It was an honor to hang out with Jensen Huang, CEO of
@nvidia
, and do a long-form podcast with him. Really fun & fascinating technical deep-dive conversation on & off the mic. One of the most brilliant & thoughtful human beings I've ever met. NVIDIA is the most valuable company
Anthropic releases Claude 3.5 Sonnet, their most capable model yet. It outperforms GPT-4o and Gemini 1.5 Pro on multiple benchmarks while being faster and more cost-effective than Claude 3 Opus.
GPT-4o brings native multimodal capabilities to ChatGPT, enabling real-time voice conversations, image understanding, and code interpretation in a single model.
An anonymous Substack post accuses compliance startup Delve of “falsely” convincing “hundreds of customers they were compliant” with privacy and security regulations.
Grok, how do you make chips? | Image: Laura Normand / The Verge
Elon Musk announced plans to build a Terafab plant in Austin, Texas, that will be jointly run by Tesla and SpaceX. The goal is to eventually build chips at scale for robotics, artificial intelligence, and space-based data centers for Musk's various companies.
Musk has expressed concern, as have other executives, about the chip industry's ability to keep up with demand as the AI industry has boomed. But building a chip fabricati...
AI was everywhere at the GDC Festival of Gaming this year. Vendors at the event pitched generative AI tools for things like making AI-driven NPCs and even entire games from a chat box. On the show floor, I spent 10 minutes playing a demo of a pixel-art fantasy world generated by Tencent's AI tools. In a briefing with Razer, I watched an AI assistant for QA automatically log issues in a shooter game. And there were many talks about AI, including a standing-room only presentation by Google Deep...
@GroqInc on meta & llama. Just sayin, DeepSeek-R1-Distill-Llama-70b is available on GroqCloud™ Tag us on socials with your projects and apps and we'll send Groq swag to some of our favorites. The new #Llama 3.1 models by @AIatMeta are running on Groq! Our 300K+ developer community can access via GroqCloud Dev Console and the models are available to the public on GroqChat. Read the full p
@GroqInc on ai hardware & chips. 1 week ago: Kimi K2 launches 72 hours later: We YOLO launch it on Groq Now: Thousands of devs are building with it Kimi K2, now on Groq: 1T parameters Full context Built for agents Unmatched pric You can now use Groq’s Kimi K2 to build super realistic speech-to-speech voice interfaces with emotion via @hume_ai 's new EVI 3. Perfect for AI companions, phone calls, or anything you’re working o
@polynoamial on ai benchmarks. There's a lot of talk of LLMs "saturating all the evals" but there's plenty of evals people could make where LLMs would do poorly: -Beat a Zelda game -Make a profit in a prediction market -Write a sta
@ARKinvest on xai & grok. In this week's newsletter What Will Prasad’s Departure From The FDA Mean For Drug Development? Tesla And xAI Join Forces To Develop Digital Optimus Air Taxi Operations Are Beginning To Take Shape xAI is charting a course for alternative compute sources what will unlock unique scaling vectors relative to the competition. An effort that's success will hinge on the team's ability to build a capab
9 sources discuss ai news & insights, including @egeberkina, @nickfloats, @DataCamp, @freeCodeCamp, @polynoamial. A blueprint breakdown of iPhone, Rolex, and Ray-Ban where engineering meets timeless design. --sref 3982351386 You asked, I delivered. Here’s the JSON code for a transparent background! Just run it like before and you’ll get the same clean results. Here's the prompt and more examples below
3 sources discuss openai & gpt, including @egeberkina, @polynoamial, @cerebras. Turning icons into iridescent legends with GPT-4o + JSON Just saw probably the coolest Halloween film made with AI today, insane stuff It was created by Simon Meyer, who broke down the whole process: Veo3 for the core scenes, Nano Banana & Higgsfield Pop
2 sources discuss ai image & video gen, including @egeberkina, @nickfloats. Since Midjourney announced that they support creating videos with specific start and end frames, I had to try it. So I retextured the images with the JSON below and then used start and end frames to Midjourney Day 1 Day 637