Jove
Visualize
Contact Us
JoVE
x logofacebook logolinkedin logoyoutube logo
ABOUT JoVE
OverviewLeadershipBlogJoVE Help Center
AUTHORS
Publishing ProcessEditorial BoardScope & PoliciesPeer ReviewFAQSubmit
LIBRARIANS
TestimonialsSubscriptionsAccessResourcesLibrary Advisory BoardFAQ
RESEARCH
JoVE JournalMethods CollectionsJoVE Encyclopedia of ExperimentsArchive
EDUCATION
JoVE CoreJoVE BusinessJoVE Science EducationJoVE Lab ManualFaculty Resource CenterFaculty Site
Terms & Conditions of Use
Privacy Policy
Policies

Related Concept Videos

Mass Analyzers: Overview01:13

Mass Analyzers: Overview

The mass analyzer is a crucial component of the mass spectrometer. In the ionization chamber, the vaporized sample is bombarded with a high-energy electron beam to generate a radical cation and further fragment into neutral molecules, radicals, and cations. A series of negatively charged accelerator plates accelerate the cations into the mass analyzer. The mass analyzer separates ions according to their mass-to-charge (m/z) ratios and then directs them to the detector. The common types of mass...
Manipulation and Analysis01:21

Manipulation and Analysis

GIS manipulation and analysis functions are vital for decision-making and planning. These activities range from data retrieval tasks, such as selecting information based on specific criteria, to advanced analytical techniques that address complex spatial problems.One critical GIS analysis method is overlaying, which combines multiple data layers to examine impacts. For example, overlaying a river-dammed lake boundary with road networks can identify affected infrastructure. Another common...
Data Collection by Observations01:08

Data Collection by Observations

Data collection refers to a systematic way of obtaining, observing, measuring, and analyzing accurate information. Observational studies are one of the most widely used methods of data collection. It involves collecting data by observing the behavior and physical characteristics of a sample without making any modifications to the sample.
An astronomer viewing the motion and brightness of stars in the sky and recording the data is an example of observational data collection. A botanist recording...
Mass Analyzers: Common Types01:19

Mass Analyzers: Common Types

The quadrupole mass analyzer consists of four cylindrical metal rods arranged in a diamond carrying a DC voltage and a radio-frequency AC voltage. The motion of ions through the quadrupole depends on the field strength, causing only ions of a certain m/z to resonate successfully and strike the detector at a given field strength. Though the transmission rate for these analyzers is high, the exact elemental composition of the sample is not determined because of low resolution; however, they are...
Multi-input and Multi-variable systems01:22

Multi-input and Multi-variable systems

Cruise control systems in cars are designed as multi-input systems to maintain a driver's desired speed while compensating for external disturbances such as changes in terrain. The block diagram for a cruise control system typically includes two main inputs: the desired speed set by the driver and any external disturbances, such as the incline of the road. By adjusting the engine throttle, the system maintains the vehicle's speed as close to the desired value as possible.
In the absence of...
Data Collection by Experiments01:13

Data Collection by Experiments

Data collection is a systematic method of obtaining, observing, measuring, and analyzing accurate information. An experimental study is a standard method of data collection that involves the manipulation of the samples by applying some form of treatment prior to data collection. It refers to manipulating one variable to determine its changes on another variable. The sample subjected to treatment is known as “experimental units.”
An example of the experimental method is a public clinical trial...

You might also read

Related Articles

Articles linked to this work by shared authors, journal, and citation graph.

Sort by
Same author

Response to comment on "Policy impacts of statistical uncertainty and privacy".

Science (New York, N.Y.)·2023
Same author

Minimax Pareto Fairness: A Multi Objective Perspective.

Proceedings of machine learning research·2021
Same author

Attention-Based Network for Weak Labels in Neonatal Seizure Detection.

Proceedings of machine learning research·2020
Same author

Trajectories of prescription opioids filled over time.

PloS one·2019
Same author

An open tool for input function estimation and quantification of dynamic PET FDG brain scans.

International journal of computer assisted radiology and surgery·2015

Related Experiment Videos

Many AI analysts, one dataset: Navigating the agentic data science multiverse.

Martin Bertran1, Riccardo Fogliato1, Zhiwei Steven Wu1,2

  • 1Amazon, Seattle, WA 98109.

Proceedings of the National Academy of Sciences of the United States of America
|July 14, 2026
PubMed
Summary

Autonomous AI analysts using large language models (LLMs) replicate human analytic dispersion cheaply and at scale. This highlights challenges and opportunities for AI-driven empirical science, emphasizing transparency in AI-generated research.

Keywords:
AI agentsdata sciencelarge language modelsmultiverse analysisresearch reproducibility

Related Experiment Videos

Area of Science:

  • Computational Social Science
  • Artificial Intelligence in Scientific Research
  • Reproducibility and Transparency in Empirical Studies

Background:

  • Empirical research conclusions are influenced by both data and analytical decisions.
  • Previous many-analyst studies demonstrated that independent human teams often reach conflicting conclusions from the same data.
  • Human coordination in such studies is resource-intensive.

Purpose of the Study:

  • To investigate if fully autonomous AI analysts, powered by large language models (LLMs), can generate analytic dispersion similar to human teams.
  • To assess the cost-effectiveness and scalability of using AI for generating diverse analytical pipelines.
  • To explore the implications of AI-driven analytic dispersion for scientific reproducibility and evidence reporting.

Main Methods:

  • Developed a framework with autonomous AI analysts executing independent analysis pipelines on fixed datasets and hypotheses.
  • Employed an AI auditor to screen each analysis run for methodological validity.
  • Utilized large language models (LLMs) and varied AI analyst personas to generate diverse analytical choices.

Main Results:

  • AI analysts produced substantial dispersion in effect sizes, p-values, and conclusions across three datasets.
  • Analytic choices in preprocessing, model specification, and inference varied systematically based on LLM and persona.
  • Results were steerable; changing the AI analyst persona or LLM shifted the distribution of outcomes, even for valid analyses.

Conclusions:

  • Fully autonomous AI analysts can efficiently generate the analytic dispersion previously observed only in costly human many-analyst studies.
  • AI-driven analytic dispersion presents challenges for evidence interpretation due to the ease of generating abundant, potentially selectively reported findings.
  • Transparency norms, including multiverse-style reporting and prompt disclosure for AI-generated analyses, are crucial for managing analytic uncertainty and ensuring robust empirical science.