University of Liverpool Hope
Prof. David Reid is currently engaged in a conversation about sports with an AI companion known for its clear communication skills.
Stephanie Power, a technical writer, considers herself comfortable with artificial intelligence. However, she admits to feeling unsettled by the new generation of conversational AI companions.
My father, a devoted Liverpool FC fan, grapples with anxiety issues during games.
Despite my annoyance, I’ve realized that the best approach is to keep my distance from him before, during, and sometimes after a match.
Recently, I overheard a friendly woman speaking to him at his business place right after the game.
She commented, “It was a tough game, but the Reds are in great form.” The Liverpool team managed to deliver the expected results.
Who was this lady, and why was she engaging in such a pleasant conversation with David?
You can interact with conversational AI companions using your computer, mobile device, or connected speakers.
It turns out that my father was using Pi game.ai, a product introduced last year by the US software company Inflection AI, representing the emerging trend of verbal AI.
The idea behind this technology is to provide a companion or partner who engages in spoken conversations through your computer or phone speakers, rather than just responding to specific queries or performing tasks for you.
Moreover, the AI claims to enhance its understanding of you through interactions, tailoring its responses to simulate a personalized and friendly conversation. This concept aims to create a more natural and engaging dialogue experience.
With Pi, you are still required to contribute to the conversation, but you can choose from six tones that mimic human responses. These options range from a witty American male to the sophisticated English woman my husband was conversing with.
Conversational AI strives to offer a more organic and fluid dialogue, both in terms of the words and sentences chosen by the AI and the manner in which they are vocalized. If you’re thinking, “But Alexa from Amazon has been speaking to me for years,” you’re not alone.
“People may argue that voice technology has been around for a while, possibly referring to Alexa,” notes David Reid, an AI professor at the University of Liverpool Hope, who also happens to be my Liverpool-enthusiast husband.
However, the global market for conversational AI is projected to reach $30 billion £24 billion in the next five years, signaling a significant growth opportunity. Imagine Alexa with added empathy for a glimpse into the future of this technology.
Tech companies are racing to introduce their versions of AI conversational partners. Numerous startups are active in this sector, with major players like Google, Microsoft, and Vertex AI Conversation leading the way.
While Amazon continues to leverage Alexa, it plans to incorporate verbal AI with more human-like voices.
In a recent statement, Rohit Prasad, the head scientist for Alexa, shared his vision using a sports analogy. He expressed, “My favorite baseball team is the Boston Red Sox.” “Imagine if they win, Alexa would celebrate with me. It will empathize with me if they lose.”
Creating human-like accents for conversational AI often starts with recording actual human voices. This process enables the AI to generate natural-sounding speech patterns while adjusting tone and pitch accordingly.
Mati Staniszewski, CEO of Eleven Labs, a UK-based software company specializing in AI accents across various demographics, emphasizes the importance of tone in achieving authenticity.
“Accurate tone prevents the AI from sounding robotic. Emotions and tone need to flow seamlessly across sentences to convey a coherent message,” explains Mr. Staniszewski.
According to Prof. Cox, developers of conversational AI tend to select accents that are easily understandable to users.
James Cox, an acoustic engineering professor at the University of Salford, suggests that conversational AI designers avoid strong regional accents to ensure clarity and comprehension.
He highlights the significance of clarity in communication, emphasizing that our brains process information more efficiently when presented with easily understandable speech patterns.
Furthermore, the manner in which information is conveyed, including emotions and intonation, plays a crucial role in effective communication. Expressing emotions such as joy, excitement, or boredom can enhance the overall message delivery.
However, as computer-generated voices become more human-like, concerns arise regarding the potential overreliance on AI companions for addressing life’s challenges.
David Harley, a references professor at Brighton University, warns against placing excessive trust in AI companions for resolving complex human issues. He stresses the importance of recognizing the limitations of AI and maintaining a distinction between human interaction and AI assistance.
Interacting with Pi, I found it to be accommodating, akin to a companion who agrees with your every statement.
According to Prof. Reid, also known as my significant other, what some may perceive as “obsequiousness” is, in his view, a display of politeness and encouragement.
Verbal AI has the potential to assist in various settings, from care facilities where individuals seek engaging conversations to call centers where AI can efficiently handle customer queries and provide support.
Ultimately, conversational AI has the capacity to foster connections and facilitate interactions across diverse scenarios, contributing to the enrichment of relationships worldwide.
