AI and I

Prompt engineering matters more than ever. But it’s evolving into something totally new:  A way for non-technical domain experts to solve complex problems with AI. I spent an hour talking to prompt wizard Jared Zoneraich, cofounder and CEO of PromptLayer, about why the death of prompt engineering is greatly exaggerated. And why the future of prompting is equipping non-technical experts with the tools to manage, deploy, and evaluate prompts quickly. We get into: His theory around why the “irreducible” nature of problems will keep prompt engineering relevant Prompt engineering best practices around prompts, evals, and datasets Why it’s important to align your prompts with the language the model speaks How to run evals when you don’t have ground truth Why he believes that the companies who have domain experts to scope out the right problems will win in the age of gen AI This is a must-watch for prompt engineers, people interested in building with AI systems, or anyone who wants to generate predictably good responses from LLMs. If you found this episode interesting, please like, subscribe, comment, and share!  Want even more? Sign up for Every to unlock our ultimate guide to prompting ChatGPT here: https://every.ck.page/ultimate-guide-to-prompting-chatgpt. It’s usually only for paying subscribers, but you can get it here for free. To hear more from Dan Shipper: Subscribe to Every: https://every.to/subscribe  Follow him on X: https://twitter.com/danshipper  Timestamps: Introduction: 00:01:08 Jared’s hot AGI take: 00:09:54 An inside look at how PromptLayer works: 00:11:49  How AI startups can build defensibility by working with domain experts: 00:15:44 Everything Jared has learned about prompt engineering: 00:25:39 Best practices for evals: 00:29:46 Jared’s take on o-1: 00:32:42 How AI is enabling custom software just for you: 00:39:07 The gnarliest prompt Jared has ever run into: 00:42:02 Who the next generation of non-technical prompt engineers are: 00:46:39 Links to resources mentioned in the episode:  Jared Zoneraich: @imjaredz PromptLayer: @promptlayer, https://www.promptlayer.com/ A couple of Steven Wolfram’s articles on ChatGPT: What Is ChatGPT Doing … and Why Does It Work?, ChatGPT Gets Its “Wolfram Superpowers”!   

This episode is sponsored by Notion. I’ve been using Notion to manage my professional and personal life for almost 10 years. As a company, they pay attention to the craft and ideas underlying the software they build, and that comes through in the experience of using Notion every day. If you’re a startup, get up to 6 months of Notion Plus with unlimited AI—worth up to $6,000—for free by going to https://ntn.so/every, selecting Every in the drop-down partner list, and using the code EveryXNotion. Notion cofounder Simon Last told me everything he’s learned from integrating AI into a platform that has over 100 million users. Simon likes to keep a low profile, even though he’s the driving force behind Notion AI, one of the most widely scaled AI applications in the world. In his first-ever podcast interview, we get into: What he would build if he started Notion from scratch today with AI How to get high quality and reliable results from AI at scale The future of human creativity in a world with machines that think   This is a must-watch for anyone interested in building reliable AI products at scale. If you found this episode interesting, please like, subscribe, comment, and share!  Want even more? Sign up for Every to unlock our ultimate guide to prompting ChatGPT here: https://every.ck.page/ultimate-guide-to-prompting-chatgpt. It’s usually only for paying subscribers, but you can get it here for free. To hear more from Dan Shipper: Subscribe to Every: https://every.to/subscribe  Follow him on X: https://twitter.com/danshipper Timestamps: Introduction: 00:01:57 How AI changes the way we build the foundational elements of software: 00:02:28 Simon’s take on the impact of AI on data structures: 00:10:07 The way Simon would rebuild Notion with AI: 00:13:05 How to design good interfaces for LLMs: 00:23:39 An inside look at how Notion ships reliable AI systems at scale: 00:28:22 The tools Simon uses to code: 00:35:41 Simon’s thoughts on scaling inference compute as a new paradigm: 00:38:16 How the growing capabilities of AI will redefine human roles: 00:49:10 Simon’s AGI timeline: 00:50:28 Links to resources mentioned in the episode: Simon Last: @simonlast Notion AI: https://www.notion.so/product/ai The AI code editor Simon uses: Cursor OpenAI’s definition of AGI that Simon ascribes to: https://openai.com/charter/

