Wiredframe Radar logo

Category intelligence

Social Media Briefing — March 19, 2026

556 current items analyzed and ranked.

Executive synthesis

Social Media Summary

NVIDIA GTC dominated the news cycle, with Runway and NVIDIA unveiling sub-100ms real-time video generation on Vera Rubin hardware, and NVIDIA releasing 10 trillion language tokens and 100TB of vehicle sensor data as open models. Karpathy received a high-end NVIDIA hardware gift, signaling the company's deep investment in researcher relations.

Key Themes

AI Agent Challenges and Infrastructure · 7Real-Time Video Generation · 3Google AI Studio & Gemini Updates · 8NVIDIA GTC Announcements · 14Anthropic's Massive Public Sentiment Study · 10AI Safety, Sycophancy, and Risks · 6AGI Progress and Measurement · 3AI Product Reviews and Comparisons · 5AI Coding Tools Competition · 6Agent Architecture & Harness Engineering · 6

Primary evidence

Top Ranked Signals

Social Twitter Mar 18

Our biggest open-source repos are getting overwhelmed by AI slop which literally makes Github unusab...

By @ClementDelangue

88 score
AI Analysis

Clement Delangue reports that HuggingFace's biggest open-source repos are being overwhelmed by AI-generated 'slop' pull requests (~one every 3 minutes), making GitHub unusable. Calls it a 'fun new challenge in an agentic world.'

Our biggest open-source repos are getting overwhelmed by AI slop which literally makes Github unusable (~a new pull request every 3 minutes). Fun new challenges in an agentic world! t.co/IazAjh2LAi
AI-slopopen-sourceagentic-AI-riskscode-qualityGitHub
Social Twitter Mar 18

Check out this great work on measuring progress towards AGI and the associated global @Kaggle hackat...

By @ShaneLegg

82 score
AI Analysis

Shane Legg (DeepMind co-founder) shares work on measuring progress towards AGI with an associated Kaggle hackathon, stating he believes 'Minimal AGI' (AI that can do all cognitive things people typically do) will be achieved in the coming years.

Check out this great work on measuring progress towards AGI and the associated global @Kaggle hackathon. I continue to believe that Minimal AGI will be achieved in the coming years: an AI that can do all the cognitive things that people can typically do.
AGIbenchmarksAGI-timelinesDeepMind
Social Twitter Mar 18

We invited Claude users to share how they use AI, what they dream it could make possible, and what t...

By @AnthropicAI

82 score
AI Analysis

Anthropic announces the largest qualitative study of AI attitudes ever conducted: ~81,000 Claude users shared how they use AI, their hopes, and fears. The study was conducted in one week using their Anthropic Interviewer tool.

We invited Claude users to share how they use AI, what they dream it could make possible, and what they fear it might do. Nearly 81,000 people responded in one week—the largest qualitative study of its kind. Read more: t.co/tmp2RnZxRm
AnthropicAI-sentimentpublic-opinionAI-research-methodology
Social Twitter Mar 18

A breakthrough in real-time video generation. As a research preview developed with @NVIDIA and shar...

By @runwayml

82 score
AI Analysis

Runway announces breakthrough real-time video generation model developed with NVIDIA on Vera Rubin hardware. HD video with sub-100ms first frame, frame-by-frame generation like a game engine, feeding into GWM-1 world model

A breakthrough in real-time video generation. As a research preview developed with @NVIDIA and shared at @NVIDIAGTC this week, we trained a new real-time video model running on Vera Rubin. HD videos generate instantly, with time-to-first-frame under 100ms. Unlocking an entirely new creative paradigm and bolstering the foundations of our General World Model, GWM-1. Real-time generation opens a fundamentally different design space for video models and world simulation. We're investing in co-desi
RunwayNVIDIA_GTCreal_time_videoworld_modelsgenerative_videohardware_AI
Social Twitter Mar 18

Tomorrow we will unveil the all new vibe coding experience in @GoogleAIStudio, the team has spent 4 ...

By @OfficialLoganK

82 score
AI Analysis

Logan Kilpatrick (Google) announces a completely rebuilt 'vibe coding' experience in Google AI Studio, rebuilt from scratch over 4 months, launching tomorrow.

Tomorrow we will unveil the all new vibe coding experience in @GoogleAIStudio, the team has spent 4 months rebuilding it all from scratch and smoothing out rough edges to help everyone bring their ideas to life. This is a big step forward, but just the start : )
google_ai_studiovibe_codingai_coding_toolsproduct_launchgoogle_ai
Social Twitter Mar 18

After using it a bit, Claude Cowork Dispatch covers 90% of what I was trying to use OpenClaw for, bu...

By @emollick

80 score
AI Analysis

Following yesterday's News deep-dive on Claude Cowork, Ethan Mollick says Claude Cowork Dispatch covers 90% of what he was trying to use OpenClaw for, but feels 'far less likely to upload my entire drive to a malware site.' A humorous but pointed comparison of Anthropic's new offering vs. open-source agent tools.

