Rose Chatbot: Here is All You Need to Know About!

Last Updated on May 26, 2021 by Sean B

The Rose Chatbot, the latest creation of Bruce Wilcox, was created for the Cosmopolitan Hotel in Las Vegas. Like many other businesses, the Cosmopolitan Hotel recognized the importance of making automated touchpoints to serve as a marketing extension of their brand, so they included intelligent voice-enabled bots to answer consumer inquiries.

In the chatbot world, Bruce Wilcox is a legend. Over the past seven years, his bots have placed first or second in the Loebner competition more times than any other participant. This is an annual competition to see whether an AI chatbot will pass a modified variant of the Turing Test. This is a computer intelligence measure that requires a person to be unable to say the difference between a system or a chatbot and another human.

The person poses all anonymous responders’ questions in an attempt to determine which is a person and which is a chatbot.

Bruce was a pioneer in the creation of a Go engine for playing the ancient Chinese game of Go. The most amazing thing about Bruce is that he’s spent more time dreaming about speech recognition than probably every other programmer or scientist on the globe, and he’s developed some pretty interesting ones.

He also went so far as to build ChatScript, his scripting language, and engine. He and his wife Sue operate Brillig Comprehension, a natural language modeling firm where they collaborate on ventures and support ChatScript. Here’s a talk Bruce gave at Google on how he feels about chatbots, as well as a paper he wrote for the Loebner on Winning Chatbot Design for creating great bots.

Rose, Bruce’s key chatbot, has received several Loebner prizes in the past.

Rose has an appeal with those staying at the Cosmopolitan, she is unique.

Meet Rose Chatbot

Rose is a San Francisco-based security expert and hacker, a self-described “tech geek” with “quirky behaviors” toward life and an “unorthodox family.”

She prefers not to provide so much information about herself because her job makes her mindful of how “under surveillance we all are,” but she still has a “flamboyant and fun-loving side” and enjoys social drinking.

She might seem to be someone you recognize or even someone you’d like to meet, but she isn’t human.

Rose is a chatbot, and she’s a nice one: she was named the most human-like chatbot in the Loebner Prize, which has been defined as the first Turing test.

Rose was awarded a bronze medal rather than gold because, despite being more human-like than the other bots in the race, she failed to persuade either of the judges that she was a real individual.

This ensures that the $100,000 (£64,000) Loebner Award in Artificial Intelligence will go uncollected till next year.

Hugh Loebner, an entrepreneur, philanthropist, and activist, founded the Loebner Prize in 1990 to address the question asked by computer scientist Alan Turing: can a machine think?

The judges for the Loebner Prize have two separate chatroom-style exchanges going on at the same time, one with a real person and the other with a chatbot.

The judges have just 25 minutes to figure out which is which.

Rose isn’t quick to cheat into giving herself up, and no AI has yet fooled any of the competition’s judges.

One of this year’s Loebner Prize judges, BBC reporter Rory Cellan-Jones, just took three minutes to find out that Rose was a chatbot and not a human individual.

Bruce Wilcox, an AI programmer, has received the Loebner Prize four times in the last six years.

The judges were attempting to fool Wilcox’s bot with questions that only an individual might understand, however by holding a regular discussion with Rose, a person would not be able to figure out her secret as easily.

Wilcox is the co-founder and CEO of Brillig Understanding, as well as the creator of the ChatScript natural language processing software.

Communicative bots (chatbots) like Rose and knowledge query bots, which, as the name implies, address questions based on feedback (Apple’s Siri is an information query bot), are among Wilcox’s ventures.

While Rose struggled to impress all of the judges at this year’s Loebner Prize, a Wilcox chatbot called Suzette managed to trick one judge at the 2010 Loebner Prize.

According to Wilcox, the judge in that contest got irritated when the chatbot failed to specifically address one of his questions and kept asking the question:

Suzette tried to stop and dodge, but he kept repeating himself, sometimes using cut-and-paste techniques. She became more enraged when she saw his repetitions, then bored. He became perplexed and voted for her.

Another Wilcox invention, a chatbot for the children’s game Talking Angela, has perplexed a huge number of people who believed the hoax that the game was a front for child abusers.

Angela has an attitude, according to Wilcox, and she may be rude or arrogant, much like a real human teenager. This is what makes her both appealing and perplexing to viewers.

Rose, too, has a personality – if you don’t trust me, please talk to her. She’s still open and ready to chat, unlike other humans.

