[GUEST ACCESS MODE: Data is scrambled or limited to provide examples. Make requests using your API key to unlock full data. Check https://lunarcrush.ai/auth for authentication information.] #  @connordavis_ai Connor Davis Connor Davis posts on X about ai, llm, $googl, open ai the most. They currently have XXXXX followers and XXX posts still getting attention that total XXXXXX engagements in the last XX hours. ### Engagements: XXXXXX [#](/creator/twitter::1955077938879299584/interactions)  - X Week XXXXXXX +133% - X Month XXXXXXX +12% ### Mentions: XX [#](/creator/twitter::1955077938879299584/posts_active)  - X Week XX -XX% - X Month XX +32% ### Followers: XXXXX [#](/creator/twitter::1955077938879299584/followers)  - X Week XXXXX +9.20% - X Month XXXXX +28% ### CreatorRank: XXXXXXX [#](/creator/twitter::1955077938879299584/influencer_rank)  ### Social Influence **Social category influence** [technology brands](/list/technology-brands) [stocks](/list/stocks) [social networks](/list/social-networks) **Social topic influence** [ai](/topic/ai) #4412, [llm](/topic/llm) #6, [$googl](/topic/$googl), [open ai](/topic/open-ai) #408, [llms](/topic/llms) #2, [breakdown](/topic/breakdown), [hidden](/topic/hidden), [space and](/topic/space-and), [specialized](/topic/specialized), [instead of](/topic/instead-of) **Top accounts mentioned or mentioned by** [@tianjinaigc](/creator/undefined) [@teleglobals](/creator/undefined) [@naftalihirschl](/creator/undefined) [@brandgrowthos](/creator/undefined) [@devon__kelley](/creator/undefined) [@littlebrainz](/creator/undefined) [@kashifch00](/creator/undefined) [@h3roai](/creator/undefined) [@pirouneb](/creator/undefined) [@flowersbyirene4](/creator/undefined) [@rookepoole](/creator/undefined) [@blue_rose_magic](/creator/undefined) [@practiklyperfct](/creator/undefined) [@lunimstudio](/creator/undefined) [@rancosine](/creator/undefined) [@thesealerai](/creator/undefined) [@aeris_echoes](/creator/undefined) [@trunkfuture](/creator/undefined) [@liwenxuanx56278](/creator/undefined) **Top assets mentioned** [Alphabet Inc Class A (GOOGL)](/topic/$googl) ### Top Social Posts Top posts by engagements in the last XX hours "The most important AI paper of 2025 might have just dropped. NVIDIA lays out a framework for Small Language Model agents that could outcompete LLMs. Heres the full breakdown (and why it matters):" [X Link](https://x.com/connordavis_ai/status/1963908330235281908) 2025-09-05T10:13Z 5729 followers, 449.5K engagements "Nobodys ready for what this Stanford paper reveals about multi-agent AI. "Latent Collaboration in Multi-Agent Systems" shows that agents dont need messages protocols or explicit teamwork instructions. They start coordinating inside their own hidden representations a full collaboration layer that exists only in the latent space. And the behaviors are insane: Agents silently hand off tasks based on whos better Roles appear out of nowhere leader executor supporter Policies encode signals that never show up in actions Teams adapt to new environments without retraining Collaboration stays stable" [X Link](https://x.com/connordavis_ai/status/1996165430126522561) 2025-12-03T10:31Z 5726 followers, 146.2K engagements "Holy shit this paper might be the most important shift in how we use LLMs this entire year. Large Causal Models from Large Language Models. It shows you can grow full causal models directly out of an LLM not approximations not vibes actual causal graphs counterfactuals interventions and constraint-checked structures. And the way they do it is wild: Instead of training a specialized causal model they interrogate the LLM like a scientist: extract a candidate causal graph from text ask the model to check conditional independencies detect contradictions revise the structure test counterfactuals" [X Link](https://x.com/connordavis_ai/status/1998337230637789512) 2025-12-09T10:21Z 5729 followers, 43.5K engagements "-I finally get how to build an AI agent that doesnt feel dumb. Most agents fail for the same reason: they think but they cant operate. They dont remember anything they cant use the right tools and they panic the moment a task goes off-script. After breaking a dozen prototypes heres the checklist I now use. Whenever an agent feels dumb one of these four is missing. X. Memory that actually survives the workflow Not summaries. Not context windows. Actual structured memory: what happened what the agent decided what failed what it should avoid next time If the agent forgets its own steps it" [X Link](https://x.com/connordavis_ai/status/1996546852779671685) 2025-12-04T11:47Z 5725 followers, 21.3K engagements "Multi-agent systems = specialization. One agent plans. One executes. One reflects. One handles memory. One fixes errors. Think of it as an AI team not an AI individual. This is how agents scale from toy demos to real workflows" [X Link](https://x.com/connordavis_ai/status/1998712859145257238) 2025-12-10T11:14Z 5731 followers, XXX engagements "Memory is the difference between ChatGPT with arms and a real autonomous worker. The paper breaks memory into: short-term (context window) long-term (RAG SQL workflow memory) experience memory (successful + failed runs) This is how agents stop repeating the same stupid mistakes" [X Link](https://x.com/connordavis_ai/status/1998712874139811917) 2025-12-10T11:14Z 5726 followers, XXX engagements "I didnt expect a 12B model to shake up video AI this hard but Vidi2 is doing exactly that. Most video