Labs
MissionCareersUpdates
Models
Terra-1Mantle-1Aether-1
SafetyResearchTeamBlogDocs
TAL Corp

"Imagine a world where the smartest mind on earth works for every person equally."

"We're building Artificial General Intelligence — for real people."

Scroll to explore
02Road Map
Moon
Execution Roadmap

The Path to Universal Intelligence.

LAUNCHOF COURSE I STILL LOVE YOULANDINGSTAGE I 2025 – 2026ALIGNMENTSTAGE II 2026 – 2027CONTEXT ENGINESTAGE III 2027 – 2028AUTONOMOUS RESEARCHSTAGE IV 2028 – 2030UNIVERSAL ACCESSSTAGE V 2030 & BeyondAGI FOR ALLTALCORP
03 · Stage I — Alignment
STAGE I2025 – 2026

Foundation of Aligned Intelligence

Safety first — before scale.

Every leap in intelligence must begin with a leap in responsibility. TAL Corp's founding phase establishes the core research pillars — alignment, interpretability, and agentic control.

Expected Output

Terra-1 ships with real-time interpretability dashboards. Open-source safety toolkit becomes the industry standard for alignment audits.

Key Deliverables
RLHF + RLAIF hybrid alignment framework
Glass-box interpretability toolkit (open-source)
Agentic safety benchmarks & red-team protocols
Terra-1 — production-grade aligned LLM
99.2%Safety Score
12K+Alignment Tests
3Open Tools
Earth
SAFETY PERIMETERALIGNMENT LAYERINTERPRETABILITY SHELL
Stage II — Cognitive Engine · 04
STAGE II2026 – 2027

Intelligence That Understands Context

From pattern recognition to situational awareness.

Models that don't just predict — they understand. Systems that grasp nuance, cultural context, and domain expertise at superhuman depth.

Expected Output

Mantle-1 adapts to who you are. A doctor in Nairobi gets Harvard-level clinical guidance. A Texas business owner gets Fortune-500 strategy.

Key Deliverables
Contextual Reasoning Engine (CRE)
Multi-domain expert system: medicine, law, engineering
Cultural intelligence layer — 50+ frameworks
Mantle-1 — context-aware multi-domain model
4M tknContext Depth
120+Domains
−94%Bias Reduction
Earth
STAGE I — ALIGNMENT LAYERSAFETY PERIMETERCOGNITIVE ENGINESTAGE II
05Stage III — Autonomous Discovery
STAGE III2027 – 2028

Autonomous Research Acceleration

AI that advances science — not just assists it.

Systems that autonomously generate hypotheses, design experiments, and accelerate discovery across medicine, materials science, and clean energy.

Expected Output

Aether-1 co-authors 200+ peer-reviewed papers in year one. Identifies 3 novel drug candidates for rare diseases.

Key Deliverables
Autonomous Discovery Loop (hypothesis → validation)
Science co-pilot — 10M+ papers synthesized
Drug-discovery pipeline — molecules in weeks
Aether-1 — research-grade autonomous reasoning
50×Discovery Speed
10M+Papers Indexed
1,200+Novel Compounds
Earth
STAGE I — ALIGNMENTSTAGE II — COGNITIVEDISCOVERY LOOPSTAGE III
06Stage IV — Universal Access
STAGE IV2028 – 2030

Universal Access. Zero Gatekeepers.

Democratizing intelligence for everyone.

The most transformative technology in history should not be locked behind corporate paywalls. Stage IV is our commitment to universal deployment.

Expected Output

A farmer's $50 phone runs the same models as Silicon Valley. Schools in 140 countries deploy TAL tutoring in local languages.

Key Deliverables
Edge-optimized models — full AI on $50 hardware
Universal Language Protocol — 200+ languages
Open API standard — zero licensing fees
TAL Community Network — 10K+ global partners
200+Languages
$50Min. Hardware
10K+Partners
Earth
STAGE I — ALIGNMENTSTAGE II — COGNITIVESTAGE III — DISCOVERYUNIVERSAL ACCESSSTAGE IV
07 Stage V — Post-Scarcity AGI
STAGE V2030 & Beyond

The Post-Scarcity Intelligence Era

The smartest mind — for every person, equally.

AGI eliminates knowledge inequality — where a child in a remote village has the same cognitive resources as the most elite institution on earth.

Expected Output

AGI becomes a public utility. A child born anywhere inherits the cognitive power of every library and lab ever built. The open charter is permanent.

Key Deliverables
Full AGI deployment for all humanity
Knowledge equality — near-zero global disparity
Self-improving alignment (safety scales with intelligence)
Permanent open charter — AGI can never be privatized
→ 0Knowledge Gap
8B+Humans Served
Open Forever
Earth
STAGE I — ALIGNMENTSTAGE II — COGNITIVESTAGE III — DISCOVERYSTAGE IV — UNIVERSALPOST-SCARCITYSTAGE V — AGI ERA
What Makes Us Different08
What Makes Us Different
Five Pillars — TAL Corp
PILLAR 01Safety asArchitectureNot an afterthought
◈ TOUCH TO OPEN
PILLAR 02Cultural Intelligence, NotJust Context WindowsUnderstanding people
◈ TOUCH TO OPEN
PILLAR 03Autonomous Discovery —End to EndFrom hypothesis to val
◈ TOUCH TO OPEN
PILLAR 04Universal Access. ZeroLicensing Fees.$50 hardware. 200+ lan
◈ TOUCH TO OPEN
PILLAR 05Permanent OpenCharterIrrevocable. Legally b
◈ TOUCH TO OPEN
09 · Founding Principle
SPOKE 1 MIN AGO0xA1F2VOICEDATA0x3C8ERETRIEVING FROM AIR
Founding Principle

The question isn't whether machines will think. The question is whether that intelligence will belong to everyone — or only to the few who build it.

"We're not the only ones building toward AGI. But we may be the only ones who've committed, from day one, to making it permanently open and universally accessible— not as an aspiration, but as an irrevocable charter."

— TAL Corp, 2026

Join the Mission →
Laplace's Demon — Applied to Acoustics