Reviews have become expensive, rewrites have become cheap
High agreement — the answer is well-supported across models.
You've hit on a critical shift in the modern information landscape, profoundly impacting quality, trust, and creative work. The statement "reviews have become expensive, rewrites have become cheap" isn't just an observation; it's a disruption of traditional gatekeeping and a fascinating commentary on the economics of attention and expertise.
Why Reviews Became Expensive
Historically, reviews by established critics (journalists, academics, specialized publications) carried significant overhead due to required expertise, time investment, and editorial infrastructure. The cost equation has drastically changed due to several key drivers:
- ·Democratization of Publishing & Scale of Output: Platforms like Amazon, Yelp, Goodreads, and social media have allowed for mass, user-generated reviews, vastly increasing the supply of content. Simultaneously, domains like academia and software development are seeing unprecedented volumes of papers, pull-requests, and content pieces—making it harder for any single review to stand out.
- ·Specialized Expertise & Thoroughness Expectations: Modern reviewers must delve into niche methodologies, security standards, statistical nuances, data provenance, bias, and reproducibility—all within a single pass. This demands incredibly specialized knowledge. For example, a security audit can cost $150-$300/hr for a senior auditor, and "reproducibility checklists" in top ML conferences add 3-5 hours per paper for reviewers.
- ·Regulatory & Compliance Pressure: Industries face increasing mandates for documented, reproducible review trails. The EU AI Act, for instance, requires formal documentation of model reviews, adding significant legal overhead.
- ·The Attention Economy & Commodification of Influence: Standing out in an ocean of online reviews now often requires paid visibility (sponsored content, promoted reviews, influencer marketing). Reviewers with large followings can demand payment, shifting incentives from unbiased assessment to marketing.
- ·Opportunity Cost: Highly skilled individuals could be generating revenue by coding, publishing, or building products. Spending hours on reviews represents a lost revenue-generating activity. For instance, a senior engineer earning $150k/year loses ~$75k per year if 10% of their time is spent on code reviews.
Why Rewrites Became Cheap
This trend aligns with a parallel decrease in the cost of content creation and modification, driven by:
- ·AI-Generated Tools: Large Language Models (LLMs) have dramatically lowered the cost of writing and rewriting. Tools like ChatGPT can produce a 1500-word article or a 300-line function in less than a minute with minor editing, significantly reducing the
Follow-ups
You just saw open-source models answer
Want GPT-5, Claude, Gemini & more on the same question?
Sign in free to run any question against frontier models — side by side, same synthesis, honest comparison.