Key problems with current systems

Recently there have been platforms that aggregate and analyze crypto-related information often reward users with points for high-quality content shared on X through leaderboard systems. While these platforms aim to incentivize valuable contributions, users express dissatisfaction with their leaderboard updates and reward systems. These platforms aim to solve crypto information fragmentation and reward valuable contributions, but their leaderboards and point systems spark mixed reactions. Supporters appreciate AI-driven transparency and merit-based rewards, but detractors see flaws in systems that favor established players, encourage low-quality content, and prioritize hype over substance. Recent updates to curb noise and farming (e.g., reputation thresholds) have alienated smaller users, amplifying discontent. Sentiment on X reflects frustration with technical issues, algorithmic biases, and shifts toward competitive, less accessible models. However, some users value the potential to revolutionize Web3 engagement. Below are the key reasons for this sentiment, based off reviews on X. Scream AI takes a radical approach and will be addressing most of these problems moving to make an equitable yap to earn structure where every user is rewarded.

Problems:

1. Unfairness towards smaller accounts in Leaderboard Requirements

• New Reputation Thresholds: Recent leaderboard updates introduce stricter criteria, such as minimum thresholds for points, influential "smart" followers, and crypto community engagement. Smaller accounts and new users feel disadvantaged, as these requirements favor established accounts with large followings, reducing opportunities for newcomers.

• Exclusion of Smaller Contributors: Users with limited followers or influence struggle to meet these criteria, leading to frustration. Critics argue the system prioritizes "big players," making it less inclusive for grassroots participants.

• Popularity Contest Perception: Critics argue leaderboards resemble a "popularity contest" rather than a meritocracy, as accounts with more influential followers rank higher. This social-driven aspect clashes with these platforms’ information-focused branding, leading to skepticism about fairness.

2. Algorithm Gaming and Exploitation

• Airdrop Farmers and Bots: Some users claim leaderboard algorithms are exploited by airdrop farmers or accounts managed by teams. For example, top-ranked accounts may post high-quality content at an unsustainable frequency (e.g., every 5 minutes), suggesting team operations rather than individual effort, undermining the merit-based ethos.

• Misinformation by Top Accounts: Highly ranked accounts sometimes spread misinformation but maintain top spots due to engagement metrics, leading to distrust in the leaderboard’s ability to surface credible voices.

3. Keyword Stuffing and Content Quality Issues

• Keyword-Driven Content: Many platforms use AI models that require posts to include specific crypto-related keywords to earn points. This leads to keyword-stuffed, low-quality, or spammy content, drowning out genuine discussions and frustrating users who value authenticity.

• Penalization of Original Content: Users report that original, relevant content may not earn points if it lacks specific keywords, while repetitive or buzzword-heavy posts rank higher, creating a perception that algorithms prioritize quantity or phrasing over quality.

4. Lack of Transparency in Scoring and Penalties

• Opaque Scoring Systems: Users often don’t understand why their posts fail to earn points, as AI-driven evaluations lack clear guidelines. For instance, content deemed "not original enough" may not score, but users receive little feedback on how to improve.Algorithms detect spam or manipulation and penalize accounts but without feedback too.

5. Overemphasis on Pre-TGE Projects

• Focus on Hype Over Substance: Platforms often emphasize pre-Token Generation Event (TGE) projects, fueling speculative hype. Projects with heavy marketing budgets dominate leaderboards, overshadowing fundamentally strong but less hyped initiatives.

• Volatile Rankings: Rolling time windows for leaderboards make rankings volatile, frustrating users who lose relative positions due to others’ contributions rather than their own performance.

6. Language Bias

• Spam and Engagement Farming: Point-to-earn models incentivize spam or over-engagement, cluttering X with low-value posts. This annoys users who prefer organic discussions and see points as a gamified distraction.

• Language and Regional Bias Concerns: Although AI models often support multiple languages, some users suspect English content or jargon-heavy posts score higher, potentially marginalizing non-English or less technical contributors.

7. Multimodal analysis

• No video and image content analysis: Some users do high quality posts that are either in video or image format which are time consuming. However they are rewarded similarily to a Yap that is in text format.