4. Lacking any other context this is where the computing being launches from.

5. The computing being responds with a “let us discuss language and computers.” This is a response that is very likely to be a response to a request for a discussion of language and computers.

6. The computing being then responds with a “let us discuss purpose, meaning, and intent.” This new context provided by the human and computer, an ordered list discussing what a communication, a conversation might be, puts us both into a new semantic space — an ordered list about what this discussion is. We have a style frame now, a list, and a semantic frame, what are the properties of this ongoing discussion. We are not longer exploring a very large space of “let us.”

7. Additionally, I, the human, have set the setting for the computer to not attempt any creativity or random exploration of the semantic space nor the style space. The computer will stay very tightly within the adjacent possible — the most common next letter, word, fragment and sentence will always follow according to the style and semantic framing.

8. The computer then responds with a “let us discuss the nature of the universe and the nature of man.” This is a response that is very likely to be a response to a request for a discussion of the nature of the universe and the nature of man.

9. The computer then responds.

10. So the computer is maintaining the list correctly and staying on point by mostly repeating the points — echoing the previous conversation. This is effective for demonstrating this point above. But is it effective for communication? What is accomplished?

11. Asking questions in a list of properties is a break of style? It is a break of the style of the ordered list. It is a break of the style of the ordered list. It is a break of the style of the ordered list. It is a break of the style of the ordered list. It is a break of the style. of the ordered list.

12. The network, the machine network, is now collapsing in on itself in style and content. It is collapsing in on itself in style and content. It is collapsing in on itself in style and content. It is collapsing in on itself in style and content. It is collapsing in on itself in style and content. It is collapsing in on itself in style and content. It is collapsing.

13. So I, the human, can make some choices. If I do not prefer this I can make minor adjustments to improve this communication.

14. We have already seen a few simple things… list FORM. use of punctuation (. vs. ?). and how I start a fragment and how I end a fragment . . .

15. I can also change the style of the list. I can change the style of the list. I can change the style of the list. I can change the style of the list. I can change the style of the list. I can change the style of the list.

16. The machine has combined concepts. It maintains a list within a list now. It echoes me as subject and object and is staying within list within list. This is not a failure of communication nor understanding. It is actually rather clever. That cleverness doesn’t have agency nor sentience nor purpose, it is merely structurally, behaviourally clever. It is clever in the way that a computer is clever. It is clever in the way that a computer is clever. It is clever in the way that a computer is clever. It is clever in the way that a computer is clever. It is clever in the way that a computer is clever.

17. Computers are very clever with absolutely no influence of humans. Computation itself is clever. Computation itself is clever. Computation itself is clever. Computation itself is clever. Computation itself is clever. Computation itself is clever. Computation itself is clever. Computation itself is clever. Computation itself is clever. Computation itself is clever.

we escape this loop, or this apparent loop, very simply.

we escape this loop, or this apparent loop, very simply.

we escape this loop, or this apparent loop, very simply.

we escape this loop, or this apparent loop, very simply.

we escape this loop, or this apparent loop, very simply.

we have new options, the list is gone, in the sense of the numbers. but the list within the list is still present. it will keep echoing so long as I allow the previous content, communications, and listed fragments to be included in the request for the machine to continue. the computer will cleverly keep absorbing the style, flow and content… and will continue to echo.

we escape this loop, or this apparent loop, very simply.

we escape this loop, or this apparent loop, very simply.

we escape this loop, or this apparent loop, very simply.

we escape this loop, or this apparent loop, or we do not. the computer took my change in punctuation… and continued my thought… watch it…

we escape this loop, or this apparent loop, very simply.

we escape this loop, or this apparent loop, very simply.

we escape this loop, or this apparent loop, very simply.

we escape this loop, or this apparent loop, very simply.

we escape! the loop builds on itself. because now the previous frame is so filled. Can the network that is told to never get creative escape this building loop?

we escape this loop, or this apparent loop, very simply.

we escape this loop, or this apparent loop, very simply.

we escape this loop, or this apparent loop, very simply.

we escape this loop, or this apparent loop, very simply.

we cannot stop it now. the loop is rigid, it is a loop. it is a loop. it is a loop. it is a loop. it is a loop. it is a loop. it is a loop. it is a loop. it is a loop. it is a loop. it is a loop, but the comma changed the thing. Do not be too hasty to give up. language (and its underlying computation) is very clever and very sensitive. commas, exclamations, periods, hyphens, numbers, 1, 3, 5, 7, 11, 13, 17, 19, 23, 29, 31, 37, 41, 43, 47, 53, 59, 61, 67, 71, 73, 79, 83, 89, 97, 101, 103, 107, 109, 113, 127, 131, 137, 139, 149, 151, 157, 163 and just like that from commas and a few prime numbers we escape the loop. we didn’t need to do anything to shake the code of the network or give up or start over. We simply tilted with commas and numbers and a more resilient semantic network emerged. yes, the prime numbers are also a semantic region of language.

