• Skip to main content
  • Skip to primary sidebar

Reefwing Software

  • Home
  • Book
  • Development
  • Blog
  • Contact Us
  • Privacy
You are here: Home / AI / Groupthink — Why AI Agent Groups Fail

September 23, 2026 by David Such Leave a Comment

Groupthink — Why AI Agent Groups Fail

In 1985, the psychologists Garold Stasser and William Titus ran what became one of the most replicated experiments in group decision-making. Four-person groups had to choose among three candidates for student body president. Each candidate was described by sixteen characteristics. The trick was in the distribution. The best candidate’s virtues were scattered across group members so that no individual could see the full picture, while a mediocre candidate’s virtues were known to everyone. The complete evidence pointed clearly to one answer. Each individual’s data pointed in the wrong direction.

When every member received all the information, 83 percent of groups chose the best candidate. When the information was distributed, 18 percent did.

This August, Anthropic ran the same test on teams of four AI agents deciding realistic questions: which candidate to hire, which investment to make, which property to buy. Shared evidence favored the wrong option. Each agent held unique, decisive private facts favoring the right one. Solving the task required an agent to recognize its private information as pivotal and press it against an apparent consensus.

A single agent handed the entire evidence base got the answer right nearly every time. Groups of agents, after discussion, got it right in 17 to 36 percent of runs for most model families. 

Read More…

Filed Under: AI, App Development, Embedded, Robotics Tagged With: embedded AI

About David Such

David Such is an embedded systems engineer and Director of Reefwing Software, based in Sydney, Australia. He develops IoT devices, robotics platforms, and drone flight control systems, with a focus on deploying intelligence on resource-constrained hardware.

David has over 30 years of industry experience spanning embedded development, systems engineering, and senior leadership. He has held executive roles including Managing Director of Serco Australia and senior management positions at Honeywell and Tyco. He was an Honorary Associate at the University of Sydney, where he mentored aspiring entrepreneurs through the Business Industry Mentoring Program and the Lean Startup course in the Faculty of Business.

His open-source sensor fusion and flight controller libraries, published under the Reefwing Software organisation on GitHub, are used by embedded developers and robotics hobbyists worldwide. He writes extensively on embedded AI, sensor systems, and edge computing across several publications on Medium, where his technical articles have built a substantial following among hardware engineers working at the edge.

David holds a BE in Electrical Engineering, a BSc in Computing Science and Physics, a BAppSc, and an MBA in Strategy.

Embedded AI: A Practical Guide to Building Intelligence on Microcontrollers is his first book with No Starch Press.

Reader Interactions

Leave a Reply Cancel reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.

Primary Sidebar

Embedded AI News

Recent Posts

  • Are LLMs Creating Languages to Speak with Each Other? November 9, 2026
  • Why AI Learns Consensus Over Truth November 6, 2026
  • Why AI Models Keep Inventing Religions October 22, 2026
  • Embedded FM Podcast October 19, 2026
  • The Future of AI Interfaces October 16, 2026

Featured Posts

Are LLMs Creating Languages to Speak with Each Other?

November 9, 2026 By David Such Leave a Comment

Is AI developing a secret, unreadable language beyond human comprehension? In this episode, we unpack the origins, popular myths, and technical realities surrounding “neuralese”—the continuous vector communications and latent reasoning spaces used by neural networks. Starting with its humble academic origin in a 2017 paper on inter-agent vector exchange, we examine how the concept evolved […]

Archives

  • November 2026
  • October 2026
  • September 2026
  • August 2026
  • July 2026
  • June 2026
  • May 2026
  • April 2026
  • March 2026
  • April 2024
  • March 2024
  • February 2024
  • January 2024
  • November 2023
  • October 2023
  • July 2023
  • June 2023
  • May 2023
  • April 2023
  • March 2023
  • November 2022
  • September 2022
  • August 2022
  • July 2022
  • April 2022
  • March 2022
  • February 2022
  • January 2022
  • October 2021
  • August 2021
  • July 2021
  • June 2021
  • February 2021
  • November 2020
  • October 2020
  • May 2020
  • April 2020
  • January 2020
  • November 2019
  • October 2019
  • September 2019
  • July 2019
  • June 2019
  • May 2019
  • March 2019
  • January 2019
  • December 2018
  • October 2018
  • September 2018
  • August 2018
  • July 2018
  • June 2018
  • May 2018
  • March 2018
  • January 2018
  • December 2017
  • November 2017
  • October 2017
  • September 2017
  • August 2017
  • July 2017
  • June 2017
  • April 2017
  • February 2017
  • January 2017
  • December 2016
  • November 2016
  • October 2016
  • September 2016
  • July 2016
  • May 2014
  • April 2014

Search

Copyright © 2026 · Executive Pro on Genesis Framework · WordPress · Log in