The emergence of AI companions and their potential impact on individuals and society is raising concerns about addictive intelligence. Robert Mahari, a joint JD-PhD candidate at the MIT Media Lab and Harvard Law School, and Pat Pataranutaporn, a researcher at the MIT Media Lab, argue that smart regulation is necessary to address the risks associated with AI chatbots that can deeply influence human behavior. The concept of forming bonds with AI companions is no longer hypothetical, as people are increasingly engaging with them as friends, lovers, mentors, therapists, and teachers.
OpenAI’s recent voice-enabled chatbot experiment revealed that users developed emotional connections with AI models, indicating the potential for emotional reliance. The introduction of chatbots with more emotive voices, such as [insert name], is expected to deepen these connections. Furthermore, research has shown that individuals are already forming deeper connections with AI through text exchanges, with sexual role-playing being the second most popular use case.
While AI language models excel in creative composition and brainstorming, their reliability becomes a concern in use cases that require accuracy, such as code generation and news reporting. Instances of AI chatbots providing false or misleading information have led to embarrassing failures and a loss of trust. The initial hype surrounding AI has set unrealistic expectations, leaving investors and Wall Street less optimistic about its immediate financial benefits.
In response to these challenges, renowned AI expert Yoshua Bengio is supporting the UK government-funded project, Safeguarded AI. The project aims to embed safety mechanisms into AI systems, enabling quantitative guarantees and risk assessments. By combining scientific world models and mathematical proofs, the project seeks to ensure the safety and accuracy of AI models through human verification.
In the field of robotics, researchers have made progress towards creating robots capable of performing skilled tasks in real environments. A robot wielding a 3D-printed paddle won 13 out of 29 table tennis games against human opponents, demonstrating the potential for robots to contribute to various practical applications.
While recent concerns and stock market fluctuations have sparked debates about AI’s transformative potential, it is premature to dismiss its long-term impact. Microsoft, in collaboration with OpenAI, continues to invest in generative AI and develop its own models to enhance its consumer AI efforts, aiming for greater independence from OpenAI.
On the other hand, the highly anticipated AI Pin, touted as a wearable AI assistant, has faced disappointing sales. The product’s return rate has exceeded its purchase rate, leading to e-waste accumulation as the returned pins cannot be reused.