The Future of AI Assistants: A Potentially Transformative Shift
In the near future, AI assistants embedded within wearable devices will seamlessly integrate into our daily lives, providing real-time guidance as we navigate various situations—from shopping in busy stores to visiting the pediatrician. This technology will actively participate in all aspects of our existence, influencing our social interactions with friends, family, coworkers, and even strangers.
However, the concept of "influence" raises concerns about privacy and autonomy. While many may feel uneasy about such pervasive assistance, society is likely to embrace these technologies, benefiting from their ability to offer timely and informed advice that feels indispensable.
Understanding Contextual AI Assistants
When we refer to "AI assistants," we often think of basic tools like Siri or Alexa, which respond to simple voice commands. Next-generation AI assistants will be far more sophisticated due to their context awareness—capable of perceiving and responding to the sights and sounds around us through advanced cameras and microphones integrated into wearable technology.
Set to make a significant impact by 2024, these context-aware AI assistants promise to deliver information tailored to our current activities. They will feel almost like a superpower—offering insights about products in storefronts, identifying plants on hikes, or suggesting recipes based on the ingredients in our kitchen. However, there is a risk of these assistants becoming overly persuasive, potentially manipulating users while guiding them through everyday tasks, particularly when utilized by corporations for targeted advertising.
The Rise of Multi-Modal Large Language Models (LLMs)
To mitigate the manipulation risk associated with AI, policymakers must focus on this critical issue, especially as the technology for context-aware assistants has only recently emerged. Multi-modal LLMs, capable of processing text, images, audio, and video inputs, represent a significant technological leap, enabling AI systems to interpret and react to our environments in real-time.
For instance, OpenAI’s ChatGPT-4, released in March 2023, is among the first mainstream multi-modal models, followed closely by Google’s Gemini LLM. Notably, Meta’s AnyMAL model takes this a step further by incorporating motion cues, allowing for an AI assistant that considers both what you see and how you move.
Wearable Technology Leading the Charge
With advances in AI technology available for consumer use, companies are rapidly developing devices equipped with cameras, microphones, and motion sensors to provide context-aware assistance. Wearable glasses, for example, present an optimal platform for this technology, ensuring that cameras capture a user's gaze while stereo microphones provide spatial awareness of surrounding sounds.
Meta currently stands out with its updated Ray-Ban smart glasses, designed to support advanced AI capabilities. On December 12, they began rolling out AI features that offer practical assistance, such as suggesting clothing combinations based on user preferences.
Humane, another key player, has developed a wearable pin equipped with similar functionalities, set to ship in early 2024. While both approaches are promising, I believe glasses provide more effective integration by aligning with a user’s line of sight and enhancing mixed-reality experiences.
Regardless of whether the technology is embedded in glasses, earbuds, or other devices, context-aware AI assistants will soon become commonplace. They promise powerful features, including real-time translation, historical context, and personalized assistance during social interactions, enhancing our communication skills and overall social awareness.
Navigating the Risks of AI Influence
As a lifelong researcher in AI and mixed reality, I have long been concerned about the potential for AI to manipulate human interactions. In my short story, Carbon Dating, I illustrated this danger through a narrative where an elderly couple’s interactions are solely guided by AI advice, highlighting the loss of authentic connection. This cautionary tale was later adapted into the film Metaverse 2030 by the UK’s Office of Data Protection Authority (ODPA).
The most pressing risks involve corporate and governmental entities imposing their agendas through conversational influence, enabling customized content targeting individuals. To raise awareness of these dangers, the Responsible Metaverse Alliance has released Privacy Lost, a short film addressing these manipulative aspects of AI.
The Choice Is Ours—For Now
Many individuals naturally find the concept of AI whispering guidance unsettling. However, as AI coaching becomes integrated into societal norms, those who resist may find themselves at a disadvantage, as essential information could be automatically provided to others during conversations. This shift could create a new social dynamic where querying personal information becomes obsolete, as it will be accessible via AI.
I refer to this phenomenon as "augmented mentality," an inevitable trend where AI tools enhance our cognitive and social capabilities. Recent advancements, such as context-aware LLMs, have removed prior technical barriers, signaling an impending shift in how we interact with AI.
With major tech companies vying to dominate this emerging market, a digital divide may arise between those who can afford AI-enhancing tools and those who cannot. The implications of AI-driven conversational influence could lead to privacy infringements and a loss of personal agency.
Need for Strong Regulation
We are on the brink of a reality where corporations could influence our thoughts and actions through AI assistants. This manipulation problem is alarming and requires rigorous regulation to safeguard individual autonomy, particularly concerning context-aware AI.
Unfortunately, recent legislative efforts such as the Executive Order on AI and the EU’s AI ACT have inadequately addressed these concerns. As we approach 2024, there is an urgent need for policymakers to prioritize the regulation of AI-powered conversational influence to ensure that users benefit from AI guidance without descending into a perilous future.
The time for action is now.
Louis Rosenberg is a pioneer in AI and augmented reality, founder of Immersion Corporation and Unanimous AI. His latest book, Our Next Reality, is available for preorder from Hachette.
Ranking
