Wiredframe Radar logo

Category intelligence

Social Media Briefing — April 10, 2026

172 current items analyzed and ranked.

Executive synthesis

Social Media Summary

Andrej Karpathy dominated discourse with a viral analysis (1.7M views) of the growing gap between casual AI users and those using frontier agentic tools like Codex and Claude Code, arguing most people drastically underestimate current AI capability.

  • Ethan Mollick provided a comprehensive frontier landscape assessment: Google, OpenAI, and Anthropic lead with possible RSI signs; xAI has fallen from the frontier; Chinese labs trail 7–9 months behind. He also flagged that all US frontier labs have abandoned open weights.
  • Sam Altman announced a new $100/month ChatGPT Pro tier driven by massive Codex demand, positioned between Plus ($20) and the existing $200 Pro — a major pricing restructure.
  • Yann LeCun revealed he *left* Meta (was not fired) largely over excessive LLM emphasis, distancing himself from Llama entirely. He also dismissed Anthropic's Mythos announcement as "BS from self-delusion."
  • Mythos skepticism was widespread: Gary Marcus called it overblown (sandboxing off, no evidence of recursive self-improvement), while Karpathy noted even the Mythos PDF itself defeated document converters. Perplexity announced Plaid financial account integration, signaling AI assistants expanding into personal finance.

Key Themes

AI Capability Gap Discourse · 8AI Landscape Assessment · 9OpenAI Codex Pricing & Tiers · 7Mythos Skepticism · 10LeCun's Meta Departure & LLM Skepticism · 7OpenAI Business & Product · 5Perplexity Finance Integration · 2Open Weights Future · 5AI Inference Infrastructure & Benchmarking · 2Gemma 4 Adoption & Ecosystem · 5

Primary evidence

Top Ranked Signals

Social Twitter Apr 9

Judging by my tl there is a growing gap in understanding of AI capability. The first issue I think ...

By @karpathy

95 score
AI Analysis

Karpathy writes an extensive analysis of the growing gap in understanding of AI capability. He identifies two groups: those who tried free/old ChatGPT and dismiss AI, and those using frontier agentic models (Codex/Claude Code) professionally who are experiencing 'AI Psychosis' from staggering improvements. He explains this gap through reinforcement learning with verifiable rewards and B2B economic incentives.

Judging by my tl there is a growing gap in understanding of AI capability. The first issue I think is around recency and tier of use. I think a lot of people tried the free tier of ChatGPT somewhere last year and allowed it to inform their views on AI a little too much. This is a group of reactions laughing at various quirks of the models, hallucinations, etc. Yes I also saw the viral videos of OpenAI's Advanced Voice mode fumbling simple queries like "should I drive or walk to the carwash". T
ai_capabilitiesagentic_aiai_discourse_gapreinforcement_learningcodexclaude_codeai_psychosispublic_perception
Social Twitter Apr 9

It is very nice to see Codex getting so much love. We are launching a $100 ChatGPT Pro tier by very ...

By @sama

90 score
AI Analysis

Building on yesterday's Social Codex buzz, Sam Altman announces OpenAI is launching a $100 ChatGPT Pro tier, noting Codex is getting 'so much love' and the new tier is by 'very popular demand'.

It is very nice to see Codex getting so much love. We are launching a $100 ChatGPT Pro tier by very popular demand.
openai_productai_pricingcodexai_business_model
Social Twitter Apr 9

So we now have a pretty good picture of the state of the frontier AI model makers. US closed source...

By @emollick

88 score
AI Analysis

Mollick provides a comprehensive state-of-the-art assessment: US closed source (Google, OpenAI, Anthropic) leads with possible RSI signs; xAI has fallen from frontier; Meta re-entered with a not-quite-frontier model; Chinese labs (Qwen, Kimi, MiniMax, Xiaomi, DeepSeek, GLM) are 7-9+ months behind; Mistral has fallen from frontier.

So we now have a pretty good picture of the state of the frontier AI model makers. US closed source models continue to lead. Google, OpenAI, and Anthropic stand well ahead of the pack, and may have signs of recursive self-improvement. xAI has fallen from frontier status for now (though promises to return shortly). Meta re-entered the space today with a not-quite-frontier closed source model, but an approach that suggests that they might be back in the race. All the other US players seem far beh
ai_landscapefrontier_modelsrecursive_self_improvementchinese_aimeta_aiopen_weightsai_geopolitics
Social Twitter Apr 9

We’re updating our ChatGPT Pro and Plus subscriptions to better support the growing use of Codex. W...

