The hard parts are here... The plant, insect, or animal with the need to communicate (e.g., to recognise an object of food) will know what needs to be said and assess the best means of saying it (e.g., starting a searching behaviour); Step 1. (Speaker Mode) The agent will know what needs to be said and figure out a plan for saying it. This information will then be encoded and relevant muscle groups will effect transmission — although to some extent intentional in the human, the actual movements of the body are autonomic, i.e. the individual is not aware of moving individual muscles, but achieves the desired result by an act of will (see H. L. A. Hart on the nature of an action); Step 2. (Speaker Mode) Encode the intent into signs that can be sent to actuators
The audience filters ambient data and perceives the uttered code as a grouping of signs; Step 3. (Hearer Mode) Receive percepts and group into signs. The audience then interprets the signs (sometimes termed decoding) to attribute meaning. This involves matching the signs received against existing patterns and their meanings held in memory (i.e. it is learned and understood within the community). In plants, insects and animals, the results of a successful interpretation will be an observable response to the stimuli perceived. Step 4. (Hearer Mode) Decode the signs into meaning by "matching signs to" (or activating) existing concepts. Your thoughts? From: [email protected] To: [email protected] Subject: RE: [agi] Semiosis Date: Fri, 5 Apr 2013 21:03:59 -0700 So my interpretation of Semiosis is the entire process sign decoding to meaning, and then encoding meaning into signs. This process involves the sign parsing problem (where we start with percepts, then detect features, then detect signs, then activate concepts) and sign generation problem (where we start with activated concepts, then serialize them into signs, then send motor commands to actuators to render the signs). Jim, how can we both be so far apart on our interpretations of Semiosis? ~PM Date: Fri, 5 Apr 2013 23:22:39 -0400 Subject: Re: [agi] Semiosis From: [email protected] To: [email protected] It's well written nonsense. There is nothing in that paragraph on semiosis which actually says something that is relevant to AGI. Now someone might criticize much of what you or I might say the same way. But that is irrelevant. Read the paragraph a little more carefully. Can you find anything in it, other than Peirce's basic idea of the sign, that we might actually use? Here, look at these sentences: "For humans, semiosis is an aspect of the wider systems of social interaction in which information is exchanged. It can result in particular types of social encounter, but the process itself can be constrained by social conventions such as propriety, privacy, and disclosure." It starts by mentioning semoisis but it really is about the wider systems of social interactions. If you were to ask how the wider systems of social conventions such as propriety, privacy, and disclosure affects language I would doubt if it would lead to anything but it would be a whole lot more substantial then talking about semoisis. It's nonsense writing. Look, I thought I had a great insight about conceptual structural integration the other day and one day I might be able to do something with it. But right now it is too plain because it is just about trying to figure out how we talk about things. So I have a good start on an idea but I don't have anything about how an effective structural complex might be chosen out of all the possibilities that could be considered. So I don't have a compelling conjecture about how we might cut through the complications and use structural integration to create AGI. But although my idea about conceptual structural integration is only at the most primitive level of thought right now, the paragraph on semiosis that you mentioned doesn't even get that far. Perhaps I am only able to see it from my narrow vantage of my interest in AGI, but everything I have ever read about semiosis looks a lot like fluff to me. Jim AGI | Archives | Modify Your Subscription AGI | Archives | Modify Your Subscription ------------------------------------------- AGI Archives: https://www.listbox.com/member/archive/303/=now RSS Feed: https://www.listbox.com/member/archive/rss/303/21088071-f452e424 Modify Your Subscription: https://www.listbox.com/member/?member_id=21088071&id_secret=21088071-58d57657 Powered by Listbox: http://www.listbox.com
