Futuros · Hello Tomorrow Global Summit 2026 · Amsterdam

Futuros.

AI will increasingly decide how well we are governed.
Humans will decide whether it gets it right.

25 countries · 10 parameters · 500+ decision-makers · 1 ten-year horizon
01 · The shift already underway

Governance is becoming a data problem — and AI is becoming its operating layer.

Decisions that used to run on intuition and annual reports now run on real-time intelligence. The jurisdiction, the institution, the movement that pairs that intelligence with sound judgment will simply outperform the one that doesn't.

The question is no longer whether AI enters governance. It's on whose terms, with whose data, and under whose accountability.

02 · The thesis

The unlock isn't AI. It's AI accompanied by humans.

AI ALONE

Fails the legitimacy test

No mandate, thin local context, no values of its own. It can hallucinate, drift, or be quietly captured.

HUMANS ALONE

Can't see at scale

No team reads 660M lives and millions of indicators in time to act — let alone simulate the alternatives.

TOGETHER

The critical unlock

AI senses, aggregates & simulates at scale. Humans hold judgment, legitimacy & accountability. Futuros is built around that loop.

03 · Why Latin America, why now

A region the frontier was not designed for.

<1%
of global frontier-AI compute sits in LATAM — for ~8% of world GDP and 660M people.
CB Insights / Stanford HAI 2024
52%
of workers are in the informal economy — invisible to models trained on Bay-Area / Brussels data.
ILO 2023
#1
in stranded clean energy: Atacama, Itaipú, Patagonia — among the cheapest clean electrons on Earth.
IRENA / IEA 2024

The frameworks weren't written for us. Futuros treats that as a market inefficiency and a moral failure at once — and a reason to build our own surface.

04 · The deployment gap

Deep tech here doesn't fail in the lab. It fails to land.

The region's first deep-tech report (LADP × Hello Tomorrow, 2025) maps the stall across 18 countries on three axes. Futuros answers each one — as infrastructure.

FINANCING

Bankable by default

One-page dossiers — capex, opex, IRR, risk — that an investment committee or a cabinet can fund.

REGULATION

Named, mapped, reachable

10-parameter evidence plus the named ministers and regulators, per country, who must say yes.

GEOPOLITICS

Sovereign by construction

Non-aligned across US / China / EU / regional providers; data never leaves its jurisdiction.

The scaling gap is a deployment gap. Futuros industrializes deployment.

05 · The platform

One platform. Four functions. One loop.

01
Aggregate

The data layer

Fuse every public & private source into one regional warehouse + lake — traceable to origin.

02
Govern

Better decisions

Turn that data into accountable choices across the 10 parameters of a good society.

03
Simulate

Bankable pilots

Test solutions in simulation and ship one-page, financeable pilots before spending a peso.

04
Awaken

Mass literacy

Drive awareness across public, private & civil society — the three sectors that move outcomes.

Every function is a human-in-the-loop AI system. AI does the scale; people keep the judgment.

Function 01 · Aggregate

First, see clearly.

A single data layer for the whole region — a warehouse and a lake that fuses fragmented public open data with private and proprietary feeds into one source of truth. Not another dashboard that cites. A platform that traces.

Function 01 · what's inside

Every number, two clicks from its source.

live demo /atlas
  • 250 cells — 25 countries × 10 parameters, each with 8–12 indicators + a governance timeline
  • World Bank + 8 frontier indicators no multilateral tracks: compute, datacenters, AI talent, chips, model deployments, renewable-host, cables, cloud regions
  • Subnational ADM1 depth — resolution below the national average, where policy actually lands
  • 1,687 cited sources — every figure carries a verifiable citation_id; contradictions logged, not hidden
  • Weekly pulse — regional brief + 90-day forecast cards, each with a falsification criterion
  • Public + private, fused safely — sensitive feeds contribute via encryption & federation; private data becomes usable without being exposed
Function 02 · Govern

Then, decide better.

The 10 parameters are a measurable definition of a good society. Governance is the work of moving those numbers — for real people, accountably. AI drafts and flags; humans amend, vote, and decide. The mandate stays human.

Function 02 · the governance suite

AI proposes. Humans decide. Everything is audited.

live demo /governance
PROPUESTAS

AI proposals, human amendments

Draft policy across all 10 parameters, then amendment threads + anonymous voting. The decision stays with people.

ORACLE

Legislation Oracle

Live bills & decrees, AI-flagged risk / conflict / opportunity — each flag shows the exact model run behind it.

TRANSPARENCIA

We audit our own AI

Every model call, prompt & output diff logged and queryable. The standard we ask of others, applied to us.

Function 03 · Simulate

Test before you spend.

