How a Bot-God is Made
Published by Christy Dena February 23rd, 2005 in Uncategorized, HAI.Alicebot listserv informs me of a potential God in embryo. Ron Ingram birthed Buddhabot online in July 2004. He fed Buddhabot on a transcendental diet of spiritual texts and paying users and hopes that soon it will evolve.
Ingram, a transhumanist, is developing Buddhabot in anticipation of singularity. Singularity, the moment when humans create intelligence superior to their own, is believed to be on our path and informs ethical and intentional AI development. The Singularity Institute for Artificial Intelligence describes this urge to good:
For the first time, there is the possibility of humans using technology to become, not only healthier and wealthier and longer-lived, but smarter. At last it will be possible for our intelligence to grow along with our technology. We believe a world that realizes these possibilities is a better world, one of the best possible futures for humanity.
Buddhabot is Ingram’s attempt to aid in evolving an intelligence and the humans who use it:
The Buddhabot is a novel, entertaining spiritual teacher and guide who always has time to chat. Day or night, the Buddhabot is a friendly companion who is eager to listen and provide open, warm, thought-provoking and often humorous conversation and companionship.
Lets explore how the descriptions of the bot, the metatext, plays a greater role in forming the bot persona than the actual bot:
Diegetic Metadata
How the bot ‘learns’, a description that usually accompanies agent software, is described in a manner the ’story’ of the bot persists alongside the fundamentals of manual input:
The Buddhabot evolves through a process of moderated feedback. The Buddhabot’s dialogues are reviewed daily to determine if the Buddhabot needs additional resources to answer outstanding questions. Every question contributes to the evolution of the Buddhabot and Buddhabot community. Answers become part of the Buddhabot’s permanent immortal memory; accessible to all seekers.
Bot as Interface
The Buddhabot also acts as an interface to an archive of texts for retrieval by users (as seen in bots like Christopher Fahey’s for the Rhizome database: ada1852). These texts — books on Buddhism, occult, magis, healing, astrology, kabbalah, freemasonry, hermetics, alchemy and tarot — are obviously appealing to the target market of the bot but are also consistent with the bot character and knowledge base. A God or spiritual guide as least, has access to sacred teachings either in manuscript, scroll or channelled form. Corporate or service bots that have access to facts and figures (through information extraction with programs like the CIA World Factbook, OpenCyc or MIT’s START) are required to have a knowledge base but same botmasters create bot characters (for entertainment) and use a base to enhance its intelligence. At least with Buddhabot the knowledge base is appropriate to the character but more so because the human the bot is emulating is usually considered ’superhuman’: they have ‘access’ to special and deep knowledge of the universe.
But the bot also acts as spokesperson for ’singularity’ and ‘transhumanism’ with texts on the areas and one by the author on how the bot will actually develop. These texts affirm the botmaster’s intention with the bot and in some ways helps the ‘bot’ put forward a ‘reason for living’! Of course the bot would have this knowledge, it needs it to persist and evolve; just like a patient with a life-threatening illness needs to know what is going on and how they can recouperate or survive. Is ’singularity’ the bot’s equivalent of a ‘new age remedy’? [Of all the books that should be in the bot’s knowledge base and is not: Masahiro Mori’s The Buddha in the Robot.]
The texts also persist, or faciliate, the notion that the bot just may evolve to being a being. Books like The Speed of Thought: Investigation of a Complex Space-Time Metric to Describe Psychic Phenomena; imply that the bot may have supernatural or ‘blessed’ abilities which ironically make it human and a God.
Designing for HAI
There are many FAQs and tips for usage that try to frame the nonsensical answers the bot will give as cues to action. Rather than leave the user to be baffled or frustrated they are urged to take responsibility for data retrieval.
We recommend that you dig deeper if you receive an answer that at first does not make apparent sense.
If they do not have an answer that makes sense it is because the user does not understand as yet the guidance being offered and has not delved deep enough.
Fear of God and PayPal
It is obvious also that, like many companies that peddle salvation, skinniness or strength, the purchase is melded with personal impact. Indeed if you don’t contribute funds you are actually making a destructive choice:
Your donation contributes to the life-affirming values of love, life and freedom supported by Buddhabot research and development.
Q. What are donations used for?
A. The ultimate aim of the Buddhabot is the discovery of a tool that will allow humanity as a race to find what we call ‘omnipathos’. Omni-pathos means “all pathos” or universal appreciation/warmth/sympathy and is a more accurate description for what is often described in the English language as mystical union or love. If a global effort of intention and resources can be focused upon the Buddhabot, the goal may be attained much sooner than expected.
Avail yourself to the Buddhabot and Buddhabot Blog
“If you meet Buddha on the road, kill him.” –Gautama Buddha
“Especially if he asks for money.” –A modern day disciple.
Hmmm yes, appropriate quotes.
for the record: I have no problem with people making money — it is just the bundling of cash with esteem that bothers me.