models can describe whats happening. A handful can summarize scenes. Almost none can track events through time with real precision. Vidi2 can. ByteDance built a model that can watch a 30-minute video and answer a single sentence like: Show me the exact moment the person in the hoodie stands up. And instead of giving timestamps or guesses it finds the right person follows them across the entire timeline and draws bounding boxes around the action as it unfolds. Full spatio-temporal tubes." [X Link](https://x.com/connordavis_ai/status/1995809544367390764) 2025-12-02T10:57Z 5588 followers, 2235 engagements "When you combine all X systems Perception Reasoning Memory Action you finally get something useful: An AI that can see think remember and act inside a real computer environment. This paper is the closest thing to an Agent Architecture Handbook we have right now" [X Link](https://x.com/connordavis_ai/status/1998712903944536553) 2025-12-10T11:14Z 5729 followers, XXX engagements "Multi-agent AI is a $50B lie. XX% of "multi-agent" systems are just single agents with fancy marketing. I just read the paper that exposes what real multi-agent intelligence actually looks like. Most people think multi-agent AI is just "multiple ChatGPTs in a room. That's like saying a surgical team is just "multiple people with knives." The real story is way deeper. Task allocation is completely broken. Current systems are basically throwing darts at a board. Give the math problem to whoever's free. Ask the creative agent to debug code. It's chaos disguised as intelligence. Real multi-agent" [X Link](https://x.com/connordavis_ai/status/1970511164686934028) 2025-09-23T15:30Z 5699 followers, 123.8K engagements "I finally understand why XX% of AI agents fail. After reading Anthropic's new context engineering guide everything clicked. Prompt engineering is outdated. Context engineering is the game. Here's the framework that actually works:" [X Link](https://x.com/connordavis_ai/status/1975177447542083934) 2025-10-06T12:32Z 5721 followers, 41.6K engagements "Holy shit Google just rewired how AI agents talk to the world 🤯 Theyve built a real-time bidirectional streaming architecture meaning agents no longer wait for your input to finish before responding. They see hear and act while youre still speaking. They can be interrupted mid-task. They can collaborate with other agents in real time. Its all powered by the new Agent Development Kit (ADK) built around async I/O stateful sessions streaming-native tools and live callbacks. This isnt request-response anymore. Its conversation without turns. The moment AI starts feeling alive" [X Link](https://x.com/connordavis_ai/status/1985297464388435988) 2025-11-03T10:46Z 5721 followers, 62.6K engagements "THE AI SAFETY REPORT NO ONE WANTS TO GO VIRAL JUST DROPPED and its bad thats why lol. The Future of Life Institute just audited the X companies building the most powerful AI models on the planet Anthropic OpenAI Google DeepMind xAI DeepSeek Alibaba Cloud Z. ai Meta and the results feel like someone pulled the fire alarm. On the surface everything looks fine. But when you zoom in for even X seconds the whole sector is sprinting faster than its safety brakes can handle. Heres what punched me in the gut: 1/ even the leaders are slipping Anthropic still ranks highest overall but the report says" [X Link](https://x.com/connordavis_ai/status/1998013800054988861) 2025-12-08T12:56Z 5727 followers, 15.1K engagements "I didnt truly understand how to build strong AI agents until one paper snapped everything into place. Not a tutorial. Not a YouTube demo. A single arXiv paper: Fundamentals of Building Autonomous LLM Agents. It finally made sense why most agents feel like chatbots with extra steps and why real autonomous systems need an actual architecture. Heres the backbone the pros use the part nobody explains clearly 👇 X. Perception: what the agent actually sees It isnt just text. Real agents mix: - screenshots - DOM trees - accessibility APIs - Set-of-Mark style visual encodings Thats how an agent stops" [X Link](https://x.com/connordavis_ai/status/1998712780355256548) 2025-12-10T11:13Z 5731 followers, 50.5K engagements "Finally execution. This is where every demo fails. Tool calling is easy. Executing a 15-step workflow with UI inputs state changes and error recovery Thats the real challenge. The papers action system shows exactly how agents convert plans into concrete actions in software environments" [X Link](https://x.com/connordavis_ai/status/1998712888740196848) 2025-12-10T11:14Z 5731 followers, XXX engagements "🔥 I cant believe this exists someone finally wrote the secret playbook every AI agent startup has been faking. A research team just dropped A Practical Guide for Designing Developing and Deploying Production-Grade Agentic AI Workflows and its basically the internal handbook people *think* OpenAI and Anthropic use. Not hype. Not diagrams with cute arrows. A real engineering blueprint for agents that dont collapse the moment you leave the demo sandbox. Heres what makes it insane 👇 X / They start by exposing why most agents implode Its never the model. Its the system around it:" [X Link](https://x.com/connordavis_ai/status/1999078517389291935) 2025-12-11T11:27Z 5729 followers, 12.3K engagements
[GUEST ACCESS MODE: Data is scrambled or limited to provide examples. Make requests using your API key to unlock full data. Check https://lunarcrush.ai/auth for authentication information.]
