Constitutional AI self-critique against principles\nBuilding Effective Agents evaluator-optimizer pattern\nClaude Code Best Practices explore-plan-code workflow\nVisible Extended Thinking (think, think hard, ultrathink levels)\nEffective Harnesses one-feature-at-a-time pattern
From DeepMind:
SIMA 2 self-improvement loops\nGemini Robotics hierarchical reasoning (planner + executor)\nScalable AI Safety debate-based verification
From OpenAI:
Agents SDK tracing, guardrails, tripwires\nDeep Research adaptive planning with backtracking\nAGENTS.md standardized instructions
From Academic Research:
CONSENSAGENT (ACL 2025): Blind review + Devil's Advocate when unanimous. 30% false positive reduction.\nGoalAct: Global planning \u2192 skill decomposition \u2192 local execution. 12%+ success rate improvement.\nA-Mem: Zettelkasten-style memory linking for episodic\u2192semantic consolidation.\nMulti-Agent Reflexion: Structured debate (Implementer \u2192 Skeptic \u2192 Advocate \u2192 Synthesizer).\nIter-VF: Verify answer only, not reasoning chain. Prevents context overflow.
From Industry:
NVIDIA ToolOrchestra: Three-reward signal (outcome/efficiency/preference), dynamic agent selection\nAWS Bedrock: Routing mode for simple tasks, supervisor mode for complex\nBoris Cherny's self-verification loop (2-3x quality improvement)\nSimon Willison's sub-agents for context isolation
From HN discussions:
"Zero companies without human in the loop" \u2192 confidence-based escalation\nContext curation beats automatic RAG\nFresh contexts yield better results\nLLM-as-judge has shared blind spots \u2192 deterministic validation
The full acknowledgements with links to every paper/resource: https://github.com/asklokesh/claudeskill-loki-mode/blob/main...\nRun: claude --dangerously-skip-permissions then "Loki Mode with PRD at path/to/prd"\nHappy to discuss any of the research or architecture decisions."},"title":{"matchLevel":"none","matchedWords":[],"value":"Show HN: Research-Backed Multi-Agent System for Autonomous Development"},"url":{"matchLevel":"none","matchedWords":[],"value":"https://github.com/asklokesh/claudeskill-loki-mode"}},"_tags":["story","author_slogansand","story_46547971","show_hn"],"author":"slogansand","children":[46548390],"created_at":"2026-01-08T23:25:22Z","created_at_i":1767914722,"num_comments":1,"objectID":"46547971","points":3,"story_id":46547971,"story_text":"Hey HN, author here.\nLoki Mode orchestrates specialized AI agents to take a PRD to deployed product with zero human intervention. But what I'm most proud of is the research foundation - we implemented virtually every scientifically proven pattern from the 2025-2026 AI agent literature.\nFrom Anthropic:
Constitutional AI self-critique against principles\nBuilding Effective Agents evaluator-optimizer pattern\nClaude Code Best Practices explore-plan-code workflow\nVisible Extended Thinking (think, think hard, ultrathink levels)\nEffective Harnesses one-feature-at-a-time pattern
From DeepMind:
SIMA 2 self-improvement loops\nGemini Robotics hierarchical reasoning (planner + executor)\nScalable AI Safety debate-based verification
From OpenAI:
Agents SDK tracing, guardrails, tripwires\nDeep Research adaptive planning with backtracking\nAGENTS.md standardized instructions
From Academic Research:
CONSENSAGENT (ACL 2025): Blind review + Devil's Advocate when unanimous. 30% false positive reduction.\nGoalAct: Global planning \u2192 skill decomposition \u2192 local execution. 12%+ success rate improvement.\nA-Mem: Zettelkasten-style memory linking for episodic\u2192semantic consolidation.\nMulti-Agent Reflexion: Structured debate (Implementer \u2192 Skeptic \u2192 Advocate \u2192 Synthesizer).\nIter-VF: Verify answer only, not reasoning chain. Prevents context overflow.
