Austin firm proposes 5 principles to combat ‘unethical uses of AI and digital humans’ – Biometric Update
Wants to prevent harmful, illegal, illicit activities ‘like hate speech and deepfakes’
Austin, Texas headquartered UneeQ, a company that provides a “digital human platform [to] design and launch AI-powered brand ambassadors that can recreate the best aspects of human interaction … to drive an emotional connection … to empower creators to make a positive difference with digital humans,” this past week called on businesses “to adopt five laws for ethical digital human design.”
The company even features a video of “Sophie,” a digital human who “highlights why the five laws are so important.”
In its announcement, the firm – whose digital humans it says can be designed to interact with real humans in “tone, conversation, and expression” — stated the five laws it has posted online “are open for discussion, debate, and suggestions for improving them and speeding adoption. They guide the ethical creation and use of AI-digital humans, including protecting against human bias and prohibiting things like pretending to be human and being used for harmful, illegal, or illicit activities like hate speech and deep fakes.”
UneeQ “wants the industry to adopt a principle-led, ethical approach to building a world with digital humans, which are already becoming part of our everyday lives, helping us work, providing education and support, and even managing our health,” emphasizing that, “(o)ver the next few years, they’ll slip seamlessly into virtually every part of our lives to help us in countless other ways – that is, if digital humans are designed with a set of principles in mind.”
Thus, the company said, it has created “five laws … to bring out the best in our digital population, as well as deter real people from designing AI to mislead, to spread fake news, and to cause other types of harm and distress.”
These five rules for the digital representatives of “the world’s largest brands [will help] scale customer experience and interactions in a more emotional and impactful way with AI-powered digital human brand and service ambassadors,” the company said, adding, “(o)nly 7 percent of great communication is made up of the words we say—the rest is how we say it using tone of voice and body language.”
In other words, artificial AI powered humans are able to “see and listen to users” and comprehend what they are saying, and in turn “use their own tone of voice and body language to create lifelike human conversations” in a manner that grasps how to react appropriately to the speech, tone, expressions, and mannerisms of the human its conversing with – and, presumably, be providing answers to its human’s questions.
“Have your say, contribute to the conversation and help build the best possible future of man and machine cooperation as a community,” the company says on its website where individuals can sign up to engage in discussing the firm’s proposed five laws for governing the creation and use of AI powered artificial humans, noting that standards for morals and ethics for synthetic humans is necessary because “digital humans today play a part in how we live, work and manage our health, as we talk to, confide in, and hold actual conversations with them.”
“Interactions and relationships between man and machine are dramatically changing, and the industry needs to establish a moral compass of ethical business practices that recognize the impact and connection digital humans have with customers,” said UneeQ founder and CEO Danny Tomsett. “We want to provide a starting point on the right recognition and employment of digital humans—and invite feedback, so we can take the necessary, proactive steps to do the right thing as more human, AI-driven interactions move into the mainstream.”
“Five years from now, when there are millions of digital humans accessible to us all, we’ll be glad that we created a strong foundation—a stable nucleus—on which the digital human existence can accumulate,” said QuHarrison Terry, a futurist and co-founder of Inevitable Human, which describes itself as a “community for future thinkers” whose “members receive an ‘almost daily’ newsletter from Terry [in which] he shares his thoughts on the emerging technologies shaping our future.”
“These five laws by UneeQ are a novel and a necessary idea. It’s how we limit any sort of pandemonium from occurring out of the grand opportunity that is digital humanity,” Terry held, clarifying that, “unlike automated chatbots, search windows, and other self-service mechanisms, digital humans—like Sophie—can interact using human-like expressions, warmth, and personality.”
UneeQ said in its statement that “AI-powered computer-generated imagery, vision, and voice, work together with any chatbot technology to determine behavior and speech in realtime. Conversations with digital humans flow naturally, as they tend to in real life. Digital humans can respond using emotion and expression beyond just text, using situational awareness to interact in a way that is appropriate to the context.”
UneeQ’s clients include health insurers, telecommunications, and financial services businesses around the world, including UBS, Vodafone, Southern Cross, and NAB’s UBank.
The company said it recently expanded into the U.S. “with a $10 million angel and venture capital funding round that included Richard Socher, the chief scientist at Salesforce, and was led by Sydney, Australia-based VC firm, Alium Capital.”
“What was once limited to the realm of sci-fi or ‘one day’ is quickly becoming ‘everyday’ as we enter an era in which it’s possible for everyone to have a dedicated digital human customer service representative, bank manager, teacher, or nurse, and the parameters that govern the design of these will become more and more important,” UneeQ said in its announcement.
Since the firm’s call for public engagement through its website about its proposed five canons for AI humans, as of this report only 106 people had signed up.
The five values UneeQ has proposed are:
• Honesty and transparency. Digital humans shall never pretend to be what they are not;
• AI for global good. Digital humans should always be designed to help not harm;
• Right to privacy. Digital humans must respect the privacy of those they interact with;
• Respectful behavior. Digital humans should be designed to uphold respectful behavior; and
• Co-design principles. Co-design must be utilized to avoid bias and promote diversity.