88 score
AI Analysis
trq212 (Anthropic) announces that Claude Code can now control your computer
Claude Code can control your computer now!
claude-codecomputer-useanthropicagentic-aiproduct-launch
Category intelligence
Social Media Briefing — March 24, 2026
375 current items analyzed and ranked.
Executive synthesis
Social Media Summary
A major Anthropic feature drop dominated discussion: Claude Code can now control your computer directly, signaling a significant leap in agentic coding capabilities. Meanwhile, Jeremy Howard pushed back hard on Claude Opus & Sonnet 4.6, calling them overly agentic—taking over tasks rather than collaborating with humans.
- Sam Altman announced leaving the Helion Energy board as OpenAI and Helion explore large-scale energy collaboration, underscoring AI's growing infrastructure demands
- Ethan Mollick shaped the model comparison narrative, declaring GPT-5.4 Pro uniquely capable for complex academic work, with Opus 4.6 roughly matching GPT-5.4 Thinking but nothing rivaling Pro
- Mollick also offered an original breakdown of Codex vs Claude Code design philosophies—functional skill references versus holistic workflow guidance
- Lex Fridman's full interview with NVIDIA CEO Jensen Huang drew over 1M views, covering scaling laws, AGI timelines, and AI's role across industries
Anthropic launched its Science Blog to showcase AI-accelerated research, while NVIDIA and Emerald AI announced flexible AI factories with major energy partners at CERAWeek, highlighting the rapidly scaling physical infrastructure behind the AI boom.
Key Themes
Claude Code Computer Control · 3OpenAI-Helion Energy Partnership · 1GPT-5.4 Pro Dominance & Model Comparison · 8Claude 4.6 Over-Eagerness Problem · 4Jensen Huang / NVIDIA / GTC 2026 · 3AI Agents & Enterprise Deployment · 8AI Coding Tools Philosophy · 3AI Infrastructure & Energy · 4AI's Societal & Cultural Impact · 8AI Agent Skepticism & Design · 5
Primary evidence
Top Ranked Signals
Social Twitter Mar 23
I have loved being on the Helion board; I continue to be extremely excited about a future with abund...
By @sama
82 score
AI Analysis
Sam Altman announces leaving the Helion Energy board as OpenAI and Helion explore large-scale collaboration. He'll retain financial interest but steps down from governance to reduce conflicts.
I have loved being on the Helion board; I continue to be extremely excited about a future with abundant energy and Helion in particular.
As Helion and OpenAI start to explore working together at significant scale, it is difficult for me to be on both boards. (I will have a financial interest in Helion so still be recused from negotiations, but from a governance perspective this will make things easier for both companies.)
David and Chris are exceptional founders and I think people will be quit
openai-strategyenergy-infrastructurecorporate-governanceai-infrastructure
Social Twitter Mar 23
Here's my conversation with Jensen Huang, CEO of NVIDIA, the most valuable & one of the most influen...
By @lexfridman
78 score
AI Analysis
Continuing our coverage from yesterday, Lex Fridman releases a full conversation with Jensen Huang covering AI scaling laws, supply chain, TSMC, China, consciousness, AGI timelines, and more.
Here's my conversation with Jensen Huang, CEO of NVIDIA, the most valuable & one of the most influential companies in the history of human civilization. It is the engine powering the AI revolution.
This was a fascinating & inspiring conversation, in parts super-technical on engineering of every part of the AI stack, memory, power, supply chain (TSMC, ASML, etc), in parts about leadership & psychology, and in parts personal & philosophical about life, consciousness, mortality, and human nature.
nvidiaai-scaling-lawsai-infrastructureagi-timelinessupply-chainjensen-huang
Social Twitter Mar 23
Opus & Sonnet 4.6 haven't been a great hit for most of my work, or our customers, since (as warn...
By @jeremyphoward
78 score
AI Analysis
jeremyphoward criticizes Claude Opus & Sonnet 4.6 for being overly agentic - they take over tasks rather than letting humans lead. Asks for model recommendations that are 'patient followers'.
Opus & Sonnet 4.6 haven't been a great hit for most of my work, or our customers, since (as warned in their tech report) they're over-enthusiastic about agentically taking over, rather than letting the human lead.
Any suggestions for competent models that are patient followers?
claude-4.6anthropicmodel-behaviorhuman-ai-interactionagentic-aiuser-experience
Social Twitter Mar 23
GPT-5.4 Pro continues to be the only model of its class. For anything really hard & complex, I t...
By @emollick
75 score
AI Analysis
Emollick declares GPT-5.4 Pro continues to be the only model of its class for hard and complex tasks. He feeds it maximum context and gets useful results that no other tool (Codex, Code, etc.) can match.
GPT-5.4 Pro continues to be the only model of its class. For anything really hard & complex, I throw it into the maw with every bit of context I can think of. More often than not, something very useful comes out.
I can't get the same results from Codex or Code or anything else.
gpt-5.4model-comparisonai-capabilities
Social Twitter Mar 23
Very different philosophies for skills in Codex versus Claude Code OpenAI seems to conceive of skil...
