Wiredframe Radar logo

Category intelligence

Social Media Briefing — May 11, 2026

432 current items analyzed and ranked.

Executive synthesis

Social Media Summary

A debate over AI's geographic center of gravity dominated the day. Yann LeCun forcefully argued that major innovations—Attention, PyTorch, AlphaGo, AlphaFold—emerged from Montréal, NYC, and London, not Silicon Valley. Ethan Mollick echoed the theme, noting AI users now span every industry globally.

Key Themes

AI Geographic Distribution · 3Local/Open-Source AI Acceleration · 1Cloudflare Email & SaaS Commoditization · 14OpenAI Next Model Hints · 3AI-Assisted Development & Code Security · 6Context Engineering & AI Tooling · 3LLM Inference Optimization · 1METR Benchmark Critique · 10AI Adoption Beyond Tech · 3Agentic AI Origins and Architecture · 3

Primary evidence

Top Ranked Signals

Social Twitter May 10

@eladgil BS. Attention was born in Montréal PyTorch in NYC. AlphaGo in London AlphaFold in London ...

By @ylecun

78 score
AI Analysis

Yann LeCun pushes back on Silicon Valley centrism in AI, listing major innovations from Montreal (Attention), NYC (PyTorch), London (AlphaGo, AlphaFold), Paris (Llama 1, DINO), and Hangzhou (DeepSeek). Claims SV is only 3 months ahead on topics it's singularly obsessed with.

@eladgil BS. Attention was born in Montréal PyTorch in NYC. AlphaGo in London AlphaFold in London ESMFold in NYC Llama 1 in Paris. Llama 2 in Paris+NYC+SV DeepSeek in Hangzhou Plus: DINO in Paris JEPA in Montréal+Paris+NYC SV is 3 mos ahead on topics SV is singularly obsessed with.
ai_geographyai_innovationsilicon_valley_critiqueglobal_ai
Social Twitter May 10

Local AI is having its moment! Below is the number of new GGUF models created each month over the p...

By @ClementDelangue

72 score
AI Analysis

Clement Delangue (Hugging Face CEO) reports explosive growth in GGUF model creation - 176K total models, with March-April doubling the previous rate to ~9.2K new models/month, signaling local AI acceleration.

Local AI is having its moment! Below is the number of new GGUF models created each month over the past 8 months & insights from our HF internal agent (May is partial):
  • 176,000 total public GGUF models on HF
  • Two distinct regimes: Oct–Feb averaged ~5.1K new GGUF models/month. Then March–April jumped to ~9.2K/month — nearly double the previous rate.
  • March was the inflection point (+55% MoM) — likely driven by a wave of new open-weight model releases being quantized to GGUF.
  • April susta
local_aiopen_source_aiggufai_ecosystem_growth
Social Twitter May 10

✉️ Trying @Cloudflare's new Email Sending feature today If you send 1,000,000 emails per month: - P...

By @levelsio

68 score
AI Analysis

Levelsio provides a detailed price comparison of email providers at 1M emails/month (Postmark $1206, Resend $650, SendGrid $600, Cloudflare $354, SES $100), announces he's trying Cloudflare Email, and argues email sending has become a commodity especially with AI making integration easy.

✉️ Trying @Cloudflare's new Email Sending feature today If you send 1,000,000 emails per month:
  • Postmark: $1,206/mo
  • Resend: $650/mo
  • SendGrid: $600/mo
  • Cloudflare: $354/mo
  • Amazon SES: $100/mo
So Postmark is now by far the most expensive email provider And SES and Cloudflare are now the cheapest email providers I know my friend @marckohlbrugge is trying out SES now so I'll try Cloudflare and see how it is, SES is cheaper but Marc said it takes a bit more managing, and since I already
Email infrastructure costsCloudflare emailDeveloper toolingAI-assisted developmentSaaS commoditization
Social Twitter May 10

Most of what's now called "agentic AI" — language models that browse, click, search, and take action...

By @burkov

62 score
AI Analysis

Burkov discusses the ICLR 2023 ReAct paper from Google Research that introduced the thought-action-observation loop, arguing it's the foundation of modern agentic AI. Notes Google invented both self-attention and the agentic loop but OpenAI drove productization.

Most of what's now called "agentic AI" — language models that browse, click, search, and take actions in an environment — traces its lineage back to this ICLR 2023 paper from Google Research. Rather than have a language model either reason silently and produce an answer (chain-of-thought prompting, where the model writes out its thinking before the final response) or just emit actions in an environment, the authors proposed to interleave the two. The model writes a thought, takes an action, se
agentic-aiai-historygoogle-researchchain-of-thought
Social Twitter May 10

If you wanna switch to @Cloudflare Email Sending today, here's my prompt for you, as always I'm unaf...