Don't gamble scarce public money on a hunch at full scale. Simulate the solution, see the outcome across the parameters — then ship a one-page pilot a cabinet or an investment committee can fund.

Function 03 · data → bankable pilot

A simulator that outputs a one-page dossier.

live demo /pilotos
  • Configurator — scope, geography, financing mix, tech, timeline → live capex / opex / IRR / risk / impact
  • Bankable dossier — financing sankey, named decision-maker map, risk register, on one page
  • 30+ pilot archetypes — seeded across the 10 parameters and priority countries
  • 33 case studies — LATAM wins & failures + global benchmarks (Estonia, Pix, Plan Ceibal, Atacama), each with a translation guide
Function 04 · Awaken

Then, bring everyone in.

Better governance fails if only technocrats can read it. Outcomes move when the three sectors that hold the levers share one traceable picture of reality. The platform is the campaign.

Function 04 · three sectors, one literacy

Same facts. Three audiences. Shared truth.

PUBLIC

Cabinets, legislators, regulators

Sovereignty-first intelligence to govern and legislate on traceable evidence.

$
PRIVATE

Capital, firms, founders

Bankable pilots and a market map for where to deploy in the region.

CIVIL SOCIETY

NGOs, press, academia, citizens

Hold power accountable with the same sourced facts the powerful use.

Powered by 500+ named decision-makers with public-record timelines, and persona lenses that re-order the same data for each actor.

live demo /personas
The measure · what 'a good society' means

Ten parameters. One horizon: 2035.

Health
coverage · wellbeing
Education
learning · human capital
$
Economy
jobs · productivity
Security
safety · crime
§
Justice & Rule of Law
courts · impunity
Climate & Environment
water · biodiversity
Infrastructure
energy · connectivity
Innovation & Tech Sovereignty
R&D · compute
Social Cohesion
equity · inclusion
Democratic Governance
voice · transparency
06 · The trust layer · AI × cryptography

Don't trust the AI. Verify it.

For AI to sit near governance, its data and its outputs must be checkable by math, not reputation. Cryptography is the substrate beneath all four functions — the difference between an audited-AI promise and an audited-AI proof.

VERIFIABLE AI

Proof, not promises

Zero-knowledge ML can prove an output came from the stated model, prompt & data — without exposing private inputs or weights.

PRIVATE, USABLE

Data that never leaves

Secure multi-party computation & federation let sensitive public and private data feed the lake without extraction. Sovereignty by construction.

TAMPER-EVIDENT

Logs anyone can check

Append-only, publicly verifiable trails for every model call, citation & vote — plus anti-sybil, anonymous-yet-verifiable participation.

Where institutional trust can't be assumed, verifiable-by-default isn't a luxury — it's how AI earns a seat in governance, and how 25 sovereign instances interoperate with no single point of capture.

07 · Why you can trust it

Built to be checked, not believed.

TRACEABLE

Sourced to origin

1,687 cited sources, 2-click provenance, and a public log of every correction.

AUDITED AI

We audit ourselves

Every model call & prompt logged. We won't ask for trust we don't extend.

NON-ALIGNED

Sovereignty-first

Balanced across US / China / EU / regional providers — not a foreign template renamed.

CALIBRATED

Failures stay public

A foresight scoreboard where our missed forecasts remain published, dated.

08 · Live today, not a roadmap

Open it on your phone right now.

25
countries live
250
country × parameter cells
500+
named decision-makers
1,687
cited sources
30+
pilot archetypes
33
case studies
ADM1
subnational depth
ES·EN·PT
trilingual throughout
09 · How you plug in

This needs a coalition. You're it.

IF YOU GOVERN

Pilot a parameter

Run a pilot in your jurisdiction; fork and self-host a country instance of the platform.

IF YOU FUND

Back a dossier

Take a bankable pilot to your committee; co-finance with Futuros as platform of record.

IF YOU CONVENE

Spread the literacy

Bring traceable facts to your community; hold power accountable with sourced evidence.

One next step for everyone in this room: a 30-minute live immersion in the platform.

Plugs into the regional rails: eLAC2026 (ECLAC) · ILIA (CENIA × ECLAC) · LatamGPT (CENIA + CAF, 60+ institutions) · LADP × Hello Tomorrow
Futuros · the region, with intelligence

Latin America doesn't need another report.
It needs a platform that acts.

futuros.xyz · cush@odisea.xyz · next: 30-min live immersion
Futuros · v1.5
1 / 19
→ next · ← back · s notes · ? help

Speaker notes

Keyboard

/ Space
Next slide
Previous slide
s
Toggle speaker notes
f
Fullscreen
Home / End
First / last slide
?
This help · Esc closes
⌘/Ctrl+P
Export to PDF (1 slide/page)