There were the ancient Egyptians, who built statues of gods that, when hit by the sun, appeared to talk to astonished crowds, and Leonardo da Vinci, who designed one of the very first robots, a robotic knight which was able to sit, stand, and move its arms.

Those, of course, were one-off curiosities, but now robotics is big business and has been for years.

The manufacturing sector has relied on robots on production lines for decades, and consumers have embraced things like automated vacuum cleaners, toys, and teaching aids.

Most of the robots we use today don't look like us, but the advent of AI has propelled the industry forward.

And 2024 saw record levels of funding for humanoid robotics.

So why the obsession with making robots human?

An often-cited reason is the embodiment theory, that is, we're creating machines that need to function in a human-designed world.

Take a door, for instance.

A robot needs to be human-sized to navigate getting through one.

Too tall or wide and it won't fit, too short and it can't reach the door handle.

To turn the handle, the robot will need dexterous hands.

And having two legs means it can step over any obstacle, climb stairs, and generally take up less space.

Then there's the societal argument.

Some say we create machines that look like us because it helps with interaction and collaboration.

We simply respond better to things if we can identify with them.

Some psychologists also think that it adds credibility.

If robots take human form, then they're seen as more intelligent.

There is a flip side, though, and it's known as the uncanny valley.

First coined by robotics professor Masahiro Mori in 1970, the hypothesis states that we generally react positively to human-like machines, but as they become more human-like, we develop unease or even repulsion.

Curiously, the hypothesis goes on to state that as robots become almost indistinguishable from humans, we once again view them positively.

Getting the uncanny valley right, then, is a big focus for developers.

One of the most high-profile proponents of human-like androids is Elon Musk.

Tesla unveiled the first generation of its robot Optimus in 2022, and the company is hoping to deploy the robots in its factories by the end of 2024.

They'll be available for purchase more widely the year after.

This is a key reason behind why we're making these robots.

We want them to work for us.

Globally, there is a massive shortage of workers in key sectors like manufacturing.

According to one estimate, up to 8 million factory jobs globally could be left unfilled by 2030.

Having robots to do some of these repetitive and mundane tasks could go some way to filling that gap.

They can also help fill dangerous jobs.

NASA, for instance, has developed Valkyrie, a humanoid robot it says can be used in challenging environments such as natural disaster areas or offshore oil rigs.

The end goal is to send Valkyrie to space, where it can focus on tasks like maintaining the spacecraft, freeing up exploration time for human astronauts.

However, there are still hurdles to overcome.

Making a human-like machine is more complex and more expensive than ordinary robotic systems, and with more components, there are more things that can go wrong.

Bias is another issue.

Like in AI, some critics point out we're creating machines that simply perpetuate stereotypes.

Robots in healthcare are often given more female characteristics, and those doing manual labor, male ones.

And then there are the worst-case scenarios, often depicted in science fiction.

Could robots gain consciousness and turn on their creators, like in Ex Machina or Blade Runner?

Could they be built with the purpose of harming humans, like in The Terminator?

One thing is certain, growing investment into the sector means we're going to see a lot more robots around us.

Billionaire venture capitalist Vinod Khosla predicts we could see a billion bipedal robots by 2040, and Goldman Sachs estimates that the global market for humanoid robots could reach $38 billion by 2035.