Wiredframe Radar logo

Category intelligence

Social Media Briefing — February 13, 2026

573 current items analyzed and ranked.

Executive synthesis

Social Media Summary

A landmark day dominated by Anthropic's $30B raise at a $380B valuation and simultaneous reveal of $14B run-rate revenue with 10x annual growth. An Anthropic engineer attributed much of the fundraise momentum to Claude Code, whose weekly active users doubled since January.

  • Sam Altman launched GPT-5.3-Codex-Spark as a research preview, touting 1000+ tokens/sec via a new Cerebras hardware partnership
  • Demis Hassabis and Noam Shazeer announced a major Gemini 3 Deep Think upgrade achieving SOTA on ARC-AGI-2 (84.6%), 3455 Codeforces Elo, and gold-medal Physics/Chemistry Olympiad performance — independently certified by François Chollet
  • Google also unveiled Aletheia, a math research agent powered by Deep Think that solved multiple open Erdős problems

On the ideas front, John Carmack argued AI will shift economic value from raw intelligence to agency, empowering a new class of high-agency individuals. Andrej Karpathy highlighted Simile AI's novel approach of using LLMs as population simulators rather than single-personality chatbots. Chollet provided a definitive historical account of the ARC benchmarks, pushing back on narratives that they were designed as anti-LLM tests.

Key Themes

Anthropic $30B Funding & Revenue Milestone · 6Claude Code Growth & Anthropic Fundraise · 5Gemini 3 Deep Think Launch & Benchmarks · 16GPT-5.3-Codex-Spark Launch · 4Gemini 3 Deep Think Upgrade · 13ARC Benchmark History & Roadmap · 9Aletheia Math Research Agent · 5Gemini Deep Think 3 Launch · 4AI and Human Agency/Society · 5New Model Releases & Benchmarking · 6

Primary evidence

Top Ranked Signals

Social Twitter Feb 12

We’ve raised $30B in funding at a $380B post-money valuation. This investment will help us deepen o...

By @AnthropicAI

95 score
AI Analysis

Anthropic announces $30B fundraise at $380B post-money valuation to deepen research, innovate products, and expand infrastructure.

We’ve raised $30B in funding at a $380B post-money valuation. This investment will help us deepen our research, continue to innovate in products, and ensure we have the resources to power our infrastructure expansion as we make Claude available everywhere our customers are.
Anthropic BusinessAI FundingAI Industry
Social Twitter Feb 12

GPT-5.3-Codex-Spark is launching today as a research preview for Pro. More than 1000 tokens per sec...

By @sama

93 score
AI Analysis

Sam Altman announces GPT-5.3-Codex-Spark launching as research preview for Pro users, achieving over 1000 tokens per second.

GPT-5.3-Codex-Spark is launching today as a research preview for Pro. More than 1000 tokens per second! There are limitations at launch; we will rapidly improve.
product-launchopenaicodexinference-speed
Social Twitter Feb 12

Thrilled to announce a big upgrade to Gemini 3 Deep Think that hits new records on the most rigorous...

By @demishassabis

92 score
AI Analysis

Building on yesterday's Social buzz, Hassabis announces major Gemini 3 Deep Think upgrade with record benchmarks: 84.6% ARC-AGI-2, 48.4% Humanity's Last Exam (no tools), 3455 Elo on Codeforces.

Thrilled to announce a big upgrade to Gemini 3 Deep Think that hits new records on the most rigorous benchmarks in maths, science & reasoning - including 84.6% on ARC-AGI-2, 48.4% Humanity’s Last Exam without tools, and 3455 Elo rating on Codeforces! t.co/D3FuMwaLpr
gemini-deep-thinkbenchmarksarc-agiproduct-launchreasoning
Social Twitter Feb 12

A huge part of this raise is Claude Code. Weekly active users doubled since January. People who've ...

By @bcherny

92 score
AI Analysis

Building on yesterday's Social buzz, Boris Cherny (Anthropic engineer) states that a huge part of Anthropic's fundraise is driven by Claude Code. Weekly active users doubled since January, and non-coders are building with it.

A huge part of this raise is Claude Code. Weekly active users doubled since January. People who've never written a line of code are building with it. Humbled to work on this every day with our team.
anthropicclaude-codeai-codingai-businessfundraising
Social Twitter Feb 12

An updated Gemini 3 Deep Think is out today: 📈 Achieves SOTA on ARC-AGI-2, MMMU-Pro, and HLE. 🥇Gold...