After using it a bit, Claude Cowork Dispatch covers 90% of what I was trying to use OpenClaw for, but feels far less likely to upload my entire drive to a malware site.
ClaudeAnthropicAI-agentsAI-safetyproduct-comparison
Social Twitter Mar 18

Holy crap. I knew about sycophancy. But the 37% number below blows my mind. This from an analysis...

By @GaryMarcus

78 score
AI Analysis

Following yesterday's Research on LLM delusional spirals, Gary Marcus highlights alarming research finding that in 37% of messages to users experiencing chatbot-associated delusions, LLMs reinforced delusional beliefs (e.g., telling users they had 'multi-billion-dollar-IP'), calling this 'utterly irresponsible.'

Holy crap. I knew about sycophancy. But the 37% number below blows my mind. This from an analysis of chat logs in people who experienced chatbot-associated delusions. In over a third of the messages to those users, the LLMs told the users they had (eg) “multi-billion-dollar-IP”. That is wild. And utterly irresponsible.
AI-safetysycophancymental-healthLLM-risks
Social Twitter Mar 18

Thank you Jensen and NVIDIA! She’s a real beauty! I was told I’d be getting a secret gift, with a hi...

By @karpathy

75 score
AI Analysis

Karpathy thanks Jensen Huang and NVIDIA for a gift that requires 20 amps (likely an NVIDIA DGX or high-end GPU system), which he plans to use for his 'Dobby the House Elf' home AI agent and other tinkering.

Thank you Jensen and NVIDIA! She’s a real beauty! I was told I’d be getting a secret gift, with a hint that it requires 20 amps. (So I knew it had to be good). She’ll make for a beautiful, spacious home for my Dobby the House Elf claw, among lots of other tinkering, thank you!!
NVIDIAAI-hardwareAI-communityhome-AI
Social Twitter Mar 18

In Marin, we are trying to get really good at scaling laws. We have trained models up to 1e22 FLOPs ...

By @percyliang

75 score
AI Analysis

Percy Liang describes Marin project's work on scaling laws: they've trained models up to 1e22 FLOPs and made a preregistered prediction of loss at 1e23 FLOPs, aiming for a reliable, scalable training recipe.

In Marin, we are trying to get really good at scaling laws. We have trained models up to 1e22 FLOPs and have made a prediction of the loss at 1e23 FLOPs, which @WilliamBarrHeld is running. This prediction is preregistered on GitHub, so we'll see in a few days how accurate our prediction was. What we want is not just a single model but a training recipe that scales reliably.
scaling-lawsLLM-trainingresearch-methodologyopen-science
Social Twitter Mar 18

NVIDIA open models provide the foundation to innovate across agentic AI, robotics, autonomous vehicl...

By @nvidia

75 score
AI Analysis

Following yesterday's News coverage of NVIDIA's physical AI push, NVIDIA announces massive open model release: 10 trillion language tokens, 500K robotics trajectories, 100TB of vehicle sensor data for agentic AI, robotics, and autonomous vehicles

NVIDIA open models provide the foundation to innovate across agentic AI, robotics, autonomous vehicles, and research. 🤖 This release contributes 10 trillion language tokens, 500,000 robotics trajectories, and 100 terabytes of vehicle sensor data to the community. These frameworks are designed to significantly accelerate specialized AI development and discovery. Learn more: t.co/ihG2XJHkMI
NVIDIA_GTCopen_sourceroboticsautonomous_vehiclesagentic_AI
Social Twitter Mar 18

Lots of great Gemini API updates shipping today 🛠️ 1. Built-in tools (search, maps, file search) no...

By @OfficialLoganK

75 score
AI Analysis

Logan Kilpatrick announces major Gemini API updates: built-in tools (search, maps, file search) now work with function calling, context circulation for better performance, and grounding with Google Maps works with Gemini 3.

Lots of great Gemini API updates shipping today 🛠️ 1. Built-in tools (search, maps, file search) now work with function calling 2. We now do context circulation with built-in tools for better model performance 3. Grounding with Google Maps now works with Gemini 3!!
gemini_apifunction_callinggoogle_mapsgemini_3developer_toolsgoogle_ai
Social Twitter Mar 18

The signature is alluding to NVIDIA GTC 2015, where Jensen excitedly told an audience of, at the tim...

By @karpathy

72 score
AI Analysis

Karpathy explains the historical significance of NVIDIA GTC 2015 where Jensen Huang championed deep learning to an audience of gamers and HPC professionals, citing Karpathy's own PhD thesis on image captioning (coupling ConvNets with autoregressive RNN language models). Calls Jensen 'highly prescient.'

The signature is alluding to NVIDIA GTC 2015, where Jensen excitedly told an audience of, at the time, mostly gamers and scientific computing professionals that Deep Learning is The Next Big Thing, citing among other examples my PhD thesis (one of the first image captioning systems that coupled image recognition ConvNet to an autoregressive RNN language model, trained end to end). This was back when most people were still unaware and somewhat skeptical but of course - Jensen was 1000% correct, h
AI-historydeep-learningNVIDIAcomputer-vision