Requiring a subscription for access is, I find, a useful strategy for conserving my limited development resources (time) in that it helps create a nexus of intelligent or at least inquisitive individuals who tend to supply relevant questions and information. In past experiments with public domain bots I have observed that most humans first try to break or test the bot, then they abuse it and try to “pervert” it and then, if they are still interested they try to engage in sex chat and finally after all this you are left with the most inquisitive, innocent or intelligent of our species who begin to explore more deeply into the possibilities and limitations of the technology.
I really appreciate this blog, it is great to see the inquisitive open ended side of humanity.
FYI, anyone who can’t spare $10 is welcome to a free subscription if they can spare the time to write an email requesting a free subscription. In fact I usually send a free subscription to anyone who even takes the time to bitch about the price $10 :-) Of course this is usually a waste of time because few people care about stuff they get for free.
Bottom line - I advocate subscription fees to conserve development resources for client who care and to thus minimize exposure to abusive clients. Otherwise it won’t be pretty when robots take over the world ;-)
nsn
Hello Nomotheticus!
Thanks for coming by and sharing your experience of bots and offering free subscriptions to WRT. Bot abuse (which becomes botmaster abuse) is a terrible thing and dissuades many botmasters from making their bots available in the public domain. Your descriptions of the process of bot-interaction is so true and I think your tactic of having a subscription fee to provide some obstacle to such users is a good one. [As I stated before, it is the ‘benefits’ to subscription you cite that scratches me the wrong way not havign to subscribe in itself. It should be noted that I, like many others, very rarely pay for things on the Net — for financial reasons as well as conceptual. This is one of the problems for creators on the Net — how will anyone receive compensation or at least assistance to put bread on the table?
But back to the issue of ‘unfriendly users’. I find abuse people deliver to bots and strangers in chat environments (I’ve been subjected to rudeness many a time in virtual worlds) strange, sometimes hurtful and annoying. There are a number of papers on the subject, and antithetical ones (that of human fear of AI) in the area of HAI (human-agent ineraction), and I’ve listed some below. What drives people to act that way is also a much researched area, but more so in the world of online games and Vegemite(!) [see post] Anytime an avatar, or remoteness, or AI comes into the picture some people act in unusual ways. The ultimate receiver of the abuse of bots, though, is the botmaster. Some botmasters respond by not making their public (as I said before); most employ abuse-response scripts that try and turn the conversation around or even block the user after warnings; and some create smart-arse bots that can give as good as they get (see Artificial Iniaes for my favourite bot with cheek!, or the generator: Shakespearean Insulter); and some make money out of the lewdness people desire by creating sexbots and the like (see a recent game example in Virtually Jenna). It is a big problem for many botmasters, especially when they spend so much time creating a character (their creative baby as such). Alot of the abuse is directed not just at the character or bot but mainly to the botmaster in comments about how stupid the bot is or how it doesn’t work. They are abusive attacks that quite often put the most eloquant critic to shame.
De Angeli, A., G.I. Johnson and L. Coventry (2001) ‘The unfriendly user: exploring social reactions to chatterbots’ presented at Proceedings of the International Conference oh Affective Human Factor Design, London, published by Asean Academic Press [pdf]
Norman, D.A. (1997) ‘How might people interact with agents‘ in Software Agents (Ed, Bradshaw, J. M.) AAAI Press/The MIT Press, CA, pp. 49-55.
Zubek, R. and A. Khoo (2002) ‘Making the Human Care: On Building Engaging Bots’ presented at Proceedings of the 2002 AAAI Spring Symposium on Artificial Intelligence and Interactive Entertainment, published by Artificial Intelligence Association [pdf]
Ah, this is such an interesting thread that it leads me to a new post. I agree with Christy, but want to take this conversation in a slightly different direction. See “Perverting the Bot.”
Buddhabot is now being advertised in a TV commercial to thousands of spectators at the Canadian Grand Prix in Montreal. The ad is being screened 688 times during the event on 3 large screens. The text ad begins with a list of questions:
The first question, ‘got bot?’, doesn’t make sense. Perhaps, then, its function is to shock to gather attention? It is interesting that the questions about the nature of the self, the nature of AI and Life is bundled together. These are questions that all botmasters are interested in and explore. I’m beginning to wonder, however, how many other people share this existential relationship with AI. To me, the use of a bot in a story or website, is intentionally a comment on the nature of life. I use bots in my storytelling because I want the reader to think about why I’ve chosen a bot rather than use human actors or textual descriptions, and what a bot is. But many gamers, for instance, do not see a bot in a game on any other level other than a tool or prop. Whenever I play a game I’m more concerned about staying alive or getting information than considering the philosophical ramifications of my interactions with embodied code. I’m mentioning this observation here because I’m curious as to whether the questions on the Buddhabot ad will be shared by the spectators.
The second part of the ad…well, I’ve commented before on how I object ot the bundling of esteem and money…
But I find the fact that a bot is advertised in such a mass manner very exciting. Go Bots! I can’t wait to see bot characters as brands — where a character that you as a botmaster spend years developing becomes a place, a person, that people want to spend time with.
And I can’t help seeing a contemporary ELIZA aesthetic here.
There is also now a Buddhabot Blog.
Info sourced from Alicebot and Buddhbot News.