Wiredframe Radar logo

Category intelligence

Social Media Briefing — February 9, 2026

361 current items analyzed and ranked.

Executive synthesis

Social Media Summary

GPT-5.3 Codex dominated discussions as Greg Brockman shared official walkthroughs and teased transformative computing capabilities, drawing massive engagement. xAI announced new image models on Grok Imagine API.

Yann LeCun clarified his Meta departure with 462K views, noting scientists aren't motivated by money. Andriy Burkov provided sharp technical counter-narrative on OpenClaw hype ("2% code, 98% hype").

Key Themes

Organizational Theory for Agentic AI · 1GPT-5.3 Codex Launch & Capabilities · 5Open Source Model Ecosystem · 4Google Search Growth vs Death Narrative · 4Product Releases · 1AI Routing & UX Design Flaws · 3AI Societal Impact · 3AI Agents & Coding Assistants · 8AI-Generated Code / Vibe Coding · 2OpenClaw Hype Critique · 8

Primary evidence

Top Ranked Signals

Social Twitter Feb 8

I think agentic AI would work much better if people took lessons from organizational theory, which h...

By @emollick

95 score
AI Analysis

Emollick argues agentic AI needs organizational theory: spans of control (humans max ~10 reports, 100 subagents likely too many), boundary objects for coordination, proper coupling. Calls for more experiments with agent organization

I think agentic AI would work much better if people took lessons from organizational theory, which has actually spent a lot of time understanding how to deal with complex hierarchies, information limits, and spans of control. Right now most agentic AI systems seem to pretend that models have basically unlimited ability to manage subagents when that is clearly not true. We need measures of spans of control for AI. A human tops out at less than 10 direct reports. I am pretty sure that 100 subagen
Agentic AIOrganizational theoryMulti-agent systemsAI architecture
Social Twitter Feb 8

Top 100 LLMs by Downloads Since August 2025 Source: @interconnectsai HuggingFace Snapshots Model lis...

By @natolambert

92 score
AI Analysis

Nathan Lambert shares comprehensive Top 100 LLMs by downloads since August 2025, showing Qwen dominance (40 models), followed by Meta (13), DeepSeek (10). Llama-3.1-8B-Instruct leads with 53.3M downloads, GPT-OSS models prominent.

Top 100 LLMs by Downloads Since August 2025 Source: @interconnectsai HuggingFace Snapshots Model list on GitHub: Interconnects-AI/tracked-models (~1.5K models) Featuring: @alibaba_qwen: 40, @AIatMeta: 13, @deepseek_ai: 10, @Microsoft: 8, @GoogleAI: 7, @mistralai: 4, @OpenAI: 2, @allen_ai: 2, @vikhyatk: 1, @NVIDIAAI: 1, @huggingface: 1, @Zai_org: 1, @TencentGlobal: 1 1. meta-llama/Llama-3.1-8B-Instruct - 53.3M 2. Qwen/Qwen2.5-7B-Instruct - 52.4M 3. Qwen/Qwen2.5-VL-3B-Instruct - 49.5M 4. Qwen/Qw
open_source_modelsmarket_analysismodel_adoptionecosystem_trends
Social Twitter Feb 8

Back in 2023 everybody was telling me "no one uses Google search anymore, it's over" From 2023 to 2...

By @fchollet

90 score
AI Analysis

Chollet presents original Google data: search volume 61% growth to 5T queries/year, revenue 28% growth to $225B (56% of Google revenue); criticizes Twitter pundit AI disruption predictions

Back in 2023 everybody was telling me "no one uses Google search anymore, it's over" From 2023 to 2025, Google search query volume has grown 61% to 5T/year, and search revenue has grown 28% to $225B (56% of Google's revenue) The track record of Twitter pundits predicting AI disruption has been abysmal
Google SearchMarket analysisAI disruption predictions
Social Twitter Feb 8

Top 50 100B+ LLMs by Downloads Since August 2025 Featuring: @deepseek_ai: 16, @alibaba_qwen: 8, @AI...

By @natolambert

90 score
AI Analysis

Nathan Lambert's Top 50 100B+ LLMs by downloads - DeepSeek dominates with 16 models, GPT-OSS-120B leads at 22.3M downloads, followed by DeepSeek-R1 at 3.8M. Notable presence of Kimi-K2 (1T params) and MiniMax models.

