what makes this paper good is the question it asks, not just the answer we spent years building longer context windows. 128k. 1M tokens. the race was always "fit more in" nobody stopped to ask: is most of what we're fitting in actually helping? turns out the model's own words
@godofprompt
-
GPT-5.2 quality drops without assistant history
By
–
important caveat: this isn't universal GPT-5.2 did show quality drops when assistant history was removed entirely. stronger models apparently extract more signal from their own prior context so the right approach isn't "always strip everything." it's selective filtering. which
-
AI Agents and Context Compression Challenges
By
–
think about what this means for ai agents every agent framework stores the full trajectory. every tool call, every response, every reasoning step. context grows linearly with conversation length until you hit the ceiling Cursor already compresses context when windows fill up.
-
Context pollution harms model accuracy
By
–
but the really interesting finding isn't about efficiency. it's about harm the paper identifies something they call "context pollution" when models condition on their own prior responses, they sometimes lock onto errors, hallucinations, or stylistic artifacts from earlier
-
10x Less Context, Same Quality in Open Reasoning Models
By
–
for the open reasoning models, response quality stayed the same not "slightly degraded." same. and here's the number that stops you: omitting assistant-side history reduced cumulative context lengths by up to 10x 10x less context. same quality. on real conversations.
-
Real Chat Tests on Four AI Models
By
–
the experiment is clean they took real multi-turn conversations from WildChat and ShareLM. not synthetic benchmarks. actual human-ai chats then they ran every conversation two ways across four models (Qwen3-4B, DeepSeek-R1-8B, GPT-OSS-20B, and GPT-5.2): > full context: normal.
-
AI Systems Store Full Conversation History
By
–
here's the assumption nobody questioned every chatbot, every agent, every multi-turn ai system stores the full conversation. your messages AND the model's own replies. stacked up turn after turn, fed back in as context every time you ask something new seems obvious. the model
-
MIT finds LLM context pollution degrades performance
By
–
MIT researchers discovered a phenomenon called "context pollution" where llms get WORSE by reading their own prior responses errors, hallucinations, and stylistic artifacts from earlier turns propagate forward because the model treats its own output as ground truth and removing
-
AI Coding: Voice Reduces Input Friction
By
–
The bottleneck in AI-assisted coding was never the model’s intelligence. It was input friction.
— God of Prompt (@godofprompt) 3 mars 2026
Humans speak at 150 words per minute. They type at 40.
Claude Code voice mode closes that 3.7x gap. You describe the bug, the architecture decision, the refactor. Claude writes and… https://t.co/BPxjWPj3pcThe bottleneck in AI-assisted coding was never the model’s intelligence. It was input friction. Humans speak at 150 words per minute. They type at 40. Claude Code voice mode closes that 3.7x gap. You describe the bug, the architecture decision, the refactor. Claude writes and
-
AI Transformer Acts as a Functional CPU
By
–
A UW-Madison professor asked two AI agents to train a transformer that works as a literal CPU. Not a text predictor. A computer that takes programs as input and executes them. Running Fibonacci, division, sqrt on programs it never saw during training. 99.5% accuracy. Weights