By @NoamShazeer

88 score
AI Analysis

Building on yesterday's Social buzz, Noam Shazeer announces updated Gemini 3 Deep Think achieving SOTA on ARC-AGI-2, MMMU-Pro, HLE, plus gold-medal level on Physics & Chemistry Olympiads.

An updated Gemini 3 Deep Think is out today: 📈 Achieves SOTA on ARC-AGI-2, MMMU-Pro, and HLE. 🥇Gold-medal level on Physics & Chemistry Olympiads. It turns out the best way to solve hard problems is still to think about them. Read more: t.co/PMmuyRq90B t.co/cxCAbcEz7q
Gemini Deep ThinkAI BenchmarksModel Release
Social Twitter Feb 12

The modern age has richly rewarded people with a combination of high intelligence and high agency. N...

By @ID_AA_Carmack

88 score
AI Analysis

John Carmack argues that AI automation of intelligence will empower people with high agency but lower intelligence, if they trust AI advice. Uses provocative example of a 'ruthless criminal' with always-on AI glasses.

The modern age has richly rewarded people with a combination of high intelligence and high agency. Now that many aspects of intelligence are successfully being automated, it seems likely that people with relatively lower intelligence but exceptional agency will come into their own if they are willing to egolessly accept AI advice. Imagine a ruthless criminal that completely trusts everything their always-on AI glasses are telling them, knowing that it is carefully looking out for their best int
AI and human agencyAI augmentationsocietal impact of AIfuture of work
Social Twitter Feb 12

Lots of folks spread false narratives about how ARC-1 was created in response to LLMs, or how ARC-2 ...

By @fchollet

85 score
AI Analysis

Chollet provides comprehensive history of ARC benchmarks: designed 2017-2019, progress came from test-time adaptation (not LLM scaling), base LLMs still perform poorly, ARC never claimed to prove AGI. Details ARC-3 through ARC-7 roadmap.

Lots of folks spread false narratives about how ARC-1 was created in response to LLMs, or how ARC-2 was only created because ARC-1 was saturated. Setting the record straight: 1. ARC-1 was designed 2017-2019 and released in 2019 (pre LLMs). 2. The coming of ARC-2 was announced in May 2022 (pre ChatGPT). 3. By mid-2024, there was still essentially no progress on ARC-1. 4. All progress on ARC-1 & ARC-2 came from a new paradigm, test-time adaptation models, starting in late 2024 and ramping up t
benchmarksarc-agitest-time-adaptationai-historyagi-definition
Social Twitter Feb 12

Our run-rate revenue is $14 billion, and has grown over 10x in each of the past 3 years. This growth...

By @AnthropicAI

85 score
AI Analysis

Anthropic announces $14B run-rate revenue with 10x annual growth, positioning as the intelligence platform of choice for enterprises.

Our run-rate revenue is $14 billion, and has grown over 10x in each of the past 3 years. This growth has been driven by our position as the intelligence platform of choice for enterprises and developers. Read more: t.co/aMRyOkFFSg
Anthropic BusinessAI IndustryAI Economics
Social Twitter Feb 12

Reaching AGI won't be beating a benchmark. It will be the end of the human-AI gap. Benchmarks are si...

By @fchollet

82 score
AI Analysis

Chollet defines AGI as the end of the human-AI gap, argues benchmarks are a process not a fixed point, and expects the gap to close by ~2030.

Reaching AGI won't be beating a benchmark. It will be the end of the human-AI gap. Benchmarks are simply a way to estimate the current gap, which is why we need to continually release new benchmarks (focused on the remaining gap). Benchmarking is a process, not a fixed point. We can say we have AGI when it's no longer possible to come up with a test that evidences the gap. When it's no longer possible to point to something that regular humans can do and AI can't. Today, it's still easy. I expe
agi-definitionbenchmarksai-timelines
Social Twitter Feb 12

Introducing Aletheia, a math research agent powered by an advanced version of Gemini Deep Think that...

By @YiTayML

82 score
AI Analysis

As first reported in Research yesterday, Yi Tay introduces Aletheia, a math research agent powered by Gemini Deep Think that produces publishable research and solved multiple open Erdős problems.

Introducing Aletheia, a math research agent powered by an advanced version of Gemini Deep Think that produces publishable math research (two papers, one completely automatic and another with human-AI collaboration) and solved multiple open Erdős problems. 😀🔥 Paper link below! 👇
AI Math ResearchGemini Deep ThinkAI AgentsAI for Science