Saturday, February 23, 2019

In findings published on Wednesday in PLoS One, scientists from the University of Bristol and UC Davis say zebras may have gotten their stripes to confuse ectoparasites, such as flies.

The researchers observed tabanid horse flies around zebras and domestic horses in captive settings and they found the flies had a harder time landing on zebras than on the monochrome coats of the horses.

Results showed the horse flies hovered over both types of animals at roughly the same rate but, over zebras, they did not always slow down fast enough to land successfully. They then equipped horses with both zebra-striped and unstriped cloths and again observed that the flies had trouble landing on the striped surfaces.

“This reduced ability to land on the zebra’s coat may be due to stripes disrupting the visual system of the horse flies during their final moments of approach. Stripes may dazzle flies in some way once they are close enough to see them with their low-resolution eyes[,]” said study co-author and Royal Society University Research Fellow Martin How.

Scientists have long wondered about the evolutionary impetus behind zebras’ coat patterns. Other hypotheses include the idea that stripes may confuse predators, facilitate social interactions within the herd, and regulate body temperature.

In Africa, where zebras evolved, flies can carry dangerous blood-borne pathogens.

Wednesday, November 30, 2005File:Turing1.jpg

More than 50 programmers, scientists, students, hobbyists and fans of the A.L.I.C.E. chat robot gathered in Guildford, U.K. on Friday to celebrate the tenth birthday of the award winning A.I. On hand was the founder the Loebner Prize, an annual Turing Test, designed to pick out the world’s most human computer according to an experiment laid out by the famous British mathematician Alan Turing more then 50 years ago. Along with A.L.I.C.E.’s chief programmer Dr. Richard S. Wallace, two other Loebner prize winners, Robby Garner and this year’s winner, Rollo Carpenter, also gave presentations, as did other finalists.

The University of Surrey venue was chosen, according to Dr. Wallace, not only because it was outside the U.S. (A.L.I.C.E.’s birthday fell on the Thanksgiving Day weekend holiday there, so he expected few people would attend a conference in America), but also because of its recently erected statue of Alan Turing, who posed the famous A. I. experiment which inspired much of the work on bots like A.L.I.C.E. University of Surrey Digital World Research Centre organizers Lynn and David Hamill were pleased to host the event because it encourages multi-disciplinary interaction, and because of the Centre’s interest in interaction between humans and computers.File:ALICE Birthday Cake.jpg

Dr. Wallace gave a keynote address outlining the history of A.L.I.C.E. and AIML. Many people commented on the fact the he seemed to have moved around a lot in the last ten years, having lived in New York, Pennsylvania, San Francisco, Maine, Amsterdam and Philadelphia, while working on the Alicebot project. The A.L.I.C.E. and AIML software is popular among chat robot enthusiats primarily because of its distribution under the GNU free software license. One of Dr. Wallace’s PowerPoint slides asked the question, “How do you make money from free software?” His answer: memberships, subscriptions, books, directories, syndicated ads, consulting, teaching, and something called the Superbot.

Rollo Carpenter gave a fascinating presentation on his learning bot Jabberwacky, reading from several sample conversations wherein the bot seemed amazingly humanlike. Unlike the free A.L.I.C.E. software, Carpenter uses a proprietary learning approach so that the bot actually mimics the personality of each individual chatter. The more people who chat with Jabberwacky, the better it becomes at this kind of mimicry.

In another interesting presentation, Dr. Hamill related present-day research on chat robots to earlier work on dialog analysis in telephone conversations. Phone calls have many similarities to the one-on-one chats that bots encounter on the web and in IM. Dr. Hamill also related our social expectations of bots to social class structure and how servants were expected to behave in Victorian England. He cited the famous Microsoft paperclip as the most egregius example of a bot that violated all the rules of a good servant’s behavior.

Bots have advanced a long way since philanthropist Hugh Loebner launched his controversial contest 15 years ago. His Turing Test contest, which offers an award of $100,000 for the first program to pass an “audio-visual” version of the game, also awards a bronze medal and $2000 every year for the “most human computer” according to a panel of judges. Huma Shah of the University of Westminster presented examples of bots used by large corporations to help sell furniture, provide the latest information about automotive products, and help customers open bank accounts. Several companies in the U.S. and Europe offer customized bot personalities for corporate web sites.

Even though Turing’s Test remains controversial, this group of enthusiastic developers seems determined to carry on the tradition and try to develop more and more human like chat bots.Hugh Loebner is dedicated to carry on his contest for the rest of his life, in spite of his critics. He hopes that a large enough constituency of winners will exist to keep the competition going well beyond his own lifetime. Dr. Wallace says, “Nobody has gotten rich from chat robots yet, but that doesn’t stop people from trying. There is such a thing as ‘bot fever’. For some people who meet a bot for the first time, it can pass the Turing Test for them, and they get very excited.”