Soapbox Artificial Intelligence

Why a data scientist warns against always trusting AI’s scientific discoveries

Data-mining algorithms aren’t good at communicating uncertainty in results, Genevera Allen says

data mining illustration

MACHINE EUREKA  It’s risky to take scientific discoveries made by AI at face value until these computer programs get better at assessing their own uncertainty, says data scientist Genevera Allen.

Dmytro Vikarchuk/Shutterstock

By Maria Temming

This article is only available to Science News subscribers.

Subscribers, enter your e-mail address to access our archives.

Not a subscriber?

Science News is a nonprofit.

Support us by subscribing now.

More Stories from Science News on Artificial Intelligence

  1. artificial neural network illustration
    Artificial Intelligence

    An AI that mimics how mammals smell recognizes scents better than other AI

    By Maria Temming
  2. robot resembling a child's head
    Artificial Intelligence

    Linking sense of touch to facial movement inches robots toward ‘feeling’ pain

    By Laura Sanders
  3. Nazca Line
    Archaeology

    An AI found a hidden Nazca Line in Peru showing a humanoid figure

    By Maria Temming
  4. pensive robot
    Artificial Intelligence

    A will to survive might take AI to the next level

    By Tom Siegfried
  5. doctor and patient
    Science & Society

    Bias in a common health care algorithm disproportionately hurts black patients

    By Sujata Gupta
  6. Poker game
    Artificial Intelligence

    Artificial intelligence has now pretty much conquered poker

    By Maria Temming
  7. Artificial intelligence gaming illustration
    Artificial Intelligence

    AI can learn real-world skills from playing StarCraft and Minecraft

    By Maria Temming
  8. number sense
    Artificial Intelligence

    A new AI acquired humanlike ‘number sense’ on its own

    By Maria Temming

From the Nature Index

Paid Content