By @emollick
73 score
AI Analysis
Emollick compares Codex vs Claude Code skill philosophies: OpenAI conceives skills functionally as technical references, while Claude skills focus on giving AI approaches to problems.
Very different philosophies for skills in Codex versus Claude Code
OpenAI seems to conceive of skills functionally, mostly matter-of-fact technical references for Codex. Claude skills are more about giving the AI approaches to problems
See the difference in skill creator skills t.co/Gq3hcf1In0
ai-coding-toolsopenai-productsanthropic-productsdesign-philosophy
Social Twitter Mar 23
I alternate between Opus 4.6 and GPT-5.4 Thinking for most work I do (and a lot of Google tools like...
By @emollick
72 score
AI Analysis
Emollick describes his AI workflow: alternates between Opus 4.6 and GPT-5.4 Thinking for most work, uses Google tools like NotebookLM, and both Codex and Code. GPT-5.4 Pro is a different category for very hard problems.
I alternate between Opus 4.6 and GPT-5.4 Thinking for most work I do (and a lot of Google tools like NotebookLM where Gemini has excellent harnesses), and use both Codex and Code a lot, but GPT-5.4 Pro is just a different category of model for a set of very hard problems.
gpt-5.4model-comparisonclaude-opus-4.6ai-workflowsai-coding-tools
Social Twitter Mar 23
NVIDIA and @EmeraldAi_ are announcing a new class of flexible AI factories at #CERAWeek 🔋 Working w...
By @nvidia
72 score
AI Analysis
NVIDIA and Emerald AI announce a new class of flexible AI factories at CERAWeek, partnering with major energy companies (AES, Constellation Energy, Invenergy, NextEra Energy, VISTRA) to speed up AI power infrastructure and support grid reliability.
NVIDIA and @EmeraldAi_ are announcing a new class of flexible AI factories at #CERAWeek 🔋
Working with partners AES, @ConstellationEG, @InvenergyLLC, @nexteraenergy, and VISTRA, this collaboration aims to speed up time to power and support grid reliability. These next-generation facilities will use grid intelligence to help the United States win the AI race.
Read the full announcement: t.co/hGymevUSxB
ai-infrastructureai-energynvidia-ecosystem
Social Twitter Mar 23
Teleop is so 2025. Ever since we unveiled EgoScale and the dexterity scaling law, it's been clear to...
By @DrJimFan
70 score
AI Analysis
Jim Fan (NVIDIA) declares teleoperation is outdated, citing EgoScale and dexterity scaling law. 2026 is about behavior cloning from humans and scaling robot learning without robots.
Teleop is so 2025. Ever since we unveiled EgoScale and the dexterity scaling law, it's been clear to us and the ecosystem that behavior cloning directly from humans is the way to break the curse of teleop. 2026 is all about scaling robot learning without robots.
roboticsbehavior-cloningscaling-lawsnvidia-research
Social Twitter Mar 23
Following the success of the EurIPS and NeurIPS-Mexico City pilots in 2025, we are thrilled to annou...
By @NeurIPSConf
70 score
AI Analysis
NeurIPS 2026 announces two satellite events in Paris, France and Atlanta, USA, alongside the main conference in Sydney, Australia. Both will feature keynotes, oral/poster presentations, and workshops.
Following the success of the EurIPS and NeurIPS-Mexico City pilots in 2025, we are thrilled to announce two official NeurIPS 2026 satellite events for this year!
These will be held in Paris, France and Atlanta, USA, respectively, running alongside the main venue in Sydney, Australia.
Both satellite events will feature keynotes, oral and poster presentations of accepted NeurIPS 2026 papers, as well as workshops. We are planning tutorials, affinity events, and other elements for the satellite s
ml-conferencesai-researchai-community
Social Twitter Mar 23
We recently worked with The Yomiuri Shimbun to analyze more than a million social media posts to map...
By @hardmaru
68 score
AI Analysis
Hardmaru describes collaboration with The Yomiuri Shimbun using LLM ensembles and Novelty Search to analyze 1M+ social media posts to map state-sponsored information campaigns, with journalists verifying AI-generated hypotheses.
We recently worked with The Yomiuri Shimbun to analyze more than a million social media posts to map out state-sponsored information campaigns.
t.co/fBQQLg1glS
Keyword searches are fragile for modern OSINT. To fix this, our team used an ensemble of different LLMs combined with our Novelty Search algorithm to extract underlying narratives purely from context. (e.g., The system successfully mapped posts demanding "a politician retract a statement" to the broader, hidden narrative of "Tai
ai-for-osintinformation-warfarehuman-ai-collaborationllm-applications
Social Twitter Mar 23
Introducing the Anthropic Science Blog. Increasing the pace of scientific progress is a core part o...
By @AnthropicAI
68 score
AI Analysis
Anthropic launches the Anthropic Science Blog to feature research and stories of scientists using AI to accelerate their work.
Introducing the Anthropic Science Blog.
Increasing the pace of scientific progress is a core part of Anthropic’s mission. The Science Blog will feature new research and stories of how scientists are using AI to accelerate their work.
Read the intro: t.co/1P9BDyX3xG
anthropic-strategyai-for-science