Wiredframe Radar logo

Category intelligence

Social Media Briefing — February 28, 2026

529 current items analyzed and ranked.

Executive synthesis

Social Media Summary

The AI community was dominated by two seismic stories: OpenAI's unprecedented $110B funding round from Amazon, NVIDIA, and SoftBank, and the heated debate over AI labs refusing US government mass surveillance requests.

  • Sam Altman announced the largest private funding round in history, detailing new Amazon partnership for a stateful runtime environment and continued Azure API exclusivity with Microsoft
  • Ilya Sutskever praised Anthropic for not backing down on surveillance, warning of harder challenges ahead; Jan Leike sharply framed the government as seeking a new "mass domestic surveillance" supplier
  • François Chollet argued AI performance remains fundamentally tied to task familiarity, with novel domains still exposing deep limitations

On the technical side, Andrej Karpathy shared detailed experiments running multi-agent research orgs (4 Claude + 4 Codex), finding agents strong at implementation but weak at creative ideation and experiment design. He also mapped the evolution from tab-complete to parallel agent teams in Cursor data. David Ha highlighted paradigm-shifting research on hypernetworks compiling documents directly into model weights as an alternative to long context windows. Anthropic disclosed a prompt caching bug affecting Claude Code rate limits and rolled out Claude Code Remote for Pro users.

Key Themes

Mass Surveillance & AI Labs' Refusal · 14OpenAI $110B Funding Round · 7OpenAI Mega Funding Round · 4AI Agents & Research Automation · 5Perplexity Computer Launch & Samsung Partnership · 16Claude Code Bug & Prompt Caching · 3AI Capabilities & Limitations · 8Google AI Product Launches · 3Claude Code Updates & Issues · 14Awni Hannun Departing Apple · 5

Primary evidence

Top Ranked Signals

Social Twitter Feb 27

We have raised a $110 billion round of funding from Amazon, NVIDIA, and SoftBank. We are grateful f...

By @sama

98 score
AI Analysis

Sam Altman announces OpenAI has raised $110 billion from Amazon, NVIDIA, and SoftBank - the largest private funding round ever.

We have raised a $110 billion round of funding from Amazon, NVIDIA, and SoftBank. We are grateful for the support from our partners, and have a lot of work to do to bring you the tools you deserve.
OpenAI FundingAI InvestmentAI Industry Landscape
Social Twitter Feb 27

It’s extremely good that Anthropic has not backed down, and it’s siginficant that OpenAI has taken a...

By @ilyasut

95 score
AI Analysis

Following yesterday's News coverage of Anthropic's refusal to comply with Pentagon demands, Ilya Sutskever praises Anthropic for not backing down and notes OpenAI has taken a similar stance. Warns of more challenging situations ahead and calls for AI leaders and competitors to unite when needed.

It’s extremely good that Anthropic has not backed down, and it’s siginficant that OpenAI has taken a similar stance. In the future, there will be much more challenging situations of this nature, and it will be critical for the relevant leaders to rise up to the occasion, for fierce competitors to put their differences aside. Good to see that happen today.
AI EthicsMass SurveillanceAI Industry UnityAI SafetyAI Leadership
Social Twitter Feb 27

Helping AI reach more people requires deep collaboration across the ecosystem. Today we’re announci...

By @OpenAI

95 score
AI Analysis

OpenAI announces massive $110B funding round with investments from SoftBank, NVIDIA, and Amazon to scale AI infrastructure

Helping AI reach more people requires deep collaboration across the ecosystem. Today we’re announcing new investment, with support from @SoftBank, @NVIDIA, and @Amazon, to scale the infrastructure needed to bring AI to everyone. t.co/xW0ItgMTLe
openai-fundingai-infrastructuremega-investmentindustry-partnerships
Social Twitter Feb 27

I had the same thought so I've been playing with it in nanochat. E.g. here's 8 agents (4 claude, 4 c...

By @karpathy

92 score
AI Analysis

Karpathy details his experiments running 8 AI agents (4 Claude, 4 Codex) as a 'research org' trying to solve a nanochat problem. Agents are good at implementing well-scoped ideas but fail at experiment design, creative ideation, controlling for confounds, and proper ablation studies. Envisions programming 'organizations' with prompts, tools, and processes as source code.

I had the same thought so I've been playing with it in nanochat. E.g. here's 8 agents (4 claude, 4 codex), with 1 GPU each running nanochat experiments (trying to delete logit softcap without regression). The TLDR is that it doesn't work and it's a mess... but it's still very pretty to look at :) I tried a few setups: 8 independent solo researchers, 1 chief scientist giving work to 8 junior researchers, etc. Each research program is a git branch, each scientist forks it into a feature branch, g
AI AgentsAI Research AutomationLLM LimitationsMulti-Agent Systems
Social Twitter Feb 27

This week we had a BUNCH of updates 🍌🍌 Here’s what we announced: — Nano Banana 2, our SoTA image ge...

