Don Patrick - Oct 20, 2014:

I understand that that is your philosophy. I would say we just have different goals, mine being the functional application of reasoning and problem-solving, to which end conversational skill is a part of the process, but not the final outcome. Leonard Nimoy’s scientific expertise does not go beyond his script. I don’t wish to create the script but the more flexible scientific expertise. Even if we look only at the topic of conversational skill, I believe that a handful of underlying social rules can cover more ground than an abundance of case-by-case scripting. That is not to say it is necessarily better.

As for the more general question, I don’t think I am the most appropriate representative of humanity to answer it

I think we have the same goals. Your emphasis is on a bottom-up approach, where I want to go top-down. I don’t discount your methods, I just don’t think it is necessary to replicate human brain functions. It would be sufficient, but not necessary. I just believe that I can take any given desired result and work out a way of accomplishing it without ever having to study biology or physiology of the human brain.

You and others seem fixated on “case-by-case” scripting. But I did that 30 years ago. That is old school news to me. You have no idea what I’m studying and working on now. Just because I enter Turing tests, and run an educational demonstration of that, and continue to work with scripted mimetic synthesis, does not make me a relic or a one-dimensional clown. I believe that learning other people’s technologies keeps me fresh when I think about new ideas of my own. I am fully aware of the limits of AIML or JFRED.

I don’t place a lot of faith in isolated enclaves of factoids, 20 question competitions, or Winograd Schemas. When a bot can take something like the MMPI test and not come off as a sociopath, that will be another tool in the kit.

I know this gets off topic a bit with regards to suggesting books. But frankly, most of us are working beyond what’s been written. We are writing it now. “Time will no longer be a dark spot on our lungs. They will no longer say ‘you had to have been there’, because the fact is, Albin, we were.”—Shadow of the Vampire.

Robby Garner - Oct 21, 2014:

You and others seem fixated on “case-by-case” scripting. But I did that 30 years ago. That is old school news to me. You have no idea what I’m studying and working on now.

I meant no offence, but I can only go on what I do know about behaviourism, your past work and recent comments, and they showed no hints to a change of direction. As it turns out I do not know, I can make no comparison.
I could make three arguments why underlying psychological mechansisms matter, they all come down to a personal preference: That for many, the end result is not the purpose.

Robby Garner - Oct 21, 2014:

I just believe that I can take any given desired result and work out a way of accomplishing it without ever having to study biology or physiology of the human brain.

Oh wait, I thought you said “psychology”. I’m not concerned with the natural biological workings of the brain either, just the information processes, the psychology and thoughts. But as far as I know that differs from behaviourism:

The primary tenet of behaviorism is that psychology should concern itself with the observable behavior of people and animals, not with unobservable events that take place in their minds. The behaviorist school of thought maintains that behaviors as such can be described scientifically without recourse either to internal physiological events or to hypothetical constructs such as thoughts and beliefs.

That said, if you have any pointers on how to mimic conversational skills, you’re welcome to share.

Ah ha! We reach! I am just as interested in how human beings function in society. I don’t claim to have produced anything that will always pass a Turing test, or that always answers questions correctly.

I’m interested in ways to model what people do in real life.  I adopted the term “mimetic synthesis” as a way to describe my belief that intelligence can be built, one facet at a time. I don’t think there has to be any general “formula” or a single algorithm for intelligence.

Just like people learn to read, do maths, interact with other people, I believe that machine behaviors can be added together and integrated to perform as an individual entity, much the way robots have learned to balance and walk on two legs and do other tasks. I don’t see why we can’t build mimetic entities that can learn or be taught social norms that human beings would recognize as intelligent creatures.

Hi Robby! Nice to see you here again.

I always end up in conversational debates with other chatbot / AI enthusiasts as to which is better or more on point with regard to intellect, a decent conversational bot OR a bot that can answer a multitude of questions second only to Watson!

For me, I place a large portion of the intelligent “PIE” slices, toward the conversational bot and in doing so, I feel the bot’s ability to stay on topic and follow logical flow, is indicitive of an intelligence. I have chatted with at least a hundred or more bots in my lifetime and to me, it seems the ones that are able to “learn”, have to biggest appeal for me.

Scripted bots are nice but when they can start cataloging conversational snippets of data, user information like birthdays, family, like / dislikes, events, etc., they are the more personable, enjoyable and desirable.

Things have greatly changed since you won the prize many years ago. Processor speed, memory, clouds, accessibility, etc. but the underlying problem is still comprehension for the chatbot. Some people say that the bot doesn’t NEED to comprehend but rather to make the “Connection” as to which words indicates what. This is almost similar to pattern matching on a grander scale but it does have some validity and some followers.

Gone are the days of (no offense) programming your bot to be “clever” when trying to fool the judges. That approach simply doesn’t cut it any more. The questions are too off the wall to calculate and a bot’s cleverness thinly disguised by a cute or clever answer is all too soon discovered.

So, it the testing / contest flawed? What does it really test…the bot’s ability or the programmer’s cleverness?

They test to see if an artificial entity can FOOL some judges into thinking the bot was actually a person!

Why not test the chatbot for it’s ability to carry on a normal conversation…a discussion of various current events, weather, history, or life in general. Birth, death, parties, parades, music, emotions, jokes, etc.

Judge them on how well the chatbot follows the topical flow when the subject changes and of course record the answers for comparison with other bots.

In conversation lies intelligence, not simply knowing how many moons some planet has or if a car is larger than a breadbox. (perhaps a model car isn’t) but I digress.

Real life experiences deposit a great deal of insight in us, provided we are receptive to it. Since I’ve gotten much older, I have now realized that my brain can only hold so much information! Yes it has limits! Therefore, in order for me to acquire new information, I have to throw out some of the old to make room. This process is called “Selective Forgetting” and I’ve become pretty good at it! You just wait! You turn’s coming!

Have a conversation about that! (I made your Robitron logo for your web site many years ago).