Subscribe now
A law for AI hiring tools is about to come into effect in New York

AI hiring tools to be audited for sexism and racism under New York law

29 June 2023

A New York City law is requiring companies to show that the AI-powered tools they use in hiring and promotion are not biased. Critics say it lets most employers off the hook


Mathematician makes breakthrough on 100-year-old problem about knots

Mathematician makes breakthrough on 100-year-old problem about knots

5 February 2021

Some tangles that look like knots are actually “the unknot”, which can be untangled into a simple loop, and a new algorithm has sped up the process of finding them


Picture of young couple using smartphones while dating together

Voice assistant recordings could reveal what someone nearby is typing

4 December 2020

Tapping on devices that is detected by voice assistants could potentially be used to deduce what a person is typing on their phone up to half a metre away


Uber and Lyft pricing algorithms charge more in non-white areas

Uber and Lyft pricing algorithms charge more in non-white areas

18 June 2020

The algorithms that ride-hailing companies such as Uber use to determine fare prices appear to set higher prices for non-white neighbourhoods


digital human

Battling AI algorithm tested on a quantum computer for first time

25 January 2019

One of the most powerful techniques in machine learning, generative adversarial networks, has been tested on a quantum computer for the first time


How to stop artificial intelligence being so racist and sexist

How to stop artificial intelligence being so racist and sexist

4 July 2018

AI his frequently be biased, but a new technique may be able put fairness right at the heart of training algorithms


Playing Atari games

AI cheats at old Atari games by finding unknown bugs in the code

27 February 2018

An AI found a bug in the Atari game Q*bert and exploited it to quickly score a million points. It used self-destruction as a winning strategy too


A picture of a dog in the snow

Fooling AI can now be done a thousand times faster

20 December 2017

By changing an image pixel by pixel, neural networks can be tricked into thinking a dog is two people skiing


We’ve figured out how to ensure quantum computers can be trusted

We’ve figured out how to ensure quantum computers can be trusted

7 November 2017

Quantum computers will be useless if we can't trust their calculations. Now, two teams have programmed quantum systems to detect their own errors


AI could put a stop to electricity theft and meter misreadings

AI could put a stop to electricity theft and meter misreadings

23 September 2017

An algorithm developed to detect electricity theft for companies in Latin America could also uncover billing errors


Sign up to our weekly newsletter

Receive a weekly dose of discovery in your inbox. We'll also keep you up to date with New Scientist events and special offers.

Sign up
Piano Exit Overlay Banner Mobile Piano Exit Overlay Banner Desktop