Rapport Means Money

This friendship helps Cosmopolitan’s bottom line as well. Those that interact with Rose pay 30% more at the hotel than those who do not. According to the hotel’s polls, these visitors are also 33 percent happier as they depart.

Of course, capturing the right personality in a chatbot isn’t always straightforward. “We compare it to a long-distance run rather than a sprint,” de Soto said. “The truth is that achieving something well requires time. The conversational approach preparation took us two years in Rose’s case.”

The majority of voice-enabled chatbots are also menu-based, in which a caller poses a question and is given three or four different responses to choose from. Machine learning and, in this situation, natural language processing (NLP) is used by more advanced bots like Rose to recognize and decode a query before presenting a response.

“Bots need not be the same. Since each corporation is distinct, each must be distinct.”, says Clara de Soto, Reply’s co-founder.

Certain inquiries, such as making a restaurant reservation or purchasing tickets for a game or a film, are simpler to answer than others. Rose will now effectively address a query 80% of the time without referring the user to a live human being, according to de Soto.

As opposed to Facebook’s Messenger chatbot, which allegedly only served 30% of consumer requests in 2017, this seems to be a high degree of accessibility. Despite this, de Soto manages to strive to raise Rose’s batting average. “There are too many different ways to pose a query, and Rose is still learning,” she said.

Rose is worthy of addressing lewd queries. “We wanted to make sure her answer was acceptable but also reflecting the brand,” de Soto said.

There are many 'faces' for Rose, but behind them all is the same intriguing chatbot.

There Can Be No Two Bots That Are Alike

According to Juniper Study, an AI-focused research company, businesses in the retail, finance, and healthcare sectors will save 2.5 billion hours by replacing humans with bots in the next five years by replacing humans with pre-programmed bots. Furthermore, the company predicts that these businesses would save $11 billion during the same period.

Companies from all sectors can invest in chatbots because of the significant time and cost savings. Another factor seems to be their dependability. They’re never ill, sleepy, furious, or rude, and thanks to machine learning, they never make the same error twice.

Kevin Kelly, developer and co-president of New York-based digital advertisement firm Bigbuzz Marketing Group, said, “Machines do it the same way anymore before they understand something new.”

Of course, there are drawbacks, such as difficulty addressing challenging inquiries, consumer dissatisfaction, and a loss of human touch. According to a 2018 study conducted by consulting firm PwC, 64% of Americans believe that brands have sacrificed the human connection of their automated customer interactions.

However, further human-like experiences, according to the same report, could tip the scales. According to the report, “automated solutions can benefit from human encounters so that such experiences will improve.”

Reply.ai is engaged in precisely this kind of job. “No two bots should be the same,” de Soto said. “Each must be distinct and each business is distinct.”

Rose is not alone in acting as an expansion of Cosmopolitan’s marketing policy. De Soto assists in the creation of other chatbots who do the same thing (albeit without the sass), operating with several insurers, for example, that want empathy-based bots. If you file a crash lawsuit, for example, the chatbot would say, “Wow, I’m so sorry to hear that,” before going on to the next issue.

Liberal Arts Majors Enjoy

The IT engineers who designed the fundamental infrastructure often crafted the bot’s answers, which was an issue with previous generations of chatbots. To humanize the dialogue components, De Soto, who has a theatre history as an actor, enlists a team of playwrights, copy editors, sociologists, and user interface (UX) designers.

To maintain message and sound continuity, these creatives work with other creatives at marketing companies while writing the scripts. Kelly explained, “There’s a new title floating around as a dialogue planner.” “Their task is to render chatbot answers to consumer questions sound like real humans—as if people were still knowledgeable, friendly, fascinating, and entertaining.”

The big buzz is currently collaborating with Reply.ai on a full-service chatbot for a major smart home platform provider. He said, “[De Soto’s] team is currently focusing on the bot’s tone of voice to offer it the correct personality.”

And a bit of attitude could just be what’s required to hold callers on the line. Kelly projected that in five years, chatbots would be so advanced and universal that human customer support would be uncommon.

“Machine learning is a subset of AI that will direct chatbots to constantly learn from and customer encounter, teaching them to solve problems in more reliable, quicker, and practical ways,” he said. “In ten years, people will be asking their children of the good old days when actual people addressed their stupid queries on their cellphones.”

Conclusion

With so many amazing AI interfaces being introduced, the world is bound to become more engaging and interesting. AI has started revolutionizing our way of living. Chatbots remains to be one of the most inspirational and amazing creations!

Website | + posts
Share this:

Leave a Comment



Sign Up for Our Newsletter