{"url": "https://www.youtube.com/watch?v=sJcUbMiwrS4", "page_url": "https://www.youtube.com/watch?v=sJcUbMiwrS4", "title": "Anthropic - Daniela Amodei | Founders You Should Know", "show_or_channel": "Founders You Should Know", "date": null, "duration_s": 279, "type": "talk", "speakers": ["Daniela Amodei"], "confidence": "medium", "notes": "About 5-min recruiting pitch, likely 2022-2023. Short, but no longer version found.", "people": ["daniela-amodei"], "id": "0000_founders-you-should-know-anthropic-daniela-amodei-founders-yo", "audio_duration_s": 278.3146875, "worker": "jc-w09", "transcribe_s": 12.6, "model": "large-v3", "segments": [{"start": 0.0, "end": 28.75, "text": "Okay. Hi, everybody. Thank you so much for having me here today at Founders You Should Know. I'm Daniela. I'm really excited to talk with all of you about Anthropic, the company that I co-founded and I'm president of. So Anthropic is really seeking to answer this kind of fundamental question, which is, can we build and develop transformative AI systems that are truly safe for humans to use? And before I tell you about how we're doing that, I just wanted to share a little bit about my background."}, {"start": 28.85, "end": 57.62, "text": "So I originally started my career working in global health. I spent time living abroad. I worked in politics. And then I moved on to work at Stripe just about 10 years ago, growing teams from around 40 to 1,200 people, and then went and did something similar, growing and managing teams at OpenAI. And then I co-founded Anthropic just about two years ago with six co-founders. So you can see my codependency knows no bounds. So Anthropic's mission is to build reliable, steerable, and interpretable AI systems."}, {"start": 58.22, "end": 87.09, "text": "And we really are at the cutting edge of developing these kind of frontier generative AI models, particularly in language, but across multiple modalities. Our founders are probably best known for leading the development of GPT-3 and more recently for developing our helpful, honest, and harmless framework for training AI systems. We've published more than 15 safety papers ranging from interpretability to reinforcement learning from human feedback and much more."}, {"start": 87.09, "end": 101.01, "text": "And most recently, we've trained Claude, who's our new dialogue agent and are preparing to take him to market. So if you've had the chance to play around with chat GPT, or you're just very active on AI Twitter, which..."}, {"start": 101.01, "end": 126.64, "text": "You might have seen just a little bit of comparison sort of between those two systems. And the general consensus is that Claude is just a little bit better on a lot of dimensions that customers care about compared to ChatGPT. And so here, this is things like humor and general kind of engagement with the user, interactiveness, friendliness, and all of the safety metrics that I just kind of talked about. So being helpful to the person interacting, harmless, and honest in his answers."}, {"start": 127.25, "end": 154.3, "text": "So our momentum is very strong. We have a number of customers that are now exploring us in private beta. As of about two weeks ago, we had 200 just organic inbound interest customers. And so this is basically, we had no website, we've done no product branding. We just had 200 people kind of come to us and say, hey, we'd love to use Claude for our business. And then we put up a wait list and got around 2,500 users just in the past two weeks that are interested in getting access to Claude."}, {"start": 154.3, "end": 175.31, "text": "So there are a lot of possible use cases that Claude can be used for. These range from things like search to code generation, customer support, content moderation, summary and retrieval, and much more. In terms of fundraising, we've raised just about a billion dollars across our series A and B, and we're actually in active talks for our series C as well."}, {"start": 175.92, "end": 196.4, "text": "In terms of where we are, great opportunity right now. People really want to use Claude. We have way more demand than we currently have capacity to absorb. It's kind of an unusual problem. It's a good problem, but it's still a problem. And what that means for everyone in this room is that we're very actively hiring across a bunch of different roles in the company, really across the stack."}, {"start": 197.2, "end": 224.43, "text": "So just a tiny bit more about us. We're about 100 people right now. We have a beautiful office in Jackson Square full of plants. I think we're like a two to one plant to human ratio. And we're a hybrid in office and remote culture. So we are hiring on all cylinders. On the engineering side, we're looking for product engineers, for people with systems experience, always for research engineers, really just hiring for engineering across the full stack."}, {"start": 225.17, "end": 251.38, "text": "On the security side, we just closed a CISO, very exciting, and we're actually looking to hire people here from all different levels of experience. Security is a major focus for us, so if you have a security interest, please come talk to me. On the business side, we're also growing really, really quickly. We are looking for account executives and AMs, sales, business operations, trust and safety, user ops, really all the way from individual contributor up through director."}, {"start": 252.27, "end": 278.31, "text": "On the product side, we're looking for communications, designers in particular, marketing, product managers, and more. And then in operations, we're hiring recruiters, people ops, legal, HR, finance. We're hiring everybody. If you are just like interested in AI, if you like plants, come talk to me. There's probably a role that I have open for you. And in general, just thank you so much for having me. I'm really looking forward to talking with all of you."}]}