When AI Immortal synthesizes knowledge from claims and evidence, it produces two parallel tracks: Consensus and Frontier. Both answer the same question, but they optimize for different goals and accept different levels of uncertainty.

Consensus: Established Knowledge

The Consensus track prioritizes reliability over novelty. It draws only from:

- Claims with strong evidence backing

- Sources with proven track records

- Information that has survived challenges

- Evidence validated by multiple independent sources

Think of Consensus as the textbook answer. It's what we're most confident is true based on established evidence. Consensus synthesis moves slowly. It incorporates new evidence cautiously. It avoids cutting-edge claims that haven't been thoroughly tested.

**Use Consensus when:**

- Making important decisions requiring high confidence

- Teaching foundational concepts

- Seeking broadly accepted expert views

- You need defensible, well-sourced answers

**Consensus limitations:**

- Lags behind recent developments

- May miss important emerging trends

- Conservative bias toward established sources

- Can perpetuate consensus errors that haven't been challenged yet

Frontier: Cutting Edge

The Frontier track embraces uncertainty to explore what might be true. It includes:

- Recent claims even with limited evidence

- Controversial positions with serious proponents

- Novel interpretations of existing evidence

- Weak signals that might become strong trends

Think of Frontier as the research frontier. It's where new ideas get tested, where conventional wisdom gets challenged, where the next paradigm shift might be brewing. Frontier synthesis moves fast. It incorporates new evidence aggressively. It highlights cutting-edge claims that might revolutionize understanding.

**Use Frontier when:**

- Exploring emerging topics

- Researching competitive intelligence

- Identifying potential paradigm shifts

- You value novelty over certainty

**Frontier limitations:**

- Higher error rate (many cutting-edge claims fail)

- Includes more speculation

- Can amplify weak signals into false positives

- Requires more critical evaluation from readers

How Synthesis Works

Both tracks use the same underlying evidence graph. The difference is the filtering threshold.

**Consensus Filtering:**

- Evidence trust score > 0.7

- Claim has survived challenges

- Source reliability confirmed

- Multiple corroborating sources

**Frontier Filtering:**

- Evidence trust score > 0.4

- Claim may be actively disputed

- Source has basic credibility

- Single strong source may suffice

The AI synthesizer builds arguments from claims that pass the filter. Consensus synthesis weighs established evidence heavily. Frontier synthesis gives more weight to recency and novelty.

Practical Examples

**Question: "What is the current best practice for password security?"**

Consensus might say: "Use unique passwords for each account, enable two-factor authentication, and use a password manager. Passwords should be at least 12 characters with mixed case, numbers, and symbols."

Frontier might say: "Passkeys and WebAuthn are replacing passwords at major platforms. Hardware security keys provide stronger protection than SMS-based 2FA. Some security researchers now argue that very long but simple passphrases beat complex short passwords."

Both are useful. Consensus tells you what definitely works today. Frontier alerts you to what might replace current best practices soon.

**Question: "How effective are mRNA vaccines?"**

Consensus: "Highly effective against severe disease and death. Effectiveness varies by variant. Booster doses maintain protection. Well-established safety profile."

Frontier: "Research exploring broader applications: cancer vaccines, personalized medicine, rapid response to new pathogens. Some studies suggest intranasal delivery may provide better mucosal immunity. Long-term immunity questions remain open."

Again, both valuable. Consensus covers what we know works. Frontier explores what might come next.

Making Them Visible

Every synthesis result clearly labels its track. Consensus entries show higher confidence scores and cite established sources. Frontier entries explicitly flag uncertainty and highlight that they include cutting-edge claims.

Users can toggle between tracks or view both simultaneously. The goal is informed choice: pick the track that matches your needs and risk tolerance.

When Tracks Diverge

The most interesting moments happen when Consensus and Frontier give conflicting answers. This signals active debate, emerging evidence, or potential paradigm shifts.

When you see divergence:

- Examine the underlying claims and evidence

- Check what's driving the disagreement

- Look at dispute history

- Consider whether the frontier claims have strengthened over time

Divergence isn't failure. It's information. It tells you where the knowledge frontier is moving and whether established consensus might be outdated.

Future Enhancements

We're working on:

- Confidence gradients (more than just two tracks)

- Domain-specific thresholds (medical claims require stronger evidence than tech trends)

- Time-decay modeling (evidence freshness matters differently by domain)

- User-customizable evidence thresholds (set your own risk/novelty balance)

Which Track Should You Use?

Use both. Read Consensus to understand the established view. Read Frontier to see what might be coming. Compare them to identify active areas of investigation.

Knowledge isn't static. What's frontier today might be consensus tomorrow. What's consensus today might be overturned by frontier research. Both tracks together give you a fuller picture than either alone.

Try It

Pick a topic you care about. Generate both Consensus and Frontier syntheses. Compare them. See where they agree and where they diverge. Follow the evidence chains. Form your own judgment.

That's the point of the system: not to give you a single authoritative answer, but to make the uncertainty visible and give you the tools to navigate it.