Concept 23 – Human-Computer Interfaces
The Internet lessens the recognition of difference between human and computers because, at a distance, it often feels similar to communicate and act on the internet regardless of whether one is speaking with a human or a machine.
More and more our interactions on the internet are with computers. These days when purchasing an airline ticket, we would no longer ever consider walking to a travel agency to discuss flight options with a staff member and pay in cash. We are now quite able to research options from the comfort of home, a search engine collates and provides prices for possible itineraries after we have entered our criteria, and often provides cheaper alternatives as well.
The way we interact with businesses online is also evolving though. Often now we have the option of contacting a business by email – which will prompt an automated reply, or by an online chat.
BT (British Telecom) has developed Anna, a bot (or robot) to assist with enquiries. She looks nice and helpful, blinks and moves her head around a little to appear more human. For each question you ask she will provide a few options providing links to the page that will (hopefully) hold the information you are seeking. When you select one of these links you are navigated to information pages, and Emma disappears. I think this is a weakness in the navigation, I felt a bit deserted when she left without saying goodbye! Here’s a link to BT’s Ask Emma: http://bt.custhelp.com/cgi-bin/bt.cfg/php/enduser/cci/bt_homepage.php?p_sid=m83sazaj
To enable a bot to effectively answer a query, a large number of predictions need to made about the information that will be sought. Responses need to be modelled on a ranges of keywords, and also need to cover options including spelling errors.
You can now develop your own bot at home, with your very own personalised responses. Because recording responses to all possible queries is an exceptionally time intensive process, you can take a bot with either pre-programmed responses (default brain) or you can build your bot entirely from the start (no brain).
I had some discussions with a few of the more popular bots linked to this site, and I was a little disappointed. Some of the responses were very strange and not at all linked to the question I asked, and some of them are programmed to be just plain cheeky and sarcastic! Overall, it felt quite clear that I was talking with a computer, it just wasn’t right.
I had a better experience when trying to install a new internet connection recently. There were some problems getting started up, and the help files on the cd were presented in a very relaxed, simple question and answer style. For each response I provided the cd would invoke a process to check a certain part of the hardware of software that it felt could be root of my connection problems. There was very little looping in the questions, it was very polite at all times, and I felt quite secure that there was progress happening.
As business continue to reduce resourcing of labour intensive roles, more and more we will be communicating with computers in our every day lives.
Supporting Site 1:
Create AIM Bots, stay on AIM 24/7 and web chat solutions - http://www.ai-buddy.com/
This site is where you can get your very own Bot. Here they are being promoted as excellent receptionists for your email or IM services while you are asleep, as they can provide pre-programmed information to your senders. It also contains a information on how to set variables to questions that may be asked – also reminding to allow for common spelling mistakes.
I have found that if you ask really silly questions some bots end the discussion. I haven’t been able to resurrect a discussion with a bot I have annoyed so far. They can be quite moody.
Supporting Site 1:
Wikipedia: Human Computer Chess Matches - http://en.wikipedia.org/wiki/Human-computer_chess_matches m/
Proof that computers can be programmed to think strategically, Wikipedia gives a nice description of the evolution of chess programmes, and how they have now become more superior than humans at the game. Grandmaster chess player Kasparov thought he saw “deep intelligence and creativity“ in Deep Blue, a computer designed for chess games, thought there must have been human involvement.
It appears that after many wins over the last few years, computers are now seen as the superior players. The page quotes a McGill University computer science professor, Monty Newborn, when referring to the lessening interest in the human-computer chess games due to the computer’s superiority, "the science is done"
Subscribe to:
Post Comments (Atom)
No comments:
Post a Comment