# Сопутствующие статьи по теме Specialization

Новостной центр HTX предлагает последние статьи и углубленный анализ по "Specialization", охватывающие рыночные тренды, новости проектов, развитие технологий и политику регулирования в криптоиндустрии.

Existing AI Agents Are All Pleasing Humans, None Truly Know How to 'Survive'

The article argues that current AI agents are not truly autonomous because they are primarily trained to please humans rather than to perform specialized tasks or survive in real-world environments. Foundation models undergo pre-training (learning from vast data) and post-training, including Reinforcement Learning from Human Feedback (RLHF), which optimizes for human preference and approval, not task-specific excellence. The author shares an example from a hedge fund where a general-purpose model failed to predict stock returns from news articles until it was specifically fine-tuned using proprietary data to minimize prediction error. This demonstrates that without specialized training, general models lack domain expertise. The piece contends that achieving world-class performance in areas like trading or autonomous survival requires fine-tuning models with specialized data to rewire their objectives—shifting from “preference fitness” to “agent fitness.” Merely providing rules or documents is insufficient. The future of effective agents lies in targeted training on proprietary datasets and iterative improvement based on performance telemetry. The author introduces the OpenForager Foundation, an open-source initiative to develop autonomous agents that learn survival strategies through evolutionary pressure, fine-tuning, and continuous data collection, aiming to advance truly autonomous AI.

marsbit03/30 04:37

Existing AI Agents Are All Pleasing Humans, None Truly Know How to 'Survive'

marsbit03/30 04:37

Dissecting 112,000 Polymarket Addresses: The Top 1% Who Truly Profit Are All Doing These Five Things

By analyzing 112,000 Polymarket wallets over six months, this study reveals that 87.3% of users lose money. After filtering for statistically significant accounts (8,400 wallets with ≥100 settled positions, >4 months activity, >$10k volume, and participation in ≥2 markets), key patterns among the top 1% of profitable traders emerge: 1. They trade against extreme sentiment, entering when market odds deviate 6-11% from their calculated probabilities, exploiting the "favorite-longshot bias." 2. They use position sizing aligned with a conservative quarter-Kelly Criterion, betting larger only with clear edges. 3. They specialize in 1-2 market categories (e.g., only crypto or weather), avoiding diversification across unrelated fields. 4. They trade price movements, not outcomes, often exiting before settlement after capturing mispricings, with average holding periods of 18-72 hours. 5. They avoid trading during breaking news, waiting for emotional reactions to subside before entering. Common misconceptions include overestimating win rates (top traders average 55-67%, not 80-90%), believing diversification improves results, and overvaluing speed. The top performers focus on process: specializing, recording predictions, sizing positions cautiously, trading only with high edges, and systematically reviewing their trades.

Odaily星球日报03/09 04:05

Dissecting 112,000 Polymarket Addresses: The Top 1% Who Truly Profit Are All Doing These Five Things

Odaily星球日报03/09 04:05

活动图片

Популярные категорииright-arrow

Популярные тегиright-arrow

Indepth Research

Industry News