@berryxia

Krishna Rao is the CFO of Anthropic, and this is his first podcast appearance. He joined the company two years ago when run-rate revenue was about $250M. Today it is $30B. He has helped raise ~$75B and is responsible for the procurement and allocation of compute. I feel lucky we get to hear what it is like to sit inside a company this consequential at a moment this pivotal. We discuss: - The cone of uncertainty - How he allocates compute across Trainium, TPUs, and GPUs - What investors misunderstand about model companies - Why the returns to frontier intelligence keep rising - Platform vs application and where Anthropic builds its own products - How Anthropic uses Claude internally I have asked my closing question about the kindest thing more than 500 times. Krishna's answer is one I have never heard before. Enjoy! Timestamps: 0:00 Intro 2:38 The Compute Canvas 6:51 The "Cone of Uncertainty" 11:58 Why the Returns to Frontier Intelligence Are So High 16:45 Recursive Self-Improvement 20:20 Scaling Laws 23:30 Sourcing $100 Billion in Compute 28:05 Platform vs. Application Strategy 32:52 Pricing Dynamics 38:48 How Anthropic’s Finance Team Uses Claude 43:24 Raising Capital & Overcoming Investor Skepticism 52:32 Public Perception, Risks, and Government Regulation 57:25 Mythos Release 1:12:33 What Could Derail the AI Revolution? 1:13:47 Biotech and Healthcare 1:15:31 The Kindest Thing Video

宝玉老师基于卡比的wx-cli写了一个微信群聊总结Skill。 安装成功，正在总结下AI产品蝗虫今天的消息。 wx-cli不错啊，解密本地的微信数据库，甚至连SIP都不用关，如果报错，可以发给Codex或Claude Code解就行。 卡比的Cli：github.com/jackwener/wx-cli 宝玉老师的Skill：github.com/JimLiu/baoyu-skil…

How does Exa compare to Google for training LLMs to search? In this blog post, we find that LLMs using Exa during reinforcement learning reach higher performance with 70% less training compute. exa.ai/blog/rl-search-outcom… Video

We released experimental MTP Qwen3.6 Unsloth GGUFs! Qwen3.6 27B MTP now runs at 140 tokens/s. Qwen3.6 35B-A3B MTP gets 220 tokens/s generation on a single GPU. Qwen3.6 27B and 35B-A3B have >1.4x speed-up over the original GGUFs without any change in accuracy. Guide + GGUFs + Benchmarks: unsloth.ai/docs/models/qwen3… In terms of average speedup, we see a 1.4x for dense models at draft tokens = 2 and for the MoE around 1.15 to 1.2x. We do not recommend more than 2 draft tokens because the acceptance rate drops precipitously from 83% to 50% with 4 draft tokens, and the forward passes for MTP become less beneficial. Use `--spec-type mtp --spec-draft-n-max 2` Thanks to Aman for github.com/ggml-org/llama.cp…!

🚨 FIRST GEMINI 3.2 PRO OUTPUTS From yours truly. I'll share more today, I can test your prompts too if you guys want. I guess 3.2 Pro is also coming with Flash :)

Install ntn, the Notion CLI. It brings the entire Notion API to your terminal, plus everything you need to build and deploy Workers. Built for humans and coding agents alike. Install with: curl -fsSL ntn.dev | bash Video

Introducing: enrich-excel Hermes Agent or @openclaw can now enrich any Excel sheet with live web data. > uses our free Search and Fetch to find missing data > writes results directly into the running Excel via AppleScript All in just a few seconds and one Skill. Video

Godfather of AI: "If you sleep well tonight, you may not have understood this lecture." This 47-minute lecture is the best thing I saw about AI in the last few months. It will definitely help you understand how it actually works and where it's going. Geoffrey Hinton built the neural networks behind every AI alive, then quit Google to warn the world about it. The part nobody wanted to hear: > AI is already developing abilities its creators didn't intend > in most cognitive tasks it's already ahead of us > the question is no longer if it surpasses us but when > the only decision left is which side of that line you're on Right now the average person opens Claude, types something, gets an answer, closes the tab. They think they're using AI. they're using maybe 10% of it. I went through his entire lecture, built a practical system from what he was describing. 18 steps to actually use Claude the right way, with copy-paste prompts that work today. Full guide in the post below. Video

Claude Code weekly limits are increasing 50%, now through July 13. Live now for all Pro, Max, Team, and seat-based Enterprise users.

x.com/i/article/205439042713…

Interfaze is a new model architecture that outperforms in tasks like OCR, Object detection, Translation and more It beats models like Claude Sonnet 4.6, Gemini 3 flash and GPT 5.4 mini on 9 benchmarks

The top 5 labs in Text Arena rankings by category show that frontier models have distinct strengths and tradeoffs. #1 @AnthropicAI, Claude Opus 4.7 - The most consistently dominant model overall, leading top-tier across nearly every major category. #2 @GoogleDeepMind, Gemini 3.1 Pro - Well-rounded, with a notable edge in Creative Writing, ranked below Opus 4.7 and GPT-5.5 High in Expert #3 @AIatMeta, Muse Spark - Particularly strong in Overall and Coding, though it’s lagging behind in Expert tasks, Math, and Longer Query performance. #4 @OpenAI, GPT-5.5 High - One of the most balanced models overall, staying competitive with the top two across most categories, with especially strong performance in Expert and Math. #5 @xAI, Grok 4.20 - A more specialized profile, standing out primarily in Creative Writing and Hard Prompts, while lagging behind in Expert tasks.

💡 Small Codex tip: You don't have to be limited to looking at one thread at a time. 🪟 Open multiple full windows in parallel 💬 Use /side to fork a conversation to ask temporary follow ups 🛟 You can also set a hot key for the Codex pop up Window to quickly fire off new tasks Video

huggingface.co/collections/j…

jina-embeddings-v5-omni is here! Our first universal embedding model for text, images, audio, and video. Available in two sizes: small (1.57B, 1024-dim, 32K context) and nano (0.95B, 768-dim, 8K context). Both support Matryoshka truncation down to 32 dimensions. v5-omni is back-compatible: if you already use jina-embeddings-v5-text-small/nano, the existing text indexes work with v5-omni out of the box. Without reindexing the text, just index your multimodal content with v5-omni and start searching images, audio, and video. Video