Sign up to read our regular email newsletters

Subscribe and save

Google's AI is best yet at answering medical and health questions

Google has built an AI that can answer medical questions. However, it's not as good as a human doctor and the company says it cannot yet perform safely in the real world

Technology 20 January 2023

By Jeremy Hsu

A laptop sits on a desk, opened to the homepage for the search engine Google. A person's hands enter from the left, with one hand's fingers placed on the laptop's keyboard.

Google’s medical AI can answer common medical questions entered in search engines.

Africa Studio/Shutterstock

Google has built the best artificial intelligence yet for answering medical questions. The Med-PaLM AI can answer multiple-choice questions from medical licensing exams and common health queries on search engines with greater accuracy than any previous AI and almost as well as human doctors.

Alan Karthikesalingam and his colleagues at Google first started working with an AI the company developed called PaLM, which was trained on a wide range of text types, including online documents, books and Wikipedia, …

Continue reading

Subscribe today

in our January Sale

VIEW JANUARY SALE OFFERS

No commitment, cancel anytime*

Offer ends 14th March 2023. *Cancel anytime within 14 days of payment to receive a refund on unserved issues.

Existing subscribers, please log in with your email address to link your account access.

Inclusive of applicable taxes (VAT)

Advertisement
  1. Vagus nerve receptors may be key to controlling inflammation
  2. A better understanding of PCOS offers fresh hope for new treatments
  3. Egyptian boy mummy was buried with a ‘second heart’ made of gold
  4. The mysterious underwater avalanches reshaping Earth
  5. Wolves are eating sea otters after depleting an Alaskan island’s deer
  1. Dogs can tell when you want to give them a treat – even if you don’t
  2. Type 2 diabetes linked with an increased risk of dying from cancer
  3. The mysterious underwater avalanches reshaping Earth
  4. Fake mushroom experiment reveals why some fungi glow in the dark
  5. The first ever vaccine against RSV could be approved in 2023
  1. The future of food: How to farm insects for a growing population
  2. Livestream: First satellite launch from the UK
  3. The scientific guide to a better Christmas dinner
  4. Porvenir massacre: Did US army have larger role in 1918 killings?
  5. Watch Orion capsule splash down as NASA’s Artemis I mission ends
  1. Dogs can tell when you want to give them a treat – even if you don’t
  2. Type 2 diabetes linked with an increased risk of dying from cancer
  3. The first ever vaccine against RSV could be approved in 2023
  4. Rare Antarctic meteorite is one of the largest ever found
  5. Egyptian boy mummy was buried with a ‘second heart’ made of gold

MORE FROM NEW SCIENTIST

Researchers have suggested that monitoring audio clips on social media could be a relatively quick, inexpensive way of tracking coronavirus cases in the community

AI detects if YouTubers are infected with omicron coronavirus variant

An artificial intelligence picked up on audio samples where the speaker was probably infected with omicron with 80 per cent accuracy, potentially offering an inexpensive way of tracking cases
Health
A woman sitting at her desk

Smart office chair recognises what position you are sitting in

An office chair fitted with sensors detects how you are sitting and uses artificial intelligence to classify it – the long-term aim is to tell you how to improve your posture
Technology
Launching vehicle loaded with an ORCA type detection unit KM3NeT - LOM-on-anchor.jpg KM3NeT the next generation neutrino telescopes KM3NeT is a research infrastructure housing the next generation neutrino telescopes. Once completed, the telescopes will have detector volumes between megaton and several cubic kilometres of clear sea water. Located in the deepest seas of the Mediterranean, KM3NeT will open a new window on our Universe, but also contribute to the research of the properties of the elusive neutrino particles. With the ARCA telescope, KM3NeT scientists will search for neutrinos from distant astrophysical sources such as supernovae, gamma ray bursters or colliding stars. The ORCA telescope is the instrument for KM3NeT scientists studying neutrino properties exploiting neutrinos generated in the Earth's atmosphere. Arrays of thousands of optical sensors will detect the faint light in the deep sea from charged particles originating from collisions of the neutrinos and the Earth. The facility will also house instrumentation for Earth and Sea sciences for long-term and on-line monitoring of the deep sea environment and the sea bottom at depth of several kilometers.

Help in the hunt for neutrinos while exploring deep-sea ecosystems

The Deep Sea Explorers project is calling for volunteers to help remove noise from data collected by a neutrino telescope at the bottom of the sea, finds Layal Liverpool
Earth
ChatGPT chatbot on phone screen

Huge AI models can be halved in size without degrading performance

A way to cut the scale of artificial intelligence models by 60 per cent could save huge amounts of energy and make them more accessible
Technology

Sign up to our newsletters