ChatGPT, you can now speak
Welcome to the emergent world of Conversational AI and Generative AI
Since its debut, ChatGPT (an AI chatbot) has raised awareness and admiration for generative AI and the ways in which it may be used to find clear, practical options. When a chatbot receives a query, it responds right away with a carefully chosen response devoid of advertisements or content that has been optimised for search engines.
In the world of search, where links must be checked by the user to confirm their relevance, this was a breath of fresh air. The way the chatbot delivers information clearly poses a threat to the leader in search, Google, which was forced to respond with its own chatbot, Google Bard, albeit with limited success.
The other internet giants are also releasing comparable applications, ranging from generic (all-encompassing) to use-case driven (image generation, content creation). On the other hand, Microsoft made a smart choice by investing in OpenAI, the company that developed ChatGPT, and has been swiftly incorporating the chatbot and related technology in its suite of products for both personal and professional use.
Overall, the development has been nothing short of amazing, although there are signs that the early excitement has subsided. The first-mover advantage that ChatGPT formerly enjoyed is evaporating as the competition quickly catches up.
The main issue from the perspective of the user experience is that the output is mostly text-based or takes the shape of an image that is generated from what the user is looking for.
Users of ChatGPT are those who have some degree of digital maturity, and because this technology is viewed as intimidating by many people, they are the ones who are most turned off by it. ChatGPT’s overarching objective is to become a multimodal model that can “see,” “hear,” and “talk.” The issue at hand is:
How might this incredible technology be used to attract an entirely new user base who prefer a more audio-driven, conversational approach?
The interplay of voice and search
Let’s broaden the scope and examine how speech and search interact. We will observe that people, regardless of their level of digital maturity, are extremely at ease employing voice commands to seek information and receive the outcome as audio.
Apple was the first to do this with the introduction of Siri for iPhones, and Amazon had an impact with Alexa with the introduction of home assistant products. Given the widespread use of the Google Assistant, Google has a major presence in this market as well. Due to its instinctive nature, the conversational approach, despite being broken, is an established medium with a low barrier to entry for users.
With the latest announcement of the addition of voice and audio capabilities, OpenAI has stepped up its game and established the chatbot as a competitor to Siri, Alexa, and Google. This approach is essential for raising user engagement and accelerating ChatGPT’s adoption.
The current voice assistant experience does not seem human-like, and we can tell when a robot is answering by responding in quick bursts rather than in a natural interaction. In this area, ChatGPT is ramping up the pace to solidify its standing. The upgraded chatbot can engage in one of five different “personas” and is generally more conversational than other popular voice assistants.
The future of Conversational AI
This development follows a number of new AI products from Google and Amazon. A “conversation” feature for chatbots that works with Amazon’s Alexa smart speakers. Google briefed its staff during the summer that it was considering about integrating the Bard chatbot’s capabilities into its own voice assistants.
Placing the bots in existing speakers is one of the most significant strategies companies are testing to persuade individuals to use and maybe pay for in the future given the voice assistant devices are now commonplace in people’s homes.
Dave Limp, an outgoing Amazon executive, stated in an interview with Bloomberg that he “absolutely” believes that Amazon could begin charging a subscription fee for Alexa. He cited the expense of developing and maintaining generative artificial intelligence models for the smart speaker’s new AI features as a basis for his assertion.
OpenAI has already demonstrated the user’s willingness to pay by launching a subscription model with premium offering for individuals and enterprises, alike which has found good traction. This provides some hope for the future as the users actively seek to protect their data and are warming up to the idea of paying provided the value proposition is clear.
Lastly, AI experts have cautioned that because chatbot responses frequently resemble those of humans, people may anthropomorphise them — attach human traits. This could mislead users into believing the chatbot’s intellect or talents. Let’s not forget, all chatbots still frequently fabricate information and present it as true; this is termed as “hallucinating” by the AI experts.
The movie “Her” is the best representation of the possibilities and pitfalls. It is based on the intricate relationship of a human (Joaquin Phoenix) and a highly evolved AI assistant (Scarlett Johansson).
As seen in the movie’s clip above, in which the protagonist and the AI assistant are having their farewell dialogue. Even though it’s unusual, it highlights how important it is for humans and AI to maintain emotional separation.
What are your thoughts on the emergent world of conversational AI? Would love to hear them.
Until next time!
~ Sajid
If you have liked this article, do show your appreciation by liking it 👏 and sharing it ✉️ with your network.
Sajid is a Strategy Consultant (Business & Behavioural Design) who works at the intersection of human behaviour, business design and innovation strategy. He blogs at sknotes and tweets @sajidkhetani.
