Inside OpenAI | Logan Kilpatrick (head of developer relations)

Brought to you by:

• Hex—Helping teams ask and answer data questions by working together
• Whimsical—The iterative product workspace
• Arcade Software—Create effortlessly beautiful demos in minutes

—

Logan Kilpatrick leads developer relations at OpenAI, supporting developers building with the OpenAI API and ChatGPT. He is also on the board of directors at NumFOCUS, the nonprofit organization that supports open source projects like Jupyter, Pandas, NumPy, and more. Before OpenAI, Logan was a machine-learning engineer at Apple and advised NASA on open source policy. In our conversation, we discuss:

• OpenAI’s fast-paced and innovative work environment

• The value of high agency and high urgency in your employees

• Tips for writing better ChatGPT prompts

• How the GPT Store is doing

• OpenAI’s planning process and decision-making criteria

• Where OpenAI is heading in the next few years

• Insight into OpenAI’s B2B offerings

• Why Logan “measures in hundreds”

Some takeaways:

1. The key to OpenAI’s ability to move quickly and ship high-quality products lies in prioritizing individuals with high agency and a sense of urgency when building their teams. This approach allows OpenAI to bypass cumbersome processes, facilitating a culture where individuals can proactively identify and address challenges without waiting for extensive approvals or consensus.

2. To get better answers from ChatGPT, be specific and give more context to your questions. Instead of asking simple questions, provide details like you would when talking to someone. AI lacks built-in context, so guiding it with detailed prompts is crucial. This not only improves immediate results but also helps in developing smarter AI systems over time.

3. GPTs unlock the potential for an “agent future,” where we assign complex tasks to AI tools. Currently, we prompt GPTs, receive rapid responses, and conclude the interaction. As GPTs advance, we’ll be able to prompt them to engage in more nuanced tasks and ask them to report back when it’s done. For example, we might prompt the AI agent to take several hours to write a detailed blog post citing references and case studies, and to delineate any trade-offs the agent made.

4. Logan expects that AI won’t replace humans, but people who leverage AI to learn and improve productivity will be more competitive in the job market. Engineers are probably the biggest beneficiaries of GPTs because it allows them to ship faster.

Listen now on Apple, Spotify, Overcast, and YouTube.

Where to find Logan Kilpatrick:

• X: https://twitter.com/OfficialLoganK

• LinkedIn: https://www.linkedin.com/in/logankilpatrick/

• Website: https://logank.ai/

Where to find Lenny:

• Newsletter: https://www.lennysnewsletter.com

• X: https://twitter.com/lennysan

• LinkedIn: https://www.linkedin.com/in/lennyrachitsky/

In this episode, we cover:

(00:00) Logan’s background

(03:49) The impact of recent events on OpenAI’s team and culture

(08:20) Exciting developments in AI interfaces

(09:52) Using OpenAI tools to make companies more efficient

(13:04) Examples of using AI effectively

(18:35) Prompt engineering

(22:12) How to write better prompts

(26:05) The launch of GPTs and the OpenAI Store

(32:10) The importance of high agency and urgency

(34:35) OpenAI’s ability to move fast and ship high-quality products

(35:56) OpenAI’s planning process and decision-making criteria

(40:22) The importance of real-time communication

(42:33) OpenAI’s team and growth

(44:47) Future developments at OpenAI

(47:42) GPT-5 and building toward the future

(50:38) OpenAI’s enterprise offering and the value of sharing custom applications

(52:30) New updates and features from OpenAI

(55:09) How to leverage OpenAI’s technology in products

(58:26) Encouragement for building with AI

(59:30) Lightning round

Referenced:

• OpenAI: https://openai.com/

• Sam Altman on X: https://twitter.com/sama

• Greg Brockman on X: https://twitter.com/gdb

• tldraw: https://www.tldraw.com/

• Harvey: https://www.harvey.ai/

• Boost Your Productivity with Generative AI: https://hbr.org/2023/06/boost-your-productivity-with-generative-ai

• Research: quantifying GitHub Copilot’s impact on developer productivity and happiness: https://github.blog/2022-09-07-research-quantifying-github-copilots-impact-on-developer-productivity-and-happiness/

• Lesson learnt from the DPD AI Chatbot swearing blunder: https://www.linkedin.com/pulse/lesson-learnt-from-dpd-ai-chatbot-swearing-blunder-kitty-sz57e/

• Dennis Yang on LinkedIn: https://www.linkedin.com/in/dennisyang/

• Tim Ferriss’s blog: https://tim.blog/

• Tyler Cowen on X: https://twitter.com/tylercowen

• Tom Cruise on X: https://twitter.com/TomCruise

• Canva: https://www.canva.com/

• Zapier: https://zapier.com/

• Siqi Chen on X: https://twitter.com/blader

• Runway: https://runway.com/

• Universal Primer: https://chat.openai.com/g/g-GbLbctpPz-universal-primer

• “I didn’t expect ChatGPT to get so good” | Unconfuse Me with Bill Gates: https://www.youtube.com/watch?v=8-Ymdc6EdKw

• Microsoft Azure: https://azure.microsoft.com/

• Lennybot: https://www.lennybot.com/

• Visual Electric: https://visualelectric.com/

• DALL-E: https://openai.com/research/dall-e

• The One World Schoolhouse: https://www.amazon.com/One-World-Schoolhouse-Education-Reimagined/dp/1455508373/ref=sr_1_1

• Why We Sleep: Unlocking the Power of Sleep and Dreams: https://www.amazon.com/Why-We-Sleep-Unlocking-Dreams/dp/1501144324

• Gran Turismo: https://www.netflix.com/title/81672085

• Gran Turismo video game: https://www.playstation.com/en-us/gran-turismo/

• Manta sleep mask: https://mantasleep.com/products/manta-sleep-mask

• WAOAW sleep mask: https://www.amazon.com/WAOAW-Sleep-Sleeping-Blocking-Blindfold/dp/B09712FSLY

Production and marketing by https://penname.co/. For inquiries about sponsoring the podcast, email podcast@lennyrachitsky.com.

Lenny may be an investor in the companies discussed.