By @OpenAI

88 score
AI Analysis

OpenAI introduces a new $100/month Pro tier positioned between Plus ($20) and existing Pro ($200), offering 5x more Codex usage than Plus, with a launch promo of 10x Plus usage through May 31st.

We’re updating our ChatGPT Pro and Plus subscriptions to better support the growing use of Codex. We’re introducing a new $100/month Pro tier. This new tier offers 5x more Codex usage than Plus and is best for longer, high-effort Codex sessions. In ChatGPT, this new Pro tier still offers access to all Pro features, including the exclusive Pro model and unlimited access to Instant and Thinking models. To celebrate the launch, we’re increasing Codex usage for a limited time through May 31st so
openai_pricingcodexsubscription_tiersproduct_launch
Social Twitter Apr 9

@alz_zyd_ I was never involved in Llama. Except to promote the open sourcing of Llama-2. Since Lla...

By @ylecun

80 score
AI Analysis

LeCun clarifies he was never involved in Llama (except promoting open-sourcing Llama-2), was at FAIR focused on long-term research, and had no technical interactions with the Llama team.

@alz_zyd_ I was never involved in Llama. Except to promote the open sourcing of Llama-2. Since Llama-2, the project was in the hands of the GenAI organization. I was at FAIR (a different org focused on long-term research) and had no technical interactions with the Llama team.
lecun_departuremeta_aillamaai_industry_dynamics
Social Twitter Apr 9

Computer now connects with Plaid to link bank accounts, credit cards, and loans. Track spending in ...

By @perplexity_ai

75 score
AI Analysis

Perplexity AI announces integration with Plaid to connect bank accounts, credit cards, and loans, enabling spending tracking, custom budget tools, and net worth visualization alongside investment portfolios.

Computer now connects with Plaid to link bank accounts, credit cards, and loans. Track spending in detail, build custom budget tools, and visualize your net worth alongside your investment portfolio. t.co/m9nws4VjKO
perplexityai_financefintechproduct_launchplaid_integration
Social Twitter Apr 9

Someone recently suggested to me that the reason OpenClaw moment was so big is because it's the firs...

By @karpathy

72 score
AI Analysis

Karpathy discusses how 'OpenClaw' was a breakthrough moment because it was the first time a large group of non-technical people experienced the latest agentic AI models, having previously only known AI through ChatGPT as a website.

Someone recently suggested to me that the reason OpenClaw moment was so big is because it's the first time a large group of non-technical people (who otherwise only knew AI as synonymous with ChatGPT as a website) experienced the latest agentic models.
ai_adoptionagentic_aipublic_perceptionopenclaw
Social Twitter Apr 9

Yesterday’s Mythos announcement from Anthropic was overblown. • Sandboxing was turned off, so test...

By @GaryMarcus

72 score
AI Analysis

Following yesterday's Social analysis by Delangue, Marcus argues Anthropic's Mythos announcement was overblown: sandboxing was off, cheap open-weight models can do similar things, and there's no evidence of a major qualitative jump. Tags journalists Friedman and Farrow.

Yesterday’s Mythos announcement from Anthropic was overblown. • Sandboxing was turned off, so test didn’t show much about the real world. • Cheap open-weight models can (already) do some similar stuff • No evidence that Mythos itself is a major qualitative jump. In short, we got played. cc @tomfriedman @RonanFarrow
mythos_skepticismanthropicai_hypeai_safetybenchmark_limitations
Social Twitter Apr 9

The US frontier labs have all walked away from open weights. They continue to occasionally release e...

By @emollick

72 score
AI Analysis

Mollick observes that US frontier labs have all walked away from open weights, only occasionally releasing smaller non-competitive open models (like Gemma 4). Says all eyes are on Chinese AI labs for open models.

The US frontier labs have all walked away from open weights. They continue to occasionally release excellent open models (Gemma 4, etc), but they are smaller models that are not competitive with their closed weights models. So all eyes are on Chinese AI labs for open models.
open_weightsai_geopoliticschinese_aius_ai_labs
Social Twitter Apr 9

Our existing $200 Pro tier still remains our highest usage option. And as a thank you to our existin...

By @OpenAI

72 score
AI Analysis

OpenAI announces the existing $200 Pro tier remains the highest option, extends 2x Codex usage promo until May 31st, and resets Codex rate limits for $200 Pro users.

Our existing $200 Pro tier still remains our highest usage option. And as a thank you to our existing Pro users on the $200 tier, we’re extending our 2x Codex usage promo (until May 31st) and we’ve reset your Codex rate limits (yes, again).
openai_pricingcodexsubscription_tiers