@connordavis_ai Connor Davis
Connor Davis posts on X about ai, llm, $googl, open ai the most. They currently have XXXXX followers and XXX posts still getting attention that total XXXXXX engagements in the last XX hours.
Engagements: XXXXXX #
- X Week XXXXXXX +133%
- X Month XXXXXXX +12%
Mentions: XX #
- X Week XX -XX%
- X Month XX +32%
Followers: XXXXX #
- X Week XXXXX +9.20%
- X Month XXXXX +28%
CreatorRank: XXXXXXX #
Social Influence
Social category influence technology brands stocks social networks
Social topic influence ai #4412, llm #6, $googl, open ai #408, llms #2, breakdown, hidden, space and, specialized, instead of
Top accounts mentioned or mentioned by @tianjinaigc @teleglobals @naftalihirschl @brandgrowthos @devon__kelley @littlebrainz @kashifch00 @h3roai @pirouneb @flowersbyirene4 @rookepoole @blue_rose_magic @practiklyperfct @lunimstudio @rancosine @thesealerai @aeris_echoes @trunkfuture @liwenxuanx56278
Top assets mentioned Alphabet Inc Class A (GOOGL)
Top Social Posts
Top posts by engagements in the last XX hours
"The most important AI paper of 2025 might have just dropped. NVIDIA lays out a framework for Small Language Model agents that could outcompete LLMs. Heres the full breakdown (and why it matters):"
X Link 2025-09-05T10:13Z 5729 followers, 449.5K engagements
"Nobodys ready for what this Stanford paper reveals about multi-agent AI. "Latent Collaboration in Multi-Agent Systems" shows that agents dont need messages protocols or explicit teamwork instructions. They start coordinating inside their own hidden representations a full collaboration layer that exists only in the latent space. And the behaviors are insane: Agents silently hand off tasks based on whos better Roles appear out of nowhere leader executor supporter Policies encode signals that never show up in actions Teams adapt to new environments without retraining Collaboration stays stable"
X Link 2025-12-03T10:31Z 5726 followers, 146.2K engagements
"Holy shit this paper might be the most important shift in how we use LLMs this entire year. Large Causal Models from Large Language Models. It shows you can grow full causal models directly out of an LLM not approximations not vibes actual causal graphs counterfactuals interventions and constraint-checked structures. And the way they do it is wild: Instead of training a specialized causal model they interrogate the LLM like a scientist: extract a candidate causal graph from text ask the model to check conditional independencies detect contradictions revise the structure test counterfactuals"
X Link 2025-12-09T10:21Z 5729 followers, 43.5K engagements
"-I finally get how to build an AI agent that doesnt feel dumb. Most agents fail for the same reason: they think but they cant operate. They dont remember anything they cant use the right tools and they panic the moment a task goes off-script. After breaking a dozen prototypes heres the checklist I now use. Whenever an agent feels dumb one of these four is missing. X. Memory that actually survives the workflow Not summaries. Not context windows. Actual structured memory: what happened what the agent decided what failed what it should avoid next time If the agent forgets its own steps it"
X Link 2025-12-04T11:47Z 5725 followers, 21.3K engagements
"Multi-agent systems = specialization. One agent plans. One executes. One reflects. One handles memory. One fixes errors. Think of it as an AI team not an AI individual. This is how agents scale from toy demos to real workflows"
X Link 2025-12-10T11:14Z 5731 followers, XXX engagements
"Memory is the difference between ChatGPT with arms and a real autonomous worker. The paper breaks memory into: short-term (context window) long-term (RAG SQL workflow memory) experience memory (successful + failed runs) This is how agents stop repeating the same stupid mistakes"
X Link 2025-12-10T11:14Z 5726 followers, XXX engagements
"I didnt expect a 12B model to shake up video AI this hard but Vidi2 is doing exactly that. Most video models can describe whats happening. A handful can summarize scenes. Almost none can track events through time with real precision. Vidi2 can. ByteDance built a model that can watch a 30-minute video and answer a single sentence like: Show me the exact moment the person in the hoodie stands up. And instead of giving timestamps or guesses it finds the right person follows them across the entire timeline and draws bounding boxes around the action as it unfolds. Full spatio-temporal tubes."
