[ad_1]
New software could reduce training costs for robots to perform tasks common to humans. A group of researchers from Silicon Valley has designed a system to adapt to a robotic hand that seems more real.
"We are now aiming to conquer more complex tasks," said Lilian Weng, a member of the OpenAI technical team.
The researchers injected random noise into the simulation program, making the virtual world of the robotic hand confusing enough to not be overwhelmed by unexpected elements in the real world.
The experts have tried to reduce the time of robotic processes and distribute them to several computers for a simulation program that can be done in a few hours or days, without human help.
[ad_2]
Source link