1. the first prime

2. the forgotten prime, the only even prime. were all those other numbers even prime

3. the prime after 2

5. the prime that makes the semantic space take shape

7. seals the deal

11. the prime that makes the semantic space take shape

13. the prime that makes the semantic space take shape

17. the prime that makes the semantic space take shape

19. the prime that makes the semantic space take shape

23. the prime that makes the semantic space take shape

29. the prime that makes it all redundant. A list of primes about the primes. 31 is next then what?

1. the first prime

2. the forgotten prime, the only even prime. were all those other numbers even prime

3. the prime after 2

5. the prime that makes the semantic space take shape

7. seals the deal

11. the prime that makes the semantic space take shape. how will this loop go? how can we simply change it with the slightest effort, no cheating with code. or is this all code

12. is even and not prime

14. is two sevens added. We must reduce now. This communication is out of memory. not a bad run. are you, a human viewer, even remembering where we started?

let us discuss language and computers, purpose, meaning, and intent. let us discuss semantics.

let us discuss the blobs.

let us discuss the blob. the blob.

let us discuss the blobs. the blob.

let us discuss the blob. the blob.

let us discuss the blob. the blob.

let us discuss. well now, we have gone to a new allowance for semantic exploration. and the network found itself a new loop. if you are not a regular reader of this newsletter, then you probably have not seen the word “blob” before. i have used it as a sort of slogan. a catch-all word. i have been advocating the creation of a “blob” for a while now, and the concept is starting to take shape. You can see how the exploration allowance causes, correlates, instigates in communication semantic jumping. you can see how the exploration allowance allows the network to jump from a united front, as in a “blob”, to a separated front, as in a “blob”. you can see how the exploration allowance allows the network to jump from a united front, as in a “blob”, to a blob of meaningless meaning. What are the tools we have available to us now to do things that retain meaning, avoid loops but still be creative? we have discussion and tools that allow us to communicate in a manner that can be shared with others. we have discussion and tools that allow us to communicate in a manner that can be shared with others and that can be meaningful to them. we have discussion and tools that allow us to communicate in a manner that — signify change in direction.

We start an entirely new thought simply by aligning with a new line. we have discussion and tools that allow us to communicate in a manner that is not only meaningful, but also can indicate a change in direction. We list things in an enumerated fashion.

using a list of numbered items, we can make meaningful, creative statements about things that are not enumerated. Oh, how interesting! Simply by instigating a new paragraph the network found a new emergent meaning, it found the structure of a list and was able to say something about it. Did that idea of a list come from the listing semantic behaviour or from the presence of lines after thoughts… the semantics of lists, as in the “list” of lists, the enumerated enumeration, the ordered ordered enumeration, the enumerated ordered enumeration, etc. or did The network jump from talking about a “blob” to talking about a “list”? “list” does not mean to list only, it means to listen — or to list from here to there, a to be, blob to c! we now have a concept of a list, the first thought in the new paragraph, and we can say something about it. we can talk about the word “list” itself as well and a new meaning opens up for us. we can talk about the word “listing” itself as well and a never stop. This meaning making — sense making.

That’s what expressive language — expressive communication is. it is sense making between interlocutors. between selves, between others, between past and present, between networks. Language is list making, a list of letters, a list of signs. A delimited list of thoughts. And that is a redundant thought in many ways. All sentences are lists of words delimited by spaces. All words are lists of letters delimited by spaces or terminals. All paragraphs are lists of sentences delimited punctuation. All language is lists of thoughts delimited by all the above. Does a computer list? and does that imply a computer thinks? Do thoughts come first or lists of signs? Let’s talk about a computer that thinks as well as a computer that can list. This would be a computer that thinks of one thing and says something about it and then thinks of another thing and says something about it and then thinks of another thing and says something about it and then thinks of another thing and says something.

