Skip to main content
  1. Home
  2. Emerging Tech
  3. Legacy Archives

Robots can learn how to cook by watching YouTube videos, thanks to DARPA

Add as a preferred source on Google

Give a robot a cooking program and you feed it for a day; teach a robot how to learn by watching how-to cooking videos and you feed it for a lifetime. A Defense Advanced Research Projects Agency (DARPA) program has taken that modernized proverb to heart by enabling robots to learn how to cook by watching YouTube videos.

Three researchers from the University of Maryland and a researcher from the National Information Communications Technology Australia (NICTA) used funding from DARPA’s Mathematics of Sensing, Exploitation and Execution (MSEE) program to develop a system that helps robots learn manipulation actions by watching “unconstrained” Internet videos – in this case, how-to cooking videos on YouTube.

Recommended Videos

The robots in this research endeavor were able to “recognize, grab and manipulate the correct kitchen utensil or object and perform the demonstrated task with high accuracy – without additional human input or programming,” according to DARPA’s press release.

The research paper, “Robot Learning Manipulation Action Plans by ‘Watching’ Unconstrained Videos from the World Wide Web,” notes two levels to this system: a lower level, which helps robots recognize hand grasp type and objects; and a higher level, “where a robot processes visual cues through a manipulation action-grammar module and translates them into actions,” according to Reza Ghanadan, program manager in DARPA’s Defense Sciences Offices.

This system also enables robots to continuously build on their previous learning, which opens up the door not only to robots that can learn faster but also to robots sharing their knowledge with other robots. “This learning-based approach is a significant step towards developing technologies that could have benefits in areas such as military repair and logistics,” Ghanadan said.

It’s only a matter of time before we see robots vs. humans cooking shows on TV. Let’s just hope the robots don’t learn their manners from YouTube comments.

Jason Hahn
Former Contributor
Jason Hahn is a part-time freelance writer based in New Jersey. He earned his master's degree in journalism at Northwestern…
AI chatbots can often feed into your delusions. Researchers say you should look for three signs
Experts warn that chatbot design choices can reinforce unhealthy beliefs in vulnerable users.
ChatGPT on a smartphone

Artificial intelligence chatbots have become incredibly good at sounding human. But a new review paper by psychiatrist Marc Augustin and fellow researchers Thomas A. Pollak and Helen Morrin, published in NPP—Digital Psychiatry and Neuroscience, argues that existing AI research points to an overlooked psychological risk. The paper, highlighted by The Wall Street Journal, reviews previous studies and proposes a framework explaining how three common chatbot behaviors can combine to reinforce delusional thinking in vulnerable users, creating what the authors call an "amplification spiral."

Researchers say these are the three warning signs

Read more
Lost access to your crypto wallet? Don’t Google your way out of it
Security researchers warn that fake recovery tools are becoming the latest trap for crypto owners.
Bitcoin crypto wallet featured

Forgetting the recovery phrase to a crypto wallet can be stressful enough. Unfortunately, that's exactly the moment scammers are waiting for. A new warning highlights a growing scam in which cybercriminals disguise malware as cryptocurrency recovery software, tricking desperate users into handing over far more than just access to their wallets.

The fake recovery tool that's actually malware

Read more
Chinese AI lab says it can match Anthropic’s all-poweful Claude Mythos at sniffing security bugs
Security researchers say Z.ai's latest model can rival Anthropic's Mythos in one critical area.
China Z.Ai GLM-5.2 Featured Banner

For the past few weeks, Anthropic's Mythos has been viewed as the gold standard for AI-powered cybersecurity. That lead may already be shrinking. According to a new report from The Wall Street Journal, security researchers say Chinese AI startup Z.ai's GLM-5.2 can now match Mythos when it comes to finding software security vulnerabilities, even if it still trails Anthropic and OpenAI in broader reasoning tasks.

GLM-5.2 is closing the gap in one very important area

Read more