X Link 2025-12-02T10:57Z 5588 followers, 2235 engagements
"When you combine all X systems Perception Reasoning Memory Action you finally get something useful: An AI that can see think remember and act inside a real computer environment. This paper is the closest thing to an Agent Architecture Handbook we have right now"
X Link 2025-12-10T11:14Z 5729 followers, XXX engagements
"Multi-agent AI is a $50B lie. XX% of "multi-agent" systems are just single agents with fancy marketing. I just read the paper that exposes what real multi-agent intelligence actually looks like. Most people think multi-agent AI is just "multiple ChatGPTs in a room. That's like saying a surgical team is just "multiple people with knives." The real story is way deeper. Task allocation is completely broken. Current systems are basically throwing darts at a board. Give the math problem to whoever's free. Ask the creative agent to debug code. It's chaos disguised as intelligence. Real multi-agent"
X Link 2025-09-23T15:30Z 5699 followers, 123.8K engagements
"I finally understand why XX% of AI agents fail. After reading Anthropic's new context engineering guide everything clicked. Prompt engineering is outdated. Context engineering is the game. Here's the framework that actually works:"
X Link 2025-10-06T12:32Z 5721 followers, 41.6K engagements
"Holy shit Google just rewired how AI agents talk to the world 🤯 Theyve built a real-time bidirectional streaming architecture meaning agents no longer wait for your input to finish before responding. They see hear and act while youre still speaking. They can be interrupted mid-task. They can collaborate with other agents in real time. Its all powered by the new Agent Development Kit (ADK) built around async I/O stateful sessions streaming-native tools and live callbacks. This isnt request-response anymore. Its conversation without turns. The moment AI starts feeling alive"
X Link 2025-11-03T10:46Z 5721 followers, 62.6K engagements
"THE AI SAFETY REPORT NO ONE WANTS TO GO VIRAL JUST DROPPED and its bad thats why lol. The Future of Life Institute just audited the X companies building the most powerful AI models on the planet Anthropic OpenAI Google DeepMind xAI DeepSeek Alibaba Cloud Z. ai Meta and the results feel like someone pulled the fire alarm. On the surface everything looks fine. But when you zoom in for even X seconds the whole sector is sprinting faster than its safety brakes can handle. Heres what punched me in the gut: 1/ even the leaders are slipping Anthropic still ranks highest overall but the report says"
X Link 2025-12-08T12:56Z 5727 followers, 15.1K engagements
"I didnt truly understand how to build strong AI agents until one paper snapped everything into place. Not a tutorial. Not a YouTube demo. A single arXiv paper: Fundamentals of Building Autonomous LLM Agents. It finally made sense why most agents feel like chatbots with extra steps and why real autonomous systems need an actual architecture. Heres the backbone the pros use the part nobody explains clearly 👇 X. Perception: what the agent actually sees It isnt just text. Real agents mix: - screenshots - DOM trees - accessibility APIs - Set-of-Mark style visual encodings Thats how an agent stops"
X Link 2025-12-10T11:13Z 5731 followers, 50.5K engagements
"Finally execution. This is where every demo fails. Tool calling is easy. Executing a 15-step workflow with UI inputs state changes and error recovery Thats the real challenge. The papers action system shows exactly how agents convert plans into concrete actions in software environments"
X Link 2025-12-10T11:14Z 5731 followers, XXX engagements
"🔥 I cant believe this exists someone finally wrote the secret playbook every AI agent startup has been faking. A research team just dropped A Practical Guide for Designing Developing and Deploying Production-Grade Agentic AI Workflows and its basically the internal handbook people think OpenAI and Anthropic use. Not hype. Not diagrams with cute arrows. A real engineering blueprint for agents that dont collapse the moment you leave the demo sandbox. Heres what makes it insane 👇 X / They start by exposing why most agents implode Its never the model. Its the system around it:"
X Link 2025-12-11T11:27Z 5729 followers, 12.3K engagements