CEO Suggests That Humans Could Be 'Adversarially Attacked' Like Neural Networks

  • 📰 futurism
  • ⏱ Reading Time:
  • 32 sec. here
  • 2 min. at publisher
  • 📊 Quality Score:
  • News: 16%
  • Publisher: 68%

Technology Technology Headlines News

Technology Technology Latest News,Technology Technology Headlines

Science and Technology News and Videos

Apropos of almost nothing, the CEO of an AI startup suggested that a bad actor could use a mysterious image to "attack" human brains the way researchers have done with neural networks.

"No reason in principle why humans couldn’t be vulnerable to the same kind of adversarial examples," he wrote, "making a model think that a panda is a gibbon with 99.7% confidence because some seemingly random noise was added to it." Told from the perspective of Robbo, a right-wing terrorist who uses the imagery for his group's racist ends, the story follows the young man as he posts one such basilisk, called "The Parrot," around an unnamed British town, wearing goggles that distort his vision so that it does not harm him the way it does those who view it with their naked eyes.

 

Thank you for your comment. Your comment will be published after being reviewed.
Please try again later.
We have summarized this news so that you can read it quickly. If you are interested in the news, you can read the full text here. Read more:

 /  🏆 85. in TECHNOLOGY

Technology Technology Latest News, Technology Technology Headlines