We keep track of when our discoveries are reported in the press and when members of the Institute report on the discoveries of others. We think it’s important to publicise our best work, because researchers are able to build upon it only to the extent that they know about it.

  • Nature

    Nature21 Mar

    Roll over, Newton

    A robotic metamaterial has been made that violates Newton’s third law of motion, allowing for the propagation of topological solitons.

  • Quanta

    Number theory5 Mar

    Elliptic curve mystery

    Quanta reports on work by Yang-Hui He, who co-discovered unexpected patterns in a property related to the curves’ integer roots using AI.

  • The telegraph

    Evolvability11 Aug 2022

    Price of immortality

    Like Orpheus in the Underworld, the London Institute is challenging mortality, says our writer Thomas Hodgkinson in The Sunday Telegraph.

  • New Scientist

    AI-assisted maths6 Jun 2022

    AI helps with maths

    An AI that can turn mathematics problems written in English into a formal proving language could make them easier for other AIs to solve.

  • Harvard Business Review

    Innovation24 Feb 2020

    Taming complexity

    Complexity may be hard to unpick, without being inherently bad. Ensure the benefits of any addition to company systems outweigh its costs.

  • APS Physics

    Thermodynamics31 Jul 2018

    Slurry in a hurry

    The 3D structures of slurries—fluids full of solid particles—can be swiftly measured using a single 2D shot and electron diffraction data.

  • The Washington Post

    Discoveries3 Mar 2018

    Whatever you say

    If you meet a conspiracy theorist, don't bother trying to change their views. Encountering the truth only makes them more pig-headed.

  • Nature Physics

    Innovation25 Sep 2017

    Yes you cayenne

    In innovation, the most apparently niche ingredients may turn out to be the most useful, as the structures of recipes become more complex.

  • Bloomberg Opinion

    Discoveries10 Jun 2017

    Moore means less

    Following Moore's law, solar power will become ever cheaper as an energy source—and there’s nothing Donald Trump can do about it.

  • Bloomberg

    Markets10 Apr 2017

    A little bird told me

    Twitter sentiment during busy periods, such as ahead of quarterly earnings releases, provides some indication if a stock will rise or fall.

  • Phys.org

    Financial risk5 Apr 2017

    Fools rush in

    Measures meant to stabilise economies may have the opposite effect, creating cyclical structures in the networks of contracts between banks.

  • Scientific American

    Innovation6 Jan 2017

    Harnessing Serendipity

    Quirky and apparently mysterious, innovation is critical to sustained economic growth—and mathematics can help us understand how it works.

  • INC.

    Tech progress12 May 2016

    The future’s bright

    Architects are designing rotating homes to increase the efficiency of solar power, while its cost is set to keep falling by 10% annually.

  • The Guardian

    Tech progress26 Jan 2016

    Here comes the sun

    The cost of solar power will continue to fall by 10% annually, meeting 20% of global energy needs far sooner than has been predicted.

  • Science News

    Markets25 Jan 2015

    A stock response

    The simultaneous study of news sentiment and browsing behaviour, even on small time-scales, can help to predict stock market fluctuations.

  • Nature World News

    Network theory4 Oct 2014

    Beauty in repairability

    The hunt for networks that best combine efficiency with repairability, to avoid breakdown, leads to structural designs that resemble snowflakes.

  • Discoveries

    Network theory3 Oct 2014

    Snowflakes don't break

    Snowflake-shaped networks, with redundant arms that come into use when main branches break down, are easiest to fix when disaster strikes.

  • Discoveries

    Fractals20 Feb 2013

    Towers of strength

    The Eiffel tower is now a longstanding example of hierarchical design due to its non-trivial internal structure spanning many length scales.