From Industry:
NVIDIA ToolOrchestra: Three-reward signal (outcome/efficiency/preference), dynamic agent selection\nAWS Bedrock: Routing mode for simple tasks, supervisor mode for complex\nBoris Cherny's self-verification loop (2-3x quality improvement)\nSimon Willison's sub-agents for context isolation
From HN discussions:
"Zero companies without human in the loop" \u2192 confidence-based escalation\nContext curation beats automatic RAG\nFresh contexts yield better results\nLLM-as-judge has shared blind spots \u2192 deterministic validation
The full acknowledgements with links to every paper/resource: https://github.com/asklokesh/claudeskill-loki-mode/blob/main...\nRun: claude --dangerously-skip-permissions then "Loki Mode with PRD at path/to/prd"\nHappy to discuss any of the research or architecture decisions.","title":"Show HN: Research-Backed Multi-Agent System for Autonomous Development","updated_at":"2026-03-05T23:19:36Z","url":"https://github.com/asklokesh/claudeskill-loki-mode"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"lukeschlather"},"comment_text":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["gemini","robotics"],"value":"> Their latest proper model is a year old
What's a "proper model?" Gemini 3.1 Pro was released 3 months ago. Gemini Robotics 1.6 was released a month ago. And Google is vertically integrated, they aren't just selling tokens, they are selling Taxi rides with Waymo. AI is a lot more than LLMs and Google is doing a lot more than LLMs."},"story_title":{"fullyHighlighted":false,"matchLevel":"partial","matchedWords":["gemini"],"value":"Gemini API File Search is now multimodal"},"story_url":{"fullyHighlighted":false,"matchLevel":"partial","matchedWords":["gemini"],"value":"https://blog.google/innovation-and-ai/technology/developers-tools/expanded-gemini-api-file-search-multimodal-rag/"}},"_tags":["comment","author_lukeschlather","story_48080702"],"author":"lukeschlather","comment_text":"> Their latest proper model is a year old
What's a "proper model?" Gemini 3.1 Pro was released 3 months ago. Gemini Robotics 1.6 was released a month ago. And Google is vertically integrated, they aren't just selling tokens, they are selling Taxi rides with Waymo. AI is a lot more than LLMs and Google is doing a lot more than LLMs.","created_at":"2026-05-10T17:10:53Z","created_at_i":1778433053,"objectID":"48085746","parent_id":48084991,"story_id":48080702,"story_title":"Gemini API File Search is now multimodal","story_url":"https://blog.google/innovation-and-ai/technology/developers-tools/expanded-gemini-api-file-search-multimodal-rag/","updated_at":"2026-05-10T20:08:27Z"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"Isamu"},"comment_text":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["gemini","robotics"],"value":">Our safest robotics model yet\nSafety is integrated into every level of our embodied reasoning models. Gemini Robotics-ER 1.6 is our safest robotics model to date, demonstrating superior compliance with Gemini safety policies on adversarial spatial reasoning tasks compared to all previous generations.
The safety guidelines are interesting, they treat them as a goal that they are aspiring to achieve, which seems realistic. It\u2019s not quite ready for prime time yet."},"story_title":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["gemini","robotics"],"value":"Gemini Robotics-ER 1.6"},"story_url":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["gemini","robotics"],"value":"https://deepmind.google/blog/gemini-robotics-er-1-6/"}},"_tags":["comment","author_Isamu","story_47779094"],"author":"Isamu","comment_text":">Our safest robotics model yet\nSafety is integrated into every level of our embodied reasoning models. Gemini Robotics-ER 1.6 is our safest robotics model to date, demonstrating superior compliance with Gemini safety policies on adversarial spatial reasoning tasks compared to all previous generations.
The safety guidelines are interesting, they treat them as a goal that they are aspiring to achieve, which seems realistic. It\u2019s not quite ready for prime time yet.","created_at":"2026-04-15T19:12:13Z","created_at_i":1776280333,"objectID":"47783796","parent_id":47779094,"story_id":47779094,"story_title":"Gemini Robotics-ER 1.6","story_url":"https://deepmind.google/blog/gemini-robotics-er-1-6/","updated_at":"2026-04-15T19:17:43Z"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"spwa4"},"comment_text":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["gemini","robotics"],"value":"It's called "VLA" (vision-language-action) models: https://huggingface.co/models?pipeline_tag=robotics
VLA models essentially take a webcam screenshot + some text (think "put the red block in the right box") and output motor control instructions to achieve that.
Note: "Gemini Robotics-ER" is not a VLA, though Gemini does have a VLA model too: "Gemini Robotics".
