The High Price and Contentious Nature of "Truth”

When journalists, pundits, professionals, and media outlets import leaning opinions into factual reporting, they become propagators of agenda-driven narratives rather than purveyors of objective truth. This compromises their role as watchdogs and truth-tellers, leaving “we the people” vulnerable to manipulation leading to chaotic thinking and faulted decisions.

“THE TRUTH” requires a robust commitment to transparency, fact-checking, and adherence to ethics, ensuring that opinions and facts are clearly delineated and fact-based.

Even lying is susceptible to inflationary pricing

The cost and approach to fact-checking statements made depend on the scale, methodology, and desired accuracy of the truth. With the explosion of real-time online output, tackling misinformation effectively is a massive challenge. The price of fact-checking varies depending on whether it’s human-driven, AI-assisted, or fully automated:

Human-Driven Fact-Checking
Cost per Fact: Fact-checking a single claim takes an average of 2-4 hours, depending on complexity, at $25-$50/hour for skilled fact-checkers.
Per Claim: $50-$200 per claim.
Annual Cost (Moderate Scale): A dedicated fact-checking team verifying ~100 claims/day (36,500 claims/year) might cost $5M-$7M annually.
Scalability: Slow, labor-intensive, and costly for large-scale operations.

AI-Assisted Fact-Checking
Cost per Claim: Initial development of a robust AI system can cost $1M-$10M, with ongoing training and maintenance costs of $500K-$1M annually.
AI can significantly reduce human workload, verifying claims for $0.10-$1 per claim after deployment.
Annual Cost (Moderate Scale): Around $2M-$4M annually for a hybrid AI-human system handling 100,000 claims/year.

Fully Automated Fact-Checking (Latency Algorithm)
Development Cost: $5M-$20M to design and deploy an advanced algorithm capable of parsing, cross-referencing, and flagging content in real time.
Additional data sourcing/licensing fees for access to databases (news archives, research papers, etc.) could add $1M-$3M/year.
Ongoing costs for cloud computing and AI training: $1M-$2M/year.
Annual Cost (Moderate Scale): Around $3M-$7M annually after initial investment.

Fact-Checking Methodologies - Two potential approaches:

Latency Delay to Publishing
Process: Introduce a publishing delay (e.g., 30 minutes to several hours) during which articles and broadcasts are cross-checked by humans or algorithms.
Advantages: Allows for thorough verification. Ensures accountability before misinformation spreads.
Challenges: Costly: Requires significant staffing or highly advanced algorithms.
Resistance from Media: Latency could undermine competitiveness in real-time news cycles.

AI Algorithms to Label Untruths
Process: Use algorithms to scan content in real-time, compare it against verified databases, and flag or label untruths as "unverified," "misleading," or "false." AI tools like machine learning models analyze semantic content for consistency with verified facts.
Advantages: Faster and scalable compared to human review. Labels allow users to make informed decisions without fully censoring content.
Challenges: Requires robust, unbiased training data. Risk of over-reliance on AI, which may mislabel nuanced claims.

Implementation: Ideal Scenario
Combining latency delay and AI labeling is likely the most effective approach:
Stage 1 (Latency Delay): Critical or breaking news items undergo rapid AI pre-checks with human oversight.
Stage 2 (Labeling): For ongoing content (e.g., opinion pieces, editorials), AI labels claims with confidence scores while encouraging readers to verify flagged statements.

The cost of implementing a scalable fact-checking system could range from $2M to $20M annually depending on methodology and scale. AI algorithms for real-time labeling offer a more affordable, scalable solution than human-only systems, but they must be paired with transparency, ethical oversight, and continuous improvement to ensure credibility.

For media outlets, the question isn't just about affordability – though the costs for fact-checking are extraordinary – but whether audiences will accept latency or trust the accuracy of AI labeling systems, given the contentious nature of "truth" in our highly polarized country.

No solution is bulletproof, perfect, or pure though a decentralized verification network of entities like PolitiFact and Snopes that cross-reference immutable record-keeping of factual claims maintained on blockchain could prove to be a starting point for consideration. In an inflationary spiral who pays… we all pay the price for trust and truth as well as for lies and deceits.

It’s a business: First Draft News, Poynter’s, Media Bias, Facta.news, Alt News, Truth or Fiction, Lead Stories, BOOM Live, Chequeado, Africa Check, Full Fact, AFP Fact Check, and FactCheck.org are current service providers.