Sergey Lonshakov: Robots Cannot Achieve Autonomy Without Goal Setting
Governments are unlikely to accept the economic autonomy of robots, as it remains impossible to determine their motives. This was discussed by Sergey Lonshakov, architect of the “Robonomics” project, during the “Podcast Society” on ForkLog’s YouTube channel.
Specifically, courts cannot process disputes involving a robot that has failed to complete a task.
“A robot’s goal setting is fundamentally unclear—why it didn’t perform as expected. The key issue in robotics ethics is the inability to fully integrate robots into the economic framework,” explained Lonshakov.
This issue leads to a second problem—who should be held accountable for the mistakes of an autonomous agent.
“Any robot is a high-tech device. In some countries, there are already legal amendments stating that if a device is the result of the work of numerous scientific teams and something goes wrong, no one can be held responsible,” noted the speaker.
Ultimately, Lonshakov hypothesized that humanoid robots, due to their complexity, increasingly resemble the bio-robotic components of humans.
“Then the answer lies in the most unsettling area—we lack goal setting as well. We are simply a very complex neurobehavioral model. In the future, if large behavior models become adequate, the issue of goal setting will be resolved,” concluded the expert.
Subscribe to the podcast:
Рассылки ForkLog: держите руку на пульсе биткоин-индустрии!