When It Comes to AI Models, Bigger Isn't Always Better

Trouble viewing? View in your browser. View all Scientific American publications.

    

November 21, 2023

Large artificial intelligence models, such as the one that powers OpenAI's ChatGPT, can require massive amounts of code and training data. While that size helps the AI produce impressively human-like text, it also weighs the program down. "As models have gotten bigger, they've also become more unwieldy, energy-hungry and difficult to run and build," tech reporting fellow Lauren Leffer writes in this week's featured story. This is why some developers are experimenting with slimmed-down AI systems—which are still capable of surprisingly strong performances.

Sophie Bushwick, Associate Editor, Technology

Artificial Intelligence When It Comes to AI Models, Bigger Isn't Always Better Artificial intelligence models are getting bigger, along with the data sets used to train them. But scaling down could solve some big AI problems By Lauren Leffer

Defense If You Had a Nuclear Weapon in Your Neighborhood, Would You Want to Know about It? The Fort Berthold Reservation in North Dakota has had nuclear missile silos on its land for decades. Now the U.S. government wants to take the old weapons out and replace them with new ones, and it's unclear how many living there know about that. By Ella Weber | 15:36

Robotics Robotics 'Revives' a Long-Extinct Starfish Ancestor Engineers and paleontologists teamed up to reconstruct an ancestor of starfish from the Paleozoic era and figure out how it moved By Lauren Leffer

Defense The New Nuclear Age The U.S. is beginning an ambitious, controversial reinvention of its nuclear arsenal. The project comes with incalculable costs and unfathomable risks

Energy This Biophysicist 'Sun Queen' Harnessed Solar Power Hungarian-American biophysicist and inventor Mária Telkes illuminated the field of solar energy. She invented a solar oven, a solar desalination kit and, in the late 1940s, designed one of the first solar-heated houses By Johanna Mayer,Katie Hafner,The Lost Women of Science Initiative

Energy Dyslexia, Dark Energy and a New Arms Race The risks of a $1.5-trillion plan to build up the U.S. nuclear arsenal By Laura Helmuth

Psychology Machine Learning Creates a Massive Map of Smelly Molecules Scientists can finally predict a chemical's odor without having a human sniff it By Simon Makin

Education To Educate Students about AI, Make Them Use It A college professor and his students explain what they learned from bringing ChatGPT into the classroom By C.W. Howell,Cal Baker,Fayrah Stylianopoulos

Artificial Intelligence Drones and AI Could Locate Land Mines in Ukraine An AI model could speed up laborious and dangerous demining efforts By Lori Youmshajekian

QUOTE OF THE DAY "A recent study found that human children absolutely crush AI tools in basic problem-solving and thinking tasks, with scientists determining that AI has one serious blind spot: innovation." Maggie Harrison, Futurism

FROM THE ARCHIVE The AI Boom Could Use a Shocking Amount of Electricity Powering artificial intelligence models takes a lot of energy. A new analysis demonstrates just how big the problem could become By Lauren Leffer | October 2023

LATEST ISSUES

Questions?   Comments?

Download the Scientific American App

To view this email as a web page, go here. You received this email because you opted-in to receive email from Scientific American. To ensure delivery please add newsletters@scientificamerican.com to your address book. Unsubscribe     Manage Email Preferences     Privacy Policy     Contact Us

Scientific American

1 New York Plaza, FDR Dr, Floor 46, New York, NY 10004

Unsubscribe - Unsubscribe Preferences