Generative AI is a hot topic in the world of technology. As someone who has reported on tech for a long time, I have seen many trends come and go. Sometimes, the excitement does not match the reality. So, when generative AI started gaining attention, I was cautious. But this time, it is different. Projects like ChatGPT and DALL-E have emerged as game-changers and they are here to stay.
The significant difference between generative AI and previous tech hype cycles is its immediacy. You don’t have to wait for years to see its potential. You can experience it right now. For example, a person with minimal tech knowledge can sit at a computer, type a few words into a dialogue field and watch as the system generates paintings and short stories. There is no need to visualize how it might look years down the road. It is already here. However, this instant gratification can make it challenging to manage expectations.
Generative AI holds significant promise for the field of robotics. Many experts in robotics see it as a key player in shaping the future of this technology. It is enabling robots to acquire new skills with remarkable ease. This means robots can become more versatile without the need to change their fundamental programming. This innovative method, called diffusion policy, utilizes generative AI techniques to instruct robots and it has already been successfully employed for over 60 different skills.
Generative AI is also simplifying robot design. Researchers are using this technology to create robots, enabling rapid evolution and bypassing the lengthy evolutionary process. A recent example from Northwestern University showcases a robot that can walk successfully It is designed in seconds by an AI algorithm. This approach, known as “instant evolution,” led the AI to choose legs for the robot. It illustrates that AI can rediscover effective solutions without human biases.