Top 50 100B+ LLMs by Downloads Since August 2025 Featuring: @deepseek_ai: 16, @alibaba_qwen: 8, @AIatMeta: 5, @MiniMax_AI: 5, @Zai_org: 5, @Kimi_Moonshot: 3, @NVIDIAAI: 2, @Baidu_Inc: 2, @OpenAI: 1, @mistralai: 1, @Xiaomi: 1, @Snowflake: 1 1. openai/gpt-oss-120b (120.4B) - 22.3M 2. deepseek-ai/DeepSeek-R1 (671B) - 3.8M 3. deepseek-ai/DeepSeek-V3 (671B) - 3.6M 4. deepseek-ai/DeepSeek-R1-0528 (671B) - 3.5M 5. meta-llama/Llama-4-Scout-17B-16E-Instruct (108.6B) - 2.1M 6. Qwen/Qwen3-235B-A22B (
open_source_modelslarge_modelsmarket_analysischinese_ai_labs
Social Twitter Feb 8

Lots of folks are apparently in utter disbelief at these numbers, because *obviously* Google search ...

By @fchollet

88 score
AI Analysis

Chollet presents data: Google search queries grew 61% to 5T/year (2023-2025), revenue up 28% to $225B; usage is accelerating as of Q4 2025; challenges readers to update priors

Lots of folks are apparently in utter disbelief at these numbers, because *obviously* Google search died in 2023, *no one* is using Google at all in 2026, so the numbers must be wrong somehow, or maybe it's just AI agents making all these queries? Nope, it's a plain fact that more people than ever are using Google to search more than ever. In fact Google search usage is *accelerating* as of Q4 2025 Look, instead of grasping at straws, ask yourself why you were wrong about this and try to updat
Google SearchMarket analysisAI disruption predictions
Social Twitter Feb 8

video walkthrough of GPT-5.3 Codex:

By @gdb

88 score
AI Analysis

Continuing Brockman's coverage from Social yesterday, Brockman shares video walkthrough of GPT-5.3 Codex

video walkthrough of GPT-5.3 Codex:
GPT-5.3 CodexOpenAIProduct launch
Social Twitter Feb 8

The Claude 4.6 Opus UX automatically decides how long to think & it seems to suffer from the sam...

By @emollick

85 score
AI Analysis

Continuing our Social coverage of Claude 4.6 Opus, Emollick reports Claude 4.6 Opus auto-think UX has same issue as early GPT-5 router: doesn't take hard non-math/coding requests seriously. Notes OpenAI's fine-grained controls help

The Claude 4.6 Opus UX automatically decides how long to think & it seems to suffer from the same issue as the early GPT-5 router, not taking hard requests seriously when they are not specifically math/coding related The OpenAI system is complex, but fine-grained controls helps. t.co/R5f1nb03Hp
Claude 4.6 OpusGPT-5AI UXRouting systems
Social Twitter Feb 8

People don't want to accept that the top used open model families in 2026 are. Overall: 1. Qwen 2....

By @natolambert

85 score
AI Analysis

Nathan Lambert states top open model families in 2026: Overall - Qwen, Llama, GPT-OSS. For big models: DeepSeek leads, followed by GPT-OSS/Qwen. Notes 'people don't want to accept' this reality.

People don't want to accept that the top used open model families in 2026 are. Overall: 1. Qwen 2. Llama 3. GPT-OSS Big models: 1. DeepSeek 2. GPT-OSS/Qwen/everyone else
open_source_modelsmarket_analysisecosystem_trends
Social Twitter Feb 8

I didn't want to comment on OpenClaw. Usually, when there's so much noise in the media, it's some or...

By @burkov

82 score
AI Analysis

Adding to yesterday's Social discussion of OpenClaw, Burkov's comprehensive technical breakdown of OpenClaw: 2% ordinary code, 98% hype. Details how it's essentially wrapper around Microsoft's Playwright with LLM dispatch

I didn't want to comment on OpenClaw. Usually, when there's so much noise in the media, it's some ordinary stuff just hyped well. So I took time to learn how it works thanks to open source. I was right. OpenClaw is 2% of ordinary stuff and 98% of hype. To put it very shortly, in case you were wondering, there are two things in it: 1. You can chat with an LLM via a text messenger. Not anything new. 2. The LLM can use tools that run on your computer. Not anything new either. Most of the "mag
OpenClawTechnical analysisAI hypePlaywright