Union Square Ventures is building an AI operating system to support their investment team.  But it’s not what you think: It’s a constellation of AI tools that captures and synthesizes the firm's collective wisdom. It’s evolving every day, and Matt Cynamon is the mad scientist in charge Matt calls himself a “regular” at USV. In practice that means he’s responsible for running experiments with AI for the firm. As an inherently curious person with the professional obligation to tinker, he’s built a suite of tools for the firm, including:  The Librarian, a chatbot trained on around 15,000 articles from USV’s blog Portfolio Tracker, a GPT that analyzes the investments made by the firm Meeting Notes, a tool that makes it possible for team members to interact with meetings   I sat down with Matt to talk about how AI is enabling him to bring his ideas to life as a generalist, get demos of the tools listed above, and exchange notes on all the other projects he has in the works at USV. We edit actionable insights extracted by an AI from meetings at USV and prepare them to be posted on the firm’s X handle live on the show. We even try out an art project at USV’s office called The Dream Machine, which generates art from conversations. Here’s a link to the episode transcript.    This is a must-watch for anyone interested in riding the AI wave by learning how to ship useful products quickly. If you found this episode interesting, please like, subscribe, comment, and share!  Want even more? Sign up for Every to unlock our ultimate guide to prompting ChatGPT here: https://every.ck.page/ultimate-guide-to-prompting-chatgpt. It’s usually only for paying subscribers, but you can get it here for free. To hear more from Dan Shipper: Subscribe to Every: https://every.to/subscribe  Follow him on X: ⁠https://twitter.com/danshipper⁠  Timestamps: Introduction: (00:00:52) How Matt became in charge of everything AI at USV: (00:01:56) How AI empowers generalists to be creators: (00:06:22) The Librarian, a chatbot trained on everything USV has published: (00:10:41) Portfolio Tracker, an AI tool to track USV’s investments: (00:21:09) The AI projects that Matt has in the pipeline at USV: (00:27:21) Meeting Notes, USV’s AI note-taking tool: (00:34:33) Prompting AI to generate a post for USV’s X handle: (00:44:57) Why it’s important to diversify ownership over data: (01:00:20) The Dream Machine, AI that generates images from conversations: (01:03:20) Links to resources mentioned in the episode: Matt Cynamon: @mattcynamon Union Square Ventures: @usv, https://www.usv.com/  More about the AI tools at USV: https://www.usv.com/people/the-librarian/, https://www.usv.com/writing/2024/02/ai-aesthetics/  The X post generated live on the show: https://x.com/usv/status/1847354782941663523

Yohei Nakajima leads a double life.  By day, he’s a general partner of a small venture firm, Untapped Capital.  By night, he’s one of the most prolific internet tinkerers in AI. (He also sometimes works on automating his job as a venture capitalist.) He’s the creator of BabyAGI (@babyAGI_), the first open-source autonomous agent that went viral in March 2023. Yohei has since released seven iterations of BabyAGI (each one named after a different animal), a coding agent called Ditto, a framework for building autonomous agents, and, most recently, BabyAGI 2o, a self-building autonomous agent (that follows OpenAI’s unfortunate naming convention). Even more incredible, Yohei isn’t a professional developer. His day job is as the general partner of Untapped Capital (@UntappedVC). I sat down with Yohei to talk about: What feeds Yohei’s drive to create new tools The evolution of BabyAGI into a more powerful version of itself  What Yohei learned about himself by tinkering on the internet Yohei’s personal philosophy about how the tools we build our extensions of ourselves Why founders in AI should think about their products from a modular lens, by addressing immediate problems while enabling growth in the future Yohei’s insight into a future where models will train themselves as you use them We experiment with Ditto live on the show, using the tool to build a game of Snake and a handy scheduling app. Yohei also screenshares a demo of BabyAGI 2o in action. This is a must-watch for anyone curious about autonomous agents, building cool AI tools on the internet, and the future of AI tooling. If you found this episode interesting, please like, subscribe, comment, and share!  Want even more? Sign up for Every to unlock our ultimate guide to prompting ChatGPT here: https://every.ck.page/ultimate-guide-to-prompting-chatgpt. It’s usually only for paying subscribers, but you can get it here for free. To hear more from Dan Shipper: Subscribe to Every: ⁠https://every.to/subscribe⁠  Follow him on X: ⁠https://twitter.com/danshipper⁠  Timestamps: Introduction: (00:00:59) BabyAGI and its evolution into a more powerful tool: (00:02:26) How better models are changing the way Yohei builds: (00:05:00) Using code building agent Ditto to build a game of Snake: (00:08:10) The ins and outs of how Ditto works: (00:13:24) How Yohei gets a lot done in little time: (00:19:21) Yohei’s personal philosophy around building AI tools: (00:21:50) How Yohei experiments with AI as a tech-forward parent: (00:33:13) Demo of Yohei’s latest release, BabyAGI 2.0: (00:39:29) Yohei’s insights on the future of AI tooling: (00:51:24) Links to resources mentioned in the episode:  Yohei Nakajima: @yoheinakajima, http://yohei.me  Untapped Capital: @UntappedVC, https://www.untapped.vc/  My first interview with Yohei, around the time he released BabyAGI: https://every.to/chain-of-thought/this-vc-is-slowly-automating-their-job  The other AI tools Yohei has created: Ditto, BabyAGI 2, BabyAGI 2o The tweet thread about AI bots being let loose on a Discord server: https://x.com/AISafetyMemes/status/1847312782049333701