Tesla's Optimus robot sorts coloured blocks and strikes yoga poses

Video suggests robots have made great strides after struggling to get on-stage at unveiling event last year

Tesla chief Elon Musk has said the Optimus robots will be 'friendly'. AFP

Tesla's Optimus robot, which the company's chief Elon Musk claimed was a cut above other robots, can now perform advanced tasks, showing the progress that has been made after its initial struggles.

Optimus, officially unveiled about a year ago with an expected cost of less than $20,000, can now sort blocks based on colour and perform yoga poses, according to a short video posted on X, the platform owned by Mr Musk and formerly known as Twitter.

The update suggests there have been vast improvements in the robots compared with when they first appeared during the Tesla AI Day, where one of the Optimus robots walked slowly on stage and another had to be wheeled on.

Mr Musk said at the time that it was the first time Optimus moved without any support.

Fast forward to today and "Optimus is now capable of self-calibrating its arms and legs", text from the video read.

"Using only vision and joint position encoders, it can precisely locate its limbs in space.

"Accurately calibrated, Optimus learns various tasks more efficiently. Its neural net runs entirely on-board, using only vision."

Notably, the video showed Optimus sorting blocks of only two colours, green and blue. It did, however, show "autonomous corrective action capabilities", specifically by picking up a block and placing it in an upright position.

Optimus, described by Tesla as a "general purpose, bi-pedal, humanoid robot capable of performing tasks that are unsafe, repetitive or boring", was touted as being the future of robots, aimed at aiding people in productivity.

The adult-sized robot, set to be about 173cm tall and weighing about 57kg, was originally to be programmed to do monotonous jobs, such as assembling car parts, moving components around factories and picking up groceries.

Sceptics have voiced concerns about the potential dangers of AI-powered robots, ranging from industrial concerns, including taking jobs away from humans, to more serious issues such as harming people.

Geoffrey Hinton, one of the leading voices in the field of AI and a pioneer of deep learning, sounded his warning on AI and robots shortly after he left Google in May.

He said he was worried that future iterations of AI could become a threat to humans because of their unexpected behaviour. He also dreads the day that truly autonomous weapons – “killer robots” – may become a reality.

Mr Musk sought to allay fears during Optimus's initial unveiling, saying the robot would be "friendly" and that it was being built for humans, by humans.

“At a mechanical level, you can run away from it – and most likely overpower it,” he said.

Proponents of AI and robots are also on a charm offensive to strike down negative notions about the technology.

At the UN AI for Good Global Summit in Geneva in July, humanoid robots held a press conference to assure the public they will not take jobs away from humans. They also called for a cautious yet enthusiastic approach to AI regulation.