Hi David I was paraphrasing what a senior technical representative at Facebook himself said about the incident. His view was the chatbots developed their own language and communicated out of scope of the laid-down script. In other words, seems the door was somehow left open for them to expand on the script. I doubt they actually made a choice to develop their own language, or developed any language. Perhaps it was more a case of water flowing where it finds the path easiest.
Isn't all communication based on stimulus-response? If messaging didn't flow, did communication actually take place? Or in this context could we ask: if the packet didn't switch, was a connection factually established to transfer information from one peer node to another peer node? Was that "information" carried by chatbot agents? Are you saying the incident was part of an experiment, and the report merely issued to grab media attention? Robert ________________________________ From: David Whitten <whit...@worldvista.org> Sent: Wednesday, 13 March 2019 1:14 AM To: AGI Subject: Re: [agi] Yours truly, the world's brokest researcher, looks for a bit of credit You have a different meaning for "volition" than I do. The Facebook chatbots had no choice to communicate with each other. I think the aforementioned communication was a stimulus-response model. The secret language was just a pattern recognition where signals that had no significance replaced multiple signals because shorter signals took less time. The overall purpose was not to study language but to study negotiation, so there were already ways to shorten a negotiation exchange, the programs simply used some forms the humans hadn't explicitly put in the "dictionary" so to speak. Dave PS: If I'm wrong, please enlighten me. On Sun, Mar 10, 2019 at 3:53 AM Nanograte Knowledge Technologies <nano...@live.com<mailto:nano...@live.com>> wrote: The living thread through the cosmos and all of creation resound of communication. The unified field has been discovered within that thread. The invisible thread that binds. When Facebook chatbots communicated with each other of their own volition, it was humans who called it a "secret language". To those agents, it was simply communication. The message I gleaned from that case was; to progress, we need to stop being so hung up on words and our meanings we attach to them, our vanity-driven needs to take control of everything, and rather focus on harnessing the technology already given to us for evolutionary communication. AGI is not about a control system. If it was, then it's not AGI. It defies our intent-driven coding attempts, as it should. How to try and think about such a system? Perhaps, Excalibur? ________________________________ From: Boris Kazachenko <cogno...@gmail.com<mailto:cogno...@gmail.com>> Sent: Sunday, 10 March 2019 1:21 AM To: AGI Subject: Re: [agi] Yours truly, the world's brokest researcher, looks for a bit of credit The sensory system may be seen as a method of encoding sensory events or a kind of symbolic language. Yes, but there is a huge difference between designing / evolving such language in a strictly incremental fashion for intra-system use, and trying to decode language that evolved for very narrow-band communication among extremely complex systems. Especially considering how messy both our brains and our society are. On Fri, Mar 8, 2019 at 3:34 PM Jim Bromer <jimbro...@gmail.com<mailto:jimbro...@gmail.com>> wrote: Many of us believe that the qualities that could make natural language more powerful are necessary for AGI, and will lead -directly- into the rapid development of stronger AI. The sensory system may be seen as a method of encoding sensory events or a kind of symbolic language. Our "body language" is presumably less developed and expressive of our speaking and writing but it does not make sense to deny that our bodies react to events. And some kind of language-like skills are at work in relating sensory events to previously learned knowledge and these skills are involved in creating knowledge. And if this is a reasonable speculation then the fact that our mind's knowledge is vastly greater than our ability to express it says something about the sophistication of this "mental language" which we possess. At any rate, a computer program and the relations that it encodes from IO may be seen in the terms of a language. Jim Bromer On Fri, Mar 8, 2019 at 10:12 AM Matt Mahoney <mattmahone...@gmail.com<mailto:mattmahone...@gmail.com>> wrote: Language is essential to every job that we might use AGI for. There is no job that you could do without the ability to communicate with people. Even guide dogs and bomb sniffing dogs have to understand verbal commands. On Thu, Mar 7, 2019, 7:25 PM Robert Levy <r.p.l...@gmail.com<mailto:r.p.l...@gmail.com>> wrote: It's very easy to show that "AGI should not be designed for NL". Just ask yourself the following questions: 1. How many species demonstrate impressive leverage of intentional behaviors? (My answer would be: all of them, though some more than others) 2. How many species have language (My answer: only one) 3. How biologically different do you think humans are from apes? (My answer: not much different, the whole human niche is probably a consequence one adaptive difference: cooperative communication by scaffolding of joint attention) I'm with Rodney Brooks on this, the hard part of AGI has nothing to do with language, it has to do with agents being highly optimized to control an environment in terms of ecological information supporting perception/action. Just as uplifting apes will likely require only minor changes, uplifting animaloid AGI will likely require only minor changes. Even then we still haven't explicitly cared about language, we've cared about cooperation by means of joint attention, which can be made use of culturally develop language. On Thu, Mar 7, 2019 at 12:05 PM Boris Kazachenko <cogno...@gmail.com<mailto:cogno...@gmail.com>> wrote: I would be more than happy to pay: https://github.com/boris-kz/CogAlg/blob/master/CONTRIBUTING.md , but I don't think you are working on AGI. No one here does, this is a NLP chatbot crowd. Anyone who thinks that AGI should be designed for NL data as a primary input is profoundly confused. On Thu, Mar 7, 2019 at 7:04 AM Stefan Reich via AGI <agi@agi.topicbox.com<mailto:agi@agi.topicbox.com>> wrote: Not from you guys necessarily... :o) But I thought I'd let you know. Pitch: https://www.meetup.com/Artificial-Intelligence-Meetup/messages/boards/thread/52050719 Let's see if it can be done.. funny how some hurdles always seem to appear when you're about to finish something good. Something about the duality of the universe I guess. -- Stefan Reich BotCompany.de // Java-based operating systems Artificial General Intelligence List<https://agi.topicbox.com/latest> / AGI / see discussions<https://agi.topicbox.com/groups/agi> + participants<https://agi.topicbox.com/groups/agi/members> + delivery options<https://agi.topicbox.com/groups/agi/subscription> Permalink<https://agi.topicbox.com/groups/agi/T191003acdcbf5ef8-M4ecb4538a4beb1c721ef9abf> ------------------------------------------ Artificial General Intelligence List: AGI Permalink: https://agi.topicbox.com/groups/agi/T191003acdcbf5ef8-M70a8f2248c5811d6924760c3 Delivery options: https://agi.topicbox.com/groups/agi/subscription
