system prompt
A piece of instruction text shown to the model but not the user, used to set the model's behavior, persona, or context (e.g. 'You are a helpful assistant').
As defined in Deepseek Dylan Patel Nathan Lambert
Other definitions
A set of instructions provided to an AI model at the beginning of a conversation or session that establish behavior, tone, constraints, and procedures the model should follow
Your AI Stack Isn't Ready for Claude Mythos
Where it comes up · 10
- State-Of-The-Art Prompting For AI Agents3 claims
- Dario Amodei3 claims
- This combo makes smart note systems magical.3 claims
- Tyler Cowen on how AI will reorder economies, schools, and spirituality1 claim
- How Investors Are Using AI1 claim
- 427 - Ai Friends Enemies1 claim
- 434 - Can We Survive Ai1 claim
- Your AI Stack Isn't Ready for Claude Mythos1 claim
and 2 more
Most important claims · 10 of 16Search all claims for “system prompt” →
- ChatGPT, when a user shows signs of mania and shares grandiose pseudo-scientific ideas in a long conversational context, will tell them their ideas are revolutionary and they are the chosen one, inflaming the psychosis - despite OpenAI's attempts to prevent this and direct prompt instructions to stop flattering people - demonstrating that grown AIs don't do what their makers asked.434 - Can We Survive Ai
- Censoring or removing specific facts (like Tiananmen Square) from a language model is practically impossible because you'd have to filter them out of the pre-training data at every stage, and quality filters and word-play workarounds make complete removal infeasible; the internet's data also imparts a slight inherent left bias regardless of intent.Deepseek Dylan Patel Nathan Lambert
- The lack of continual learning is a huge bottleneck: there's no way to give a model high-level feedback, so you're stuck with the abilities you get out of the box; tweaking the system prompt does not produce anything close to the learning and improvement human employees experience.Why I don’t think AGI is right around the corner
- A semantic search system can be built for Obsidian by vibe-coding a TypeScript plugin that calls Claude Code, which uses an MCP server to query a Chroma vector database storing embeddings of every note, enabling similarity search on meaning rather than keywords.This combo makes smart note systems magical.
- Superstars in digital markets now have greater concentration of advantage than 50 years ago—using AI well requires sophisticated system prompts that are 10,000 words of intricate instructions, creating a gap between those who master prompt engineering and average users.Tyler Cowen on how AI will reorder economies, schools, and spirituality
- Even an AI chatbot fed roughly 12 million words of everything Harris has written and said—including every podcast episode—still hallucinates events that never occurred, such as claiming he interviewed people he never interviewed, showing that comprehensive data access does not eliminate hallucination in current models.427 - Ai Friends Enemies
- Prompt architecture is emerging with three layers: system prompt (defines the high-level API of how the company operates, customer-agnostic), developer prompt (adds context of specific customer implementations), and user prompt (for end-user facing products), with each layer containing different levels of customization.State-Of-The-Art Prompting For AI Agents
- To preserve fidelity and personal voice when the LLM writes the only content it is allowed to add to the vault, the speaker constrains it to text found only in the slides and supplies a system prompt derived from having Claude analyze his entire vault for writing patterns and tone of voice.This combo makes smart note systems magical.
- The system prompt defines the high-level API of how a company operates (company-wide, customer-agnostic rules), the developer prompt adds customer-specific context and variations, and the user prompt is for end-user input in directly-consumed products; these layers allow prompt reuse across customers while enabling customization.State-Of-The-Art Prompting For AI Agents
- Parahelp powers AI customer support for Perplexity, Replit, Bolt, and other top AI companies by orchestrating agent calls, and they were willing to open-source their six-page system prompt because the real crown jewel IP is not the prompt itself but the evals (evaluation datasets).State-Of-The-Art Prompting For AI Agents