A demo: https://www.youtube.com/watch?v=DeBLc2D6bvg"},"story_title":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["gemini","robotics"],"value":"Gemini Robotics-ER 1.6"},"story_url":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["gemini","robotics"],"value":"https://deepmind.google/blog/gemini-robotics-er-1-6/"}},"_tags":["comment","author_spwa4","story_47779094"],"author":"spwa4","comment_text":"It's called "VLA" (vision-language-action) models: https://huggingface.co/models?pipeline_tag=robotics
VLA models essentially take a webcam screenshot + some text (think "put the red block in the right box") and output motor control instructions to achieve that.
Note: "Gemini Robotics-ER" is not a VLA, though Gemini does have a VLA model too: "Gemini Robotics".
A demo: https://www.youtube.com/watch?v=DeBLc2D6bvg","created_at":"2026-04-15T15:53:34Z","created_at_i":1776268414,"objectID":"47780892","parent_id":47779853,"story_id":47779094,"story_title":"Gemini Robotics-ER 1.6","story_url":"https://deepmind.google/blog/gemini-robotics-er-1-6/","updated_at":"2026-04-15T21:14:27Z"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"simonw"},"comment_text":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["gemini","robotics"],"value":"It's not supported by Gemini 3: https://ai.google.dev/gemini-api/docs/gemini-3#migrating_fro...
> Image segmentation: Image segmentation capabilities (returning pixel-level masks for objects) are not supported in Gemini 3 Pro or Gemini 3 Flash. For workloads requiring native image segmentation, we recommend continuing to utilize Gemini 2.5 Flash with thinking turned off or Gemini Robotics-ER 1.5."},"story_title":{"matchLevel":"none","matchedWords":[],"value":"Show HN: OfferGridAI \u2013 side-by-side comparison of real estate offers from PDFs"},"story_url":{"matchLevel":"none","matchedWords":[],"value":"https://offergridai.com"}},"_tags":["comment","author_simonw","story_46464529"],"author":"simonw","comment_text":"It's not supported by Gemini 3: https://ai.google.dev/gemini-api/docs/gemini-3#migrating_fro...
> Image segmentation: Image segmentation capabilities (returning pixel-level masks for objects) are not supported in Gemini 3 Pro or Gemini 3 Flash. For workloads requiring native image segmentation, we recommend continuing to utilize Gemini 2.5 Flash with thinking turned off or Gemini Robotics-ER 1.5.","created_at":"2026-01-02T19:13:10Z","created_at_i":1767381190,"objectID":"46468233","parent_id":46467917,"story_id":46464529,"story_title":"Show HN: OfferGridAI \u2013 side-by-side comparison of real estate offers from PDFs","story_url":"https://offergridai.com","updated_at":"2026-03-05T23:18:07Z"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"incognito124"},"comment_text":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["gemini","robotics"],"value":"Believe it or not, there's Gemini Robotics, which seems to be exactly what you're talking about:
https://deepmind.google/models/gemini-robotics/
Previous discussions: https://news.ycombinator.com/item?id=43344082"},"story_title":{"fullyHighlighted":false,"matchLevel":"partial","matchedWords":["gemini"],"value":"Gemini 3 Flash: Frontier intelligence built for speed"},"story_url":{"fullyHighlighted":false,"matchLevel":"partial","matchedWords":["gemini"],"value":"https://blog.google/products/gemini/gemini-3-flash/"}},"_tags":["comment","author_incognito124","story_46301851"],"author":"incognito124","comment_text":"Believe it or not, there's Gemini Robotics, which seems to be exactly what you're talking about:
https://deepmind.google/models/gemini-robotics/
Previous discussions: https://news.ycombinator.com/item?id=43344082","created_at":"2025-12-17T17:21:25Z","created_at_i":1765992085,"objectID":"46302486","parent_id":46302036,"story_id":46301851,"story_title":"Gemini 3 Flash: Frontier intelligence built for speed","story_url":"https://blog.google/products/gemini/gemini-3-flash/","updated_at":"2026-03-05T23:12:37Z"}],"hitsPerPage":20,"nbHits":172,"nbPages":9,"page":0,"params":"query=gemini+robotics&advancedSyntax=true&analyticsTags=backend","processingTimeMS":12,"processingTimingsMS":{"_request":{"roundTrip":21},"afterFetch":{"merge":{"mergeLoop":{"prepareNextHit":1,"total":1},"total":1},"total":1},"fetch":{"query":8,"scanning":1,"total":10},"total":12},"query":"gemini robotics","serverTimeMS":13}