The Future of AI: A Global Brain to Control Millions of Robots | Festina Lente - Your leading source of AI news | Turtles AI

 Stuart Russell, eminent AI expert argues that the future of AI may see the creation of a global brain, which will control millions of robots rather than individual autonomous robots. His work focuses on the ethics of AI and the social and political implications of such developments.

Key points:

•  Advanced AI could be centralized in a single global system.
•  Russell is an activist against the use of autonomous weapons and for the regulation of AI.
•  His vision of AI aims to improve compatibility between machines and human values.
•  He believes AI must be carefully governed to avoid social and labor risks.

Stuart Russell, a pioneer in the field of AI, is known for his forward-looking vision of future technological developments, which could involve the integration of a global brain, a single central control system, capable of directing millions of robots. Unlike a model in which robots act individually, this structure would allow for more efficient and safe management of resources, reducing the need to create individual autonomous intelligences for each machine. This approach, which Russell explores both in academia and through the “Center for Human-Compatible AI” at UC Berkeley, is based on the belief that it is possible to develop human-compatible AI systems while minimizing the risks of misinterpretation of goals by machines.

One of Russell’s main concerns is the lack of inherent understanding on the part of AIs. Although these are capable of mimicking human behaviors, they do not possess true emotional awareness or understanding, as stated during various talks. This becomes especially crucial when it comes to entrusting machines with decisions that affect human lives, as in the case of autonomous weapons, against which Russell is strongly committed.

Russell also warns about the potential impact of AI on the labor market, pointing out how automation could wipe out entire industries, such as transportation, without adequate government intervention. This kind of AI-accelerated “creative destruction” is likely to be too sudden and disorderly to be absorbed naturally by society. The future he envisions, however, is not without opportunity: the reduction of traffic accidents and the use of AI in health care are just some of the expected benefits, provided the technology is regulated ethically and safely.

Russell concludes that to avoid the dystopian scenarios so feared, humanity must rethink the way it defines and assigns goals to machines. AI that follows ill-defined goals to the letter can pose catastrophic risks. Only with a solid ethical foundation and strict regulation can we ensure that AI not only improves our lives, but does so while respecting our values and safety.

Russell earned an honors degree in physics from Oxford University in 1982, followed by a doctorate in computer science from Stanford University in 1986. He later joined the faculty at the University of California, Berkeley, where he currently directs the Center for Human-Compatible AI, working to ensure that advanced AI technologies remain under human control. He has also earned numerous awards, including the Order of the British Empire (OBE) in 2021 and the position of Honorary Fellow at Wadham College, Oxford.

In addition to his academic contributions and pioneering research in areas such as probabilistic reasoning and machine learning, he is best known for co-writing with Peter Norvig “Artificial Intelligence: A Modern Approach,” the world’s reference text on AI. Russell is actively engaged in ethical and political issues regarding the use of AI technologies, particularly with regard to autonomous weapons and their potential negative impact on humanity.

 The future of AI is not yet written, but it will depend on the decisions we make today in terms of regulation, ethics and the design of machine targets.