By @GoogleAI

82 score
AI Analysis

Google AI summarizes a major week of announcements: Nano Banana 2 (image gen), Veo 3.1 and Lyria 3 templates in Gemini, Flow redesign, Opal agent step, Producer AI joining Labs, Gemini training for 6M US educators, Samsung Galaxy S26 AI features

This week we had a BUNCH of updates 🍌🍌 Here’s what we announced: — Nano Banana 2, our SoTA image generation model, which you can use to generate images with real-world knowledge, add text in many languages, and bring your most creative ideas to life — Nano Banana, Veo 3.1, and Lyria 3 templates in @GeminiApp to empower easy image, video, and music track creation — @FlowbyGoogle got a major redesign, taking a big step towards becoming your AI creative studio — Opal by @GoogleLabs introduce
google-ai-productsimage-generationai-educationon-device-aicreative-ai
Social Twitter Feb 27

Claude Code Remote is rolling out now for Pro users

By @bcherny

82 score
AI Analysis

Anthropic employee bcherny announces Claude Code Remote is rolling out for Pro users. Very high engagement (2,417 likes, 182K views).

Claude Code Remote is rolling out now for Pro users
anthropicclaude-codeproduct-launchdeveloper-tools
Social Twitter Feb 27

We've reset rate limits for all Claude Code users. Yesterday we rolled out a bug with prompt cachin...

By @trq212

82 score
AI Analysis

Anthropic has reset rate limits for all Claude Code users after a prompt caching bug caused usage limits to be consumed faster than normal. Hotfixed in version 2.1.62.

We've reset rate limits for all Claude Code users. Yesterday we rolled out a bug with prompt caching that caused usage limits to be consumed faster than normal. This is hotfixed in 2.1.62. Make sure you upgrade to the latest and hope you enjoy using Claude Code this weekend!
claude_codeprompt_cachingbug_fixanthropicdeveloper_tools
Social Twitter Feb 27

BREAKING: OpenAI raises another massive round: $110B at $730B pre-money valuation - $50B from Amazo...

By @TheRundownAI

80 score
AI Analysis

TheRundownAI reports OpenAI's $110B funding round at $730B pre-money valuation: $50B from Amazon, $30B from SoftBank, $30B from Nvidia

BREAKING: OpenAI raises another massive round: $110B at $730B pre-money valuation
  • $50B from Amazon
  • $30B from SoftBank
  • $30B from Nvidia
t.co/FRNpqc8i8d
openai-fundingmega-investmentvaluation
Social Twitter Feb 27

Cool chart showing the ratio of Tab complete requests to Agent requests in Cursor. With improving ca...

By @karpathy

78 score
AI Analysis

Karpathy discusses the evolution of AI-assisted coding from Tab-complete → Agent → Parallel agents → Agent Teams, referencing Cursor data. Advises 80% productive work / 20% exploration of next paradigm.

Cool chart showing the ratio of Tab complete requests to Agent requests in Cursor. With improving capability, every point in time has an optimal setup that keeps changing and evolving and the community average tracks the point. None -> Tab -> Agent -> Parallel agents -> Agent Teams (?) -> ??? If you're too conservative, you're leaving leverage on the table. If you're too aggressive, you're net creating more chaos than doing useful work. The art of the process is spending 80% of the time gettin
AI AgentsDeveloper WorkflowsAI-Assisted Coding
Social Twitter Feb 27

Perplexity's APIs (standalone search, search-grounded LLMs) are being used in all Samsung devices an...

By @AravSrinivas

78 score
AI Analysis

Arav Srinivas announces Perplexity APIs are now used in all Samsung devices and some Magnificent 7 companies. Also announces first developer event 'Ask' on March 11.

Perplexity's APIs (standalone search, search-grounded LLMs) are being used in all Samsung devices and some of the Mag 7 companies as of today. We're excited to announce our first developer event, Ask. March 11. Apply here if you're a dev building exciting agentic products!
PerplexitySamsung partnershipenterprise AIsearch APIsdeveloper events
Social Twitter Feb 27

We are excited to partner with Amazon to bring a new generation of products to market, especially ar...

By @sama

75 score
AI Analysis

Altman announces Amazon partnership for new enterprise products including a 'stateful runtime environment' and use of Trainium chips.

We are excited to partner with Amazon to bring a new generation of products to market, especially around new enterprise products like the stateful runtime environment. We are also very excited to make great use of Tranium.
OpenAI FundingAI InfrastructureEnterprise AICloud Partnerships
Social Twitter Feb 27

Even after the steep progress of the past 3 months, it remains that AI performance is tied to task f...

By @fchollet

75 score
AI Analysis

Chollet argues AI performance remains tied to task familiarity: densely sampleable domains see unbounded improvement, while novel/unfamiliar domains remain low and need new ideas, not just more data/compute.

Even after the steep progress of the past 3 months, it remains that AI performance is tied to task familiarity. In domains that can be densely sampled (via programmatic generation + verification), performance is effectively unbounded, and will keep increasing from current levels. In novel, unfamiliar domains, performance remains low and further progress still requires new ideas, not just more data and compute.
AI CapabilitiesLLM LimitationsAGIIntelligence Science