By @levelsio

62 score
AI Analysis

Levelsio shares a comprehensive AI-generated prompt for migrating transactional email to Cloudflare Email Service, covering prerequisites, domain reputation strategy, audit steps, helper function creation, phased migration, and important caveats about the beta service.

If you wanna switch to @Cloudflare Email Sending today, here's my prompt for you, as always I'm unaffiliated, not paid, not sponsored, but I like it, make sure you remove the space before the .com in the API url I added to avoid it becoming a link in this tweet: # Prompt: Migrate transactional email to Cloudflare Email Service Paste this into Claude Code (or Cursor, or any agent) running inside your project. --- I want to migrate this codebase's outbound email from its current provider (Post
AI-assisted developmentCloudflare emailDeveloper toolingEmail infrastructureAI coding agents
Social Twitter May 10

Agentic Search for Context Engineering @helloiamleonie's hot take: context engineering is about 80%...

By @aiDotEngineer

62 score
AI Analysis

AI.Engineer account shares a workshop on 'Agentic Search for Context Engineering' - Leonie's take that context engineering is 80% agentic search. Covers tool landscape: shell tools, semantic search, query execution, agent skills, and low floor vs high ceiling tool design.

Agentic Search for Context Engineering @helloiamleonie's hot take: context engineering is about 80% agentic search. The arrow from context sources to context window is doing most of the work. t.co/Iu2nkU2NdZ The workshop covers the full tool landscape: shell tools, semantic search, general-purpose query execution, agent skills, and when each one breaks. Including the part where an agent fakes semantic search by chaining grep synonyms. Low floor tools (specialized, easy to call correc
context_engineeringagentic_searchai_agentsai_toolingprompt_engineering
Social Twitter May 10

The slowest part of running an LLM just got 6x faster without losing a single token. LLMs generate ...

By @AlphaSignalAI

58 score
AI Analysis

Related: Reddit featured a BeeLlama.cpp fork with DFlash yesterday, DFlash: A lightweight block diffusion model used as a speculative decoding drafter achieves 6x lossless LLM acceleration, outperforming EAGLE-3 by 2.5x. Works with vLLM, SGLang, MLX and supports Qwen, LLaMA, Gemma.

The slowest part of running an LLM just got 6x faster without losing a single token. LLMs generate text one token at a time. That sequential bottleneck wastes GPU power and slows everything down. Speculative decoding fixes part of this. A small draft model guesses ahead, then the big model verifies in parallel. The catch: most drafters are still sequential, so the speedup is capped. DFlash flips the approach. It is a lightweight block diffusion model used as the drafter, predicting
llm-inferencespeculative-decodingdiffusion-modelsopen-source
Social Twitter May 10

Just tried this in my codebase, burned ~$70 worth of tokens and resulted in 30+ PRs, all non-critica...

By @tdinh_me

58 score
AI Analysis

Tony Dinh reports spending ~$70 in AI tokens to run an automated security audit on his codebase, resulting in 30+ PRs with legitimate but non-critical security issues.

Just tried this in my codebase, burned ~$70 worth of tokens and resulted in 30+ PRs, all non-critical but totally legit security issues.
AI code securityAI-assisted developmentDeveloper productivityAI cost-benefit
Social Twitter May 10

The personification of Claude — in name (the only AI with a human one), in training, in Anthropic’s ...

By @emollick

55 score
AI Analysis

Ethan Mollick observes that Anthropic's personification of Claude (human name, constitution, fanfiction/cartoons) feels consequential for the medium term.

The personification of Claude — in name (the only AI with a human one), in training, in Anthropic’s philosophy (see Claude Constitution), in fanfiction (see the Claude cartoons), etc — feels quite consequential in the medium term, for better and for worse.
anthropicai_anthropomorphismai_brandingclaude
Social Twitter May 10

Meta's AI can now verify code patches with 93% accuracy without running them. LLMs read code the w...

By @AlphaSignalAI

55 score
AI Analysis

Meta's AI can verify code patches with 93% accuracy without execution. A structured template forcing models to trace execution paths and prove conclusions improved patch verification from 78% to 93%. Opens path to execution-free feedback for RL pipelines.

Meta's AI can now verify code patches with 93% accuracy without running them. LLMs read code the way you skim a book. They recognize familiar words and fill in the rest from memory. A new paper tests whether AI agents can verify code changes without running the software. The results expose how models actually fail. When comparing two patches, a standard model saw a common function name and assumed it meant the usual system tool. It never checked local files. It missed that the proj
code-verificationmeta-aistructured-promptingsoftware-engineering