Human and robotic hands meet in the middle and touch the knuckle.
Getty
Artificial intelligence is moving forward with speed. Both momentum and money are focused on performance. Fast model, increase integration, and more accurate predictions. But as the industry sprints towards artificial general information (AGI), one question remains in the background. What happens to humans?
A recent report from Elon University’s imagining the Digital Future Center surveyed nearly 300 global technology experts. The resulting report, “Being Human in 2035,” concluded that most people are concerned that deepening their adoption of AI systems over the next decade will negatively change the way humans think, feel, act and engage with each other.
MIT Media Lab is trying to answer equally surprising questions. How can AI support human prosperity rather than replacing it?
This is a central issue with the lab’s newly launched Advance Humans, using the AI (AHA) program.
A star-studded symposium has launched the concept and the various areas of research it tackled, leading the way as a bold, multi-year initiative to not only improve AI, but also to enhance human prosperity in an AI-saturated world. Speakers included Arianna Huffington, who said that AI was like a “GPS for the soul,” and Tristan Harris, who warned about systems that exploit human vulnerability under the aid of aid. Both agreed that AI should not be optimized for efficiency alone, but should be designed to foster wisdom, resilience and reflection.
This reflected a deeper vision of AI that reinforces the development of AI on human interior design.
Pat Pataranutaporn, co-leading of the AHA program, sums this up to the framed audience and says, “If we are to underestimate human intelligence and undermine human dignity, what is the point of advancing artificial intelligence?
Researchers with a lack of AI
Safety and integrity dominate the debate on AI ethics, but AHA is interested in long-term human outcomes as it incorporates sections of events that cover interior life, social life, professional life, brain life and creative life.
From excessive dependence and skill atrophy to growing emotional attachment and isolation, people are already restructuring their lives around AI. However, there is little research effort dedicated to systematically understanding these changes. This goes without saying that this is about designing AI to mitigate them. Aha aims to do just that. This initiative is based on six research domains.
Understanding and Agency: An AI system that supports, rather than replacing critical thinking and reasoning. Mental and physical health: A tool that enhances long-term health and provides humanized support. Curiosity and Learning: Adaptive learning that triggers the creativity and exploration of each person’s learning journey: a system that amplifies human agents and expression rather than homogenizes. A sense of purpose: AI that helps people reflect, plan and align their values and goals while contributing to society. Healthy Social Life: AI that strengthens the connections of real people rather than reduce it.
Moonshot’s idea
The AHA’s ambitions are consistent with its Moonshot project. These include:
Atlas of human interaction: Promote comprehensive database mapping of how humans relate to AI across contexts. The prototype maps papers on human interaction to identify positive outcomes. Benchmarks for Human Prosperity: This develops a set of benchmarks that provide a rigorous metric and assessment framework for measuring how AI systems contribute to human prosperity across cognitive, emotional and social aspects. The September 17th and 18th workshop at Media Lab invites experts to begin designing benchmarks. AI Impact Global Observatory: An observatory for tracking the emotional, cognitive and social impacts of AI adoptions around the world, enabling large-scale randomised experiments on how people respond to AI in different countries and locations.
The message is clear. It’s time to measure not only the performance of machines, but also the happiness of humans.
Why now?
As AI becomes increasingly integrated into health, education, work and social life, the choices made by today’s engineers and designers will shape cognitive habits, emotional norms and social structures for decades.
However, as Aha contributors pointed out through the symposium, AI is still optimized for business metrics and safety concerns rather than psychological nuances, emotional growth, or long-term happiness.
MIT’s AHA initiative is not a criticism of AI. This is a call for better designs to design not only smarter machines, but systems that reflect us as our best self.
Professor Pattie Maes, co-leading the AHA program and director of the Fluid Interface Group, said after the event, “We’re creating AI and AI, and we don’t want to make the same mistakes as social media. It’s important to think of AI as a problem of human design, not just as a technical problem for engineers and entrepreneurs to solve, but also as a problem of human design, requiring the expertise of human interaction designers, psychologists and social scientists, and requires that AI beneficially impact the human experience.