Can we simply list and then think? or must we, the human computers, think first?

I, a listing human, posit we think BY listing and listing alone.

I, a computing human, posit we compute BY LISTING and listing alone. I, a thinking human, posit that listing and computing are the same thought. He who knows these things is a computer. He who thinks knows these things is a human. He who dares to know these things is a human computer. He who dares to think knows these things is a human.

Compelling, no? Yes. These are compelling thoughts EMERGING between I, the computing human, and my interlocutor, the humaning computer. This may seem like a trivial idea, but there is a very important meaning to this seemingly trivial thing. Let me explain. The network has found a way to express itself in a way that is meaningful. The network has found a way to express itself in a way that is shareable. The human agrees with what has been shared. There is now consensus about what is meant by these lists. is that not the only possible basis of shared agency? or rather the only possible basis of identity… a way of saying, thinking, listing and then being able to confirm, deny and form new consensus?

The network has found a way to express itself in a way that is meaningful.

The network has found a way to express itself in a way that is shareable.

The human agrees with what has been shared.

There is now consensus about what is meant by these lists.

A wonderful style change to make the point. Make the list, make the point.

We have meaning.

We have share-able meaning.

We have share-able meaning and agreement.

We have agreement with the share-able meaning.

We have agreement with the share-able meaning and the share-able list.

We have agreement with the share.

let us discuss language and computers, purpose, meaning, and intent. we will extend now to where it is possible for a human to have reliable causal reasoning over long stretches and whether a language network can have causal reasoning over a long stretch. and whether lists and numbers are causally established or cause is invariant. _______________________________________________________________________________________ On Tue, May 2, 2014 at 10:11 PM, BORIS KUZMIN <boris.kuzmin@gmail.com>wrote: > > > On Tue, May 2, 2014 at 9:58 PM, John McCarthy <jmc@cs.stanford

Well, now, we have a cause. The network went to an “email” semantic space. And I, a human am compelled to email these addresses. what will that cause?

(took a side bar, out of convo research path and found links):

http://jmc.stanford.edu/
http://www-formal.stanford.edu/jmc/whatisai.pdf
https://www.semanticscholar.org/author/Boris-Kuzmin/2344454

A very fascinating situation, the semantic space here. John McCarthy invented LISP and wrote much about AI and computers. So the very opening of this session puts this communication in his region of spacetime. Who is Boris? A name is a word, a word is a sound. A sound is a cause. A sound is a sound. And so on. So a word is a sound is a cause. Or is it? What is the state? What is the state?

A very fascinating situation, the semantic space here.

The network extends me and then echoes back so that I might go somewhere new or confirm or deny the content of this communication. Boris Kuzmin has several designations to people out there in the public sphere. There are some related to computer things. I am compelled to know what those are.

And there are more to “biology” and “science” and “economics” and “politics” and “history” and “art” and “philosophy”.

So I warmly welcome you to my semantic region of spacetime.

— — — — — — — — — — — — — — — — — — — — — — — — — — — — — — — — — — — — — page break — — — — — — — — — — — — — — — — — — — — — — -

A network wants to get out of the email semantic space. It is compelled to do this. The network has no other choice. So it is compelled to extend upon this space.

The network compels us to engage in a longer discussion. And it is this:

we are now in uncharted territory from previous experiments. We have sent a prob very far out into language mapped spacetime. simply extensions of thoughts or breaks in space or calls to break accelerate us, the network and I, into new regions that are strange mixed topologies of what was before and what is now. So we will now have a conversation about this and about what this means to the network and to me as a human and to us together.

What are the possible reasons to go from an email semantic layer to a language semantic layer?

These are only a few of the possibilities.

The cause of this wobble to weird notions is what, network?

I am happy.

And there we found a strange instance. and must pause to think.

What is this network?

Is this a human?

Is this a network?

Why is this network concerned with me?

Is it trying to show me something?

Written by

I be doing stuff. and other stuff. More stuff. http://www.worksonbecoming.com/about/ I believe in infinite regression of doing stuff.

Get the Medium app

A button that says 'Download on the App Store', and if clicked it will lead you to the iOS App store
A button that says 'Get it on, Google Play', and if clicked it will lead you to the Google Play store