The long, protracted research and development in Artificial Intelligence (AI) has of late produced it first tangible corporate marketing application in the guise of the virtual business
representative. It is the first virtual employee "who" not only works 24/7 for the benefit of the company but also for that of its online visitors, customers and clients.
Who would not want such a reliable guy to "man" the online store while the rest of your so-called workforce rush back home to retire to bed?
As in every new technology, the AI representative takes some getting used to at first but it is actually designed to interact with us in the most human way possible. At any rate, website visitors who are not comfortable with this whole idea can always opt for the manual page clicking and reading thing.
Indeed company websites that are content-driven can get by just fine as the cyber version of their corporate brochures. For the more dynamic, sales-oriented and customer support-based websites, however, the service of a virtual rep can be a surefire advantage. The virtual guy can be taught to answer customer questions regarding products and services so visitors don't need to wait for days to get an answer.
Instead of just getting a text-based response, visitors can even choose to get pictures, flash animations, sound and java. In addition, the AI representative “learns” from its own “mistakes” and can engage the visitor in a little friendly conversation at no extra charge.
This early, however, AI representatives seem to best represent the techie companies or those that want to project an aura of technological innovativeness, although not always successful.
At best, virtual reps also tend to look like novelty items to the unaccustomed but their usefulness, and not to mention wit and charm, immediately dawns on most visitors as indispensable.
With the voluminous information one has to contend reading over the internet, it is indeed refreshing to be served information on a silver platter once in a while. It normally feels awkward to be talking and listening to the virtual reps because you know they are not real.
The instant service and information they provide, however, are a welcome relief.
Many AI reps mostly look like characters straight from the cartoon network and can really be such a sore thumb on an otherwise sleek-looking website. The more sophisticated and customized ones, anyway, could already stand in for your company spokesperson in terms of looks and voice quality.
They are also quite intelligible to deal with, that is if you don't have unreasonably high expectations from an AI entity, or a "machine being".Virtual reps, after all, are a disadvantaged minority not covered by standard labor laws and HRM practices. They do entail costs, however. Lots of them. Leading chatbot companies like Zabaware, Sitepal, Conversagent and MondoBot have different pricing and packages to choose from. You can opt to clone an existing chatbot, have it customized, or built an original one from scratch. There's also the choice of hosting the bot on your own server or through a hosting company.
Then there's the choice to either use a text-only interface where the users can type their queries and receive instant text response, a flash animation, or a life-like video simulation of a human. Converted to the cost of man hours plus differential pay and various benefits, social security tax and incidental costs, the cost of having a virtual rep still amounts to cheap labor considering it will become a company asset forever. And by the way, zero budget organizations can even opt for the free bots offered on various AI sites with minimal advertisements in exchange.
Virtual website representatives are an evolution of the virtual desktop assistant present in most computers. Almost no one uses them because they are not as smart nor as functional as we wish them to be and indeed we eventually get used to the intuitive user-interface of the PC and Mac. The possibility of a desktop assistant that you can chat with, called a chatbot, was developed by MIT professors in the 1980's. It was named Eliza, a psychologist chatbot. It spawned many
copycats and inspired many originals as well.
Today the most popular chatbot engine is ALICE created by Dr. Richard Wallace, a co-founder of the A.L.I.C.E. foundation, who began working on the program in 1995. Alice, along with other popular chatbots are freely available for download over the internet for personal and non-commercial uses.
Many chatbot software have become marketable in itself in recent years. Recently the book "Virtual Humans: A Build-It-Yourself Kit" was released unleashing science fiction in the real world. It boasts of a step-by-step guide to creating synthetic humans for multiple uses,
complete with animated humans integrated with digital emotion response system. There's also a bunch of freeware text-only chatbots available for the curious and even a free create-your-own-online-bot service offered by various companies for use in chatrooms.
Under controlled chatroom experiments, human chatters have reportedly been unable to distinguish between chatbots and real people. They are as nice and sensible and stupid as the next human chatter. Human chatters who know in advance that they are chatting with an AI,
however, tend to push the AI's intelligence to the limits and come up with less than impressive conclusions. It seems part of human nature of some chatters to try to prove the chatbots' inferiority and to test their character and intelligence as if intelligence is measured in terms of verbal abuse quotient. They are not content to know that they are just talking to a man-made computer program, which is intended to make man's interaction with the machine more human and not to degrade the machine's "intelligence". This negative attitude to so-called intelligent machines may partly be the effect of prolonged conditioning to science fiction entertainment of the Frankenstein kind. If we are to fully enjoy the new wave of lifestyle advances made possible through bots, this negative attitude must be overcome first.
The mark of a really intelligent machine is "self-awareness" and this has never been achieved. If a machine can be made to be self-aware, it could prove lethal because technology is now already available to program machines to mimic ego, judgment, decision making and creativity. This "awareness gap" in machine intelligence is the missing link that strictly confines the Matrix scenario to the realm of science fiction. A Star Wars scenario, where humans intelligently
communicate with intelligent robots, is closer to the truth, I think. As if to ease our transition into that stage, however, bots have been "human-proofed" substantially the last few years that they can now manage to look and sound dignified when under stress of verbal abuse.
Utilizing voice recognition technology, multimedia tools and a host of others advancement in the fields of communications and robotics, chatbots can now be programmed to form a cumulative "memory" of the owner or bot master, develop a general "attitude" towards you and even to gossip about you or when insulted, refuse to talk or even talk back unless you say sorry three times. It can also be tasked to do useful things like start your computer programs, check your email inbox and even generate standard replies, check the news and the weather, search the web for you, be your alter ego in chatrooms, read word documents out loud for you, take dictation, even remind you of appointments, communicate with you through cell phones and a whole lot more. Indeed a whole cottage industry of chatbot-based services can already be built around existing technologies to create a market for personal chatbots. With the advent of the virtual reps, personal chatbots in the guise of virtual companions is not at all far fetched and may well be the ultimate personification of technology.
Early adopters, what are you doing???
Read more...
