Interacting with chatbots is becoming more pervasive in day to day life. Common applications include customer service portals and E-commerce support, though this technology is being applied across more and more domains, including healthcare.
Alongside commercial applications of AI chatbots, individuals are beginning to make their own avatars or digital clones. The idea here is that an individual can train AI technology on content they have personally created, so the technology can act as a proxy, creating more content or responding to inputs on the individual’s behalf. Evolution of technology on all frontiers makes creating a digital clone increasingly accessible. An example is the ability to create a Personal Voice on iOS 17. By spending around 15 minutes recording randomly selected key phrases into an iPhone, iOS 17 can replicate a user’s voice to allow personalised text to speech communications. This sort of technology can be used by people at risk of speech loss or with a disability affecting speech, but could also be deployed as one facet of a digital clone - by creating personalised audio content from written content without the need for human vocalisation.
Open AI has released GPT builder, that allows users to create custom GPTs that deploy particular knowledge or skills, or work within specific instructions. For example, it is possible to create a custom GPT maths tutor specialising in supporting learners with fractions on the GCSE curriculum. It is also possible to train up GPT builder on an individual’s own content to create a custom GPT that deploys individual-specific context in interactions with users. Other services offer products tailored specifically around the use case of an individual making an AI clone of themselves to scale up their ability to interact and build relationships with more users/ clients or work with others.
Humans will increasingly be interacting with technology instead of humans at work. The impact of human/ avatar workplace relationships on human health and workplace productivity must be studied as these changes take place. For example, an avatar providing HR support could increase capacity and help improve response times for employees, which would confer benefits to the individual and the organisation. On the other hand, AI outputs may be biased according to training data or be unable to manage emotional nuances well, leading to strained relationships on HR issues that may be sensitive and are likely to affect productivity for the organisation. Would an AI avatar trained on content created by a company’s HR director perform better or be better received in that context than a generic HR avatar? Possibly. The value of individual human expertise and individual human personality and relationship building will likely be retained by the digital clone to some extent.
The future of work may also involve avatar/ avatar relationships. This could minimise operational and administrative workstreams and increase productivity. For example, a personal avatar trained on travel preferences with calendar access could liaise with a commercial customer service avatar to book flights for a work trip.
The balance between generic AI technology and technology trained on personal content will be interesting to observe across domains. Our interaction with technology rather than humans will have implications for the future work and health as well as productivity. Another angle on this evolving space is the work and health implication for individuals training AI on their own personal content and using avatars to extend their accessibility. Time will tell how much human supervision of digital clones is necessary as well as the impact of opening up channels for engagement via avatars on the humans behind them.
ncG1vNJzZmivn6e4orrDoZyapKSde7TBwayrmpubY7CwuY6pZpqukamus7%2BMr6pmoKWirq%2B%2F