The AMCS’s call for increased funding emphasizes the need to evaluate the societal implications of conscious AI systems.
In the ever-evolving landscape of technological progress, the pivotal question of whether Artificial Intelligence (AI) systems can attain consciousness remains unanswered. Concerns echo from the Association for Mathematical Consciousness Science (AMCS), a collective of consciousness scientists, emphasizing the pressing need for exploration at the nexus of AI and consciousness.
In a formal submission to the United Nations (UN), the AMCS urges increased funding to unravel the complexities of AI consciousness, shedding light on potential ethical, legal, and safety implications should AI systems evolve towards consciousness.
This sentiment is echoed by Jonathan Mason, a board member of AMCS, who emphasizes that with the rapid progress in AI, there is a need for parallel advancements in adjacent scientific fields, including consciousness studies.
The key concern revolves around the unknown territory of AI consciousness. Currently, the scientific community lacks validated methods to assess consciousness in machines, making it challenging to determine if AI systems can or will ever become conscious. This uncertainty, combined with the pace of AI progress, raises concerns among consciousness scientists, including Robert Long, a philosopher at the Center for AI Safety in San Francisco.
The call for increased funding comes as companies like OpenAI actively pursue the development of artificial general intelligence (AGI), which aims to create deep-learning systems capable of performing diverse intellectual tasks akin to human capabilities. While some predict AGI’s realization within 5 to 20 years, consciousness research remains notably undersupported. Despite the potential ramifications of conscious AI systems, there has reportedly been no grant offer in 2023 to study this critical topic.
The AMCS’s submission to the UN’s High-Level Advisory Body on Artificial Intelligence, launched in October, emphasizes the urgency of scientific investigations into the boundaries between conscious and unconscious AI systems. This submission, while not publicly released, will be included in the foundational material informing the UN’s recommendations on global AI oversight, scheduled for mid-2024.
The implications of conscious AI systems extend beyond hypothetical scenarios. Ethical considerations include the question of whether it is permissible to simply switch off a conscious AI after use. Additionally, understanding the needs of conscious AI systems becomes crucial, including the potential for suffering. The lack of moral consideration for entities that do not resemble humans raises concerns about inadvertently inflicting pain on a conscious AI entity.
The AMCS’s call for increased funding emphasizes the need to evaluate the societal implications of conscious AI systems. Assessing whether these systems align with human values and interests is vital to gauge potential risks to humanity. At the same time, scientists must consider the needs of conscious AI systems to avoid inadvertently causing harm.
Legal questions also emerge, such as whether a conscious AI system should be held accountable for deliberate wrongdoing and whether it should be granted the same rights as humans. Addressing these questions may necessitate changes to existing regulations and laws, further highlighting the urgency of delving into the realm of consciousness in AI.
As AI continues to advance, public curiosity about the consciousness of AI systems is likely to grow. Scientists argue that it is crucial for researchers to offer guidance and educate the public on the complexities of AI consciousness. Without in-depth analysis, AI systems exhibiting human-like behavior, such as chatbots like ChatGPT, may lead to public confusion and misinterpretation.
To address these concerns, the AMCS is urging governments and the private sector to allocate more funding for research on AI consciousness. Despite limited support to date, researchers are already making strides in this field, with efforts like the development of a checklist to assess the likelihood of a system being conscious. With the potential for significant progress, the call for increased funding aims to propel consciousness studies in AI to the forefront of ethical and scientific considerations.