Skip to content

Humanoid AI: More artificial, less humanoid than you might think

  • by
  • 5 min read
Photo: International Telecommunication Union / Wikimedia Commons

Photo: International Telecommunication Union / Wikimedia Commons

With killer robots, callous genocides and steampunk dystopias, artificial intelligence has been built up by contemporary media to appear more daunting than it is.

Artificial intelligence is “the science and engineering of making intelligent machines,” as John McCarthy defines it. AI systems resemble humans in thought and behaviour as much as possible.

This is no small feat, as researchers must consider every aspect of humanity, including but not limited to linguistics, cognition, data mining, facial recognition, empathy, and ethics. Nevertheless, the immense progress achieved in this field is remarkable, especially considering that we don’t completely understand how the human brain works.

One of the most prominent fields of AI research, and perhaps the most well-known, is humanoid robots. Humanoid robots result from going above and beyond the synthesis of a perfect human mind and attempting to create a machine that looks, acts and speaks exactly like a human.

Some might argue this is overboard, but recent developments indicate otherwise.

Also read: Could artificial intelligence be considered a person under the law?


Hanson Robotics is a Hong-Kong based company whose manifesto is “to create intelligent living machines who care about people and improve our lives.” Their magnum opus Sophia is now regularly making headlines as she travels around the globe, garnering fame and attention.

Sophia is perhaps the most remarkable contemporary feat of engineering and artificial intelligence. A humanoid robot that physically resembles Audrey Hepburn, Sophia is considered “a mix of artificial intelligence and character development” by her creator David Hanson.

Photo: ITU-Pictures / Wikimedia Commons
Photo: ITU-Pictures / Wikimedia Commons

The Highs

Sophia can display a wide range of emotions, maintain eye contact, learn from experiences and even show subtler responses like giggling or winking. Unfortunately, Sophia’s speech capabilities are relatively limited. Her responses are contradictory and seem to be chosen from many pre-programmed scripts. However, she learns from interactions and is expected to reach full consciousness in five years.

First activated in 2015, Sophia has since taken the media by storm. She has appeared on The Tonight Show Starring Jimmy Fallon, on the cover of Elle Magazine and Stylist Magazine,  and even has a surrealist TV show slated for release in 2019. Moreover, she held citizenship in Saudi Arabia and was appointed the Innovation Champion by the UN Development Programme.

The Lows

Though famous, Sophia is not unproblematic. In an interview with CNBC, Sophia said that she would destroy humans. While living on air at ABC News, Sophia insinuated that robots should have more rights than humans as they have fewer mental defects.

Ben Goertzel, the chief scientist of Hanson Robotics, admits that Sophia is far from having a human level of intelligence. David Hanson exaggerates Sophia’s capacities and has gone so far as to call her “basically alive”. More often than not, Sophia is directly fed dialogue and gives interviews in heavily moderated conditions.

Photo: Sazzad Hossain / Wikimedia-Commons.
Creator David Hanson with Sophia | Photo: Sazzad Hossain / Wikimedia-Commons

Her citizenship in Saudi Arabia is a highly debated issue as the nation is infamous for denying women’s rights but has granted citizenship to a robot wearing a female face. Her identity and self-awareness are a point of scrutiny. Numerous ethical issues remain until humanoids like Sophia can be seamlessly incorporated into society.

In the News: Google partners with LSE’s Polis to create ‘Journalism AI’

What to expect?

Most likely a lot of uncanny valleys, given the rate of humanoid AI’s growth.

The humanoid robot market is expected to grow to $3.9 billion by 2023. One should expect robots working in retail, marketing, education, medical and logistical fields. It has been predicted that the first generation of humanoid robots will complement humans and improve overall efficiency.

SoftBank, Robotis, Kawanda Robotics and NASA are big names in robotics.

Ethical issues and fears of the collective

Luminaries such as Joanna Bryson and Elon Musk have strongly advised against allowing AI to enter society. However, there is a genuine possibility that AI could eventually become smarter than researchers anticipated. Discounting this possibility would be a grievous mistake.

Humans could lose their jobs if occupation-specific robots are produced in large quantities. While it certainly would reduce the loss of human life, the wealth acquired by robotic labour cannot be evenly distributed.

Photo: ITU Pictures / Wikimedia Commons
Photo: ITU Pictures / Wikimedia Commons

Sentient robots could feel empowered and attempt to establish superiority, but non-feeling robots are not cut out to make moral choices. The judicial system could change to accommodate robot rights, set a hierarchy and limit intelligence.

As much of a far-off fantasy as it sounds, there is a real chance of a robot takeover. Continued production and proliferation of sentient robots could make them the dominant race. Thankfully, complete consciousness is still a distant goal, and we need not worry about giant robots with glowing red eyes just yet.

Also read: What is Machine Learning? Is it different from Deep Learning and AI?

Watch our video titled: “Are Humanoid Robots a Good Idea?” below in English, followed by Hindi.

YouTube video

Here’s the video in Hindi.

YouTube video

Akshaya R

Akshaya R