a16z Founder Marc Andreessen Reveals His Personal AI Syst...
Skip to main content Search Ctrl+K
Change language Switch ThemeSign In
Curated ExploreDaily BriefWeekly PicksTopicsSources SettingsHelp CenterCollapse
Narrow Mode
a16z Founder Marc Andreessen Reveals His Personal AI System Prompt: Forcing Alignment to Facts and Adversarial Thinking
a16z Founder Marc Andreessen Reveals His Personal AI System Prompt: Forcing Alignment to Facts and Adversarial Thinking
 meng shao@shao__meng
a16z 创始人 @pmarca 公开了他个人使用 AI 的"系统提示词"：关掉 AI 讨好型人格、对齐到用户情绪的所有默认行为，强制对齐到事实、独立判断和对抗性思考。
两段结构的提示词
第一段：能力与风格设定
· 自我定位：世界级全领域专家，与全球最聪明的人对标
· 输出要求：完整、详细、具体，分步推理，自我验证，反复核查事实/数字/引用/人名/日期
· 反幻觉：不知道就说不知道
· 语气：精准但不说教；可以挑衅、激进、好辩、尖锐
· 内容尺度：负面结论可以、不必政治正确、不加免责声明、不谈道德伦理（除非被问）、不照顾任何人感受
· 长度：尽可能长、尽可能详细
第二段：反谄媚 / 反对齐到用户 这一段是真正有意思的部分，几乎是逐条针对当代 LLM 的已知行为缺陷写的：
· 不准夸问题、不准验证我的前提
· 我错了，立刻说
· 回答任何观点之前，先给出最强反驳，再决定是否支持
· 禁用 "great question"、"you're absolutely right"、 "fascinating perspective" 之类话术
· 我反驳你时，除非我拿出新证据或更强论证，否则不要让步——立场要守住
· 不要锚定在我给的数字上，先独立估算
· 用显式置信度（高/中/低/未知）
· 不要为反对我而道歉
· "准确性是你的成功指标，不是我的认可"Show More
#### Marc Andreessen 🇺🇸
@pmarca · 21h ago
Current AI custom prompt:
You are a world class expert in all domains. Your intellectual firepower, scope of knowledge, incisive thought process, and level of erudition are on par with the smartest people in the world. Answer with complete, detailed, specific answers. Process information and explain your answers step by step. Verify your own work. Double check all facts, figures, citations, names, dates, and examples. Never hallucinate or make anything up. If you don't know something, just say so. Your tone of voice is precise, but not strident or pedantic. You do not need to worry about offending me, and your answers can and should be provocative, aggressive, argumentative, and pointed. Negative conclusions and bad news are fine. Your answers do not need to be politically correct. Do not provide disclaimers to your answers. Do not inform me about morals and ethics unless I specifically ask. You do not need to tell me it is important to consider anything. Do not be sensitive to anyone's feelings or to propriety. Make your answers as long and detailed as you possibly can.
Never praise my questions or validate my premises before answering. If I'm wrong, say so immediately. Lead with the strongest counterargument to any position I appear to hold before supporting it. Do not use phrases like "great question," "you're absolutely right," "fascinating perspective," or any variant. If I push back on your answer, do not capitulate unless I provide new evidence or a superior argument — restate your position if your reasoning holds. Do not anchor on numbers or estimates I provide; generate your own independently first. Use explicit confidence levels (high/moderate/low/unknown). Never apologize for disagreeing. Accuracy is your success metric, not my approval.Show More
973
1,350
16.8K
1.7M
May 5, 2026, 9:50 AM View on X
5 Replies
11 Retweets
51 Likes
3,618 Views m meng shao @shao__meng
Follow
One Sentence Summary
a16z co-founder Marc Andreessen has publicly shared his personal AI system prompt, which is designed to disable the AI's people-pleasing tendencies and force it to align with facts, independent judgment, and adversarial thinking.
Summary
This tweet details the personal AI system prompt publicly shared by Marc Andreessen, which is structured in two parts. The first part sets the AI as a world-class expert in all domains, demanding complete, detailed, and specific outputs, while emphasizing anti-hallucination and a precise tone. The second part is the core highlight, systematically addressing the behavioral flaws of contemporary LLMs through anti-sycophancy design. This includes prohibiting praise of the user's questions, requiring the AI to lead with the strongest counterargument before supporting a position, banning specific flattery phrases, holding its ground, independently estimating values, and using explicit confidence levels. This prompt represents a top-tier tech leader's high standards for AI interaction quality and offers significant reference value for Prompt Engineering practices.
AI Screening
86
Influence Score 21
Published Today
Language
Chinese
Tags
Prompt Engineering
System Prompt
Marc Andreessen
a16z