Skip to content
Ramirez Global Institute · A research program of RamirezX

The fact base for a world preparing for advanced AI.

The Ramirez Global Institute publishes free, independent research on what increasingly capable artificial intelligence can do, what it changes for businesses and communities, and what it takes to adopt it safely.

Our standard

No client or sponsor commissions, funds or approves a finding. Every study is free to read.

AI-assisted, human-reviewed

Each study names the AI systems that helped produce it.

Latest research

Institute Briefs answer one question in two to four pages.

The first is in preparation.

Institute Brief · In preparation

What the 2026 evidence says about the path to advanced AI

A plain-language reading of the two largest public assessments of AI progress, the International AI Safety Report 2026 and the Stanford AI Index, written for owners and operators who need to plan around them. What has been measured, what is still uncertain, and what it means for an organization that is not an AI lab.

Study
RGI-2026-001
Theme
Capability and trajectory
Sources
Public and free

What the Brief will cover

  • Where measured capability stands today
  • What “superintelligence” means in research, and as a name
  • Where the evidence disagrees
  • Five questions every operator should be asking

Signal

Gold medal

Frontier AI systems now reach gold-medal level on International Mathematical Olympiad questions, while still failing some simple tasks.

International AI Safety Report 2026

Signal

12

Companies that published or updated AI safety frameworks in 2025. Most risk management remains voluntary.

International AI Safety Report 2026

Signal

100+

Experts, nominated by more than 30 countries and international organizations, behind the year’s main public assessment of AI risk.

International AI Safety Report 2026

Research themes

Six standing lanes.

Each holds many studies over many years.

Theme 01

Capability and trajectory

What can frontier AI actually do, how fast is that changing, and what does the evidence say about the path toward superintelligence?

First study: the quarterly Frontier Tracker.

Theme 02

The AI-native firm

How does a small or mid-sized operating company reorganize its work around AI agents, and what changes in cost, speed and staffing?

First study: field notes from working companies, published with consent.

Theme 03

Main Street economics

Where does AI raise wages and productivity in regional economies, and where does it displace work?

First study: county-level AI exposure in California’s Central Valley and Florida.

Theme 04

Safety, trust and governance

What does responsible adoption of more capable AI look like for an organization that is not an AI lab?

First study: a plain-language readiness checklist, mapped to the NIST AI Risk Management Framework.

Theme 05

Civilization-scale coordination

If intelligence becomes abundant, how do institutions, communities and markets coordinate differently?

First study: an essay series, clearly labeled as argument.

Theme 06

Science and discovery

How is AI changing the speed and cost of scientific discovery, in which fields, and what would it take to accelerate it further?

First study: a map of published AI-assisted discoveries by field, with measured time or cost savings where the papers report them.

The Frontier Tracker

A quarterly digest of measured AI capability.

The first edition is in preparation; this is what it will track.

MeasureWhat it tells youPublic sourceCadence
Benchmark frontierBest published scores on reasoning, math and coding tests, and how fast they moveStanford AI Index; Epoch AIQuarterly
Training computeHow much computing power goes into the largest modelsEpoch AI open datasetsQuarterly
Expert risk assessmentWhere the international expert consensus on capability and risk standsInternational AI Safety ReportAnnual
Safety commitmentsWhich developers publish safety frameworks, and what they commit toCompany publicationsQuarterly
Business adoptionShare of U.S. firms using AI to produce goods or servicesU.S. Census Bureau Business Trends and Outlook SurveyQuarterly

How we work

An independent institute is believed only if its rules are public. These are ours.

The charter

Independent

No client, sponsor or partner commissions, funds or approves a finding. RamirezX funds the Institute.

Free

Every published study is free to read.

Fact-based

Every claim is sourced and dated. Estimates are labeled as estimates.

Human-signed

AI drafts and checks; a named person approves every word published.

Disclosed

Each study states which AI systems helped produce it and how.

Corrected in public

Errors are fixed with a dated correction note.

Five gates before anything is published

  1. Gate 1

    Question

    A real, answerable question inside one of the six themes.

  2. Gate 2

    Evidence

    Every source public, legitimately obtained and logged.

  3. Gate 3

    Challenge

    A dedicated critic argues against the draft; every objection is answered.

  4. Gate 4

    Verification

    Every cited source opened and every number re-run.

  5. Gate 5

    Release

    Approved in writing by the Editor-in-Chief.

On the word “superintelligence”: U.S. federal agencies now use “Super Intelligence” as a name for AI. In our research the word means a system that outperforms the best humans across nearly all intellectual work, a level no system has reached yet. We study the path toward it and say plainly which meaning we use.

People

A small human team sets direction and signs off.

A standing team of AI research agents does the scanning, analysis, drafting and checking.

Founding Chair and Editor-in-Chief

Eric Ramirez

Founder, RamirezX

Institute Council

Seats open

Advisers in AI, economics and operations who review the research agenda each quarter. Inquiries welcome.

Fellows

By invitation

Researchers and practitioners who lead a study under their own name.

Newsletter

The Scale

The Institute newsletter: each new finding and every quarterly Frontier Tracker, sent when it is published. No sponsors. Unsubscribe any time.

We use your address only to send The Scale. Unsubscribe any time. Privacy.

The Ramirez Global Institute is a research program of RamirezX LLC. Its publications are research, not investment, legal or tax advice.