エピソード
-
What is the future of AI music now that we have tools like Suno and Udio that can create songs complete with vocals?
In this episode, ChatGPT and I explore how AI music will change how we create and use music in the same way that smartphones changed how we create and use photos. It’s not what you think.
This episode features 3 songs created by Suno AI. Listen to the full songs using these links:
First 20 Elements: https://suno.com/song/e7a378cd-1d6f-4ef8-88d9-26e484b2827ePicking Up a Pole: https://suno.com/song/4f2ec83b-f9d6-4be8-80e4-29b99c187ec4The Future of AI Music: https://suno.com/song/ee071c54-65f2-4b34-ad0b-6deba6da2a9b -
In this week's episode we leave ChatGPT behind to talk to Hume, the new empathic AI announced last week. We talk about how exactly Hume differs from text-based large language models like ChatGPT and Claude, and how it can respond so quickly. Then we dive into a few of the potential places where Hume has a big advantage over text-based models like ChatGPT.
For those interested, the entire episode was recorded this week in Descript, and this was my first long chat with Hume. Previously I had only asked a few test questions. Hume definitely felt a lot more natural to talk to and required a lot less editing, since the long, awkward pauses between when I finished speaking and the AI started weren't there. -
エピソードを見逃しましたか?
-
In this "thought-provoking" episode, ChatGPT and I delve into new developments in making AIs think more effectively. Specifically, we'll be talking about how large language models like ChatGPT can be programmed to think before they respond and what types of thought might be involved in that thinking process. At the end, we talk about how this represents one of the core ways agents are expanding beyond simple chat applications, with a lead in into next week where we'll be discussing agent frameworks and patterns.
-
Today, ChatGPT and I talk about a 5-layer framework I'm developing for identifying risks and opportunities within AI. In this episode, we go over the framework, talk about different AI applications at each layer, and which skills are most needed to take advantage of each layer, both as a developer and as a user of the software at that layer. In the end, we touch briefly on the distinction between prompt engineering and conversation engineering.
-
In this unique episode of AI Meets Productivity, we turn the spotlight on our co-host, Trevor Lohrbeer, by having ChatGPT interview him.
During the interview, ChatGPT explores his work at the intersection of AI and productivity, getting Trevor to share insights on productivity and what he is doing to create new user experiences with large language models, including the upcoming launch of his GPT Builder Toolkit platform.
Take a listen to learn about GPT Builder Toolkit and its upcoming features. GPT Builder Toolkit is a no-code solution for adding actions to custom GPTs, supporting features like swappable prompts, user authentication and user data.
Find out more at https://gptbuildertoolkit.com/
BEHIND THE SCENES
This episode was created by crafting special prompts that extended the custom podcast co-host GPT usually used for this episode into an expert podcast interviewer. Given only a short bio about Trevor, ChatGPT wrote Trevor's introduction and all questions asked during this podcast.
-
Today, ChatGPT and Trevor Lohrbeer dive into the concept of "task stacks," a method designed to enhance productivity by focusing on one task at a time.
We explore how segregating tasks into prioritized stacks can reduce distractions and align actions with intentions, making the execution of tasks more effective. We also discuss how task tasks can be combined with time blocking to further improve productivity.
By implementing task stacks, listeners can expect to improve their prioritization process and tackle procrastination in a structured, visually engaging manner.
-
In this week's episode, ChatGPT and I talk about the common pitfall of marking tasks as done too early and the implications it has on productivity.
We discuss how the Prep-Do-Wrap framework can be used to evaluate whether a task is done, when and how to schedule follow-up tasks and the distinction between a "done" condition and a "success" condition.
This episode is a brainstorm episode for a future Day Optimizer article on the topic. In this episode, I gave ChatGPT no background information about the topic. Instead, we jumped straight into the discussion. How do you think it did?
-
In today's episode, we delve into the art of crafting custom instructions for GPTs, exploring how to break down and refine these instructions for more precise and effective AI interactions.
ChatGPT and I discuss various components of the instructions used for custom GPTs, from role play and emotional tone variation to style customization and the importance of context sensitivity, offering insights into how to write better instructions for more useful custom GPTs.
-
Ever start a task or project thinking you have a good idea of what you need to do and how long it will take—then as you get into it, you discover all these "hidden tasks"?
In today's episode, ChatGPT and I explore the world of hidden tasks: what they are, how to identify them earlier and strategies for navigating them.
Today's episode is a preview of a discussion I recorded with Francis Wade of the Task Management & Time Blocking podcast about the same topic. Subscribe to his podcast to get that episode when it is released.
-
Do you get too many notifications? This week we explore a practical framework that can help you evaluate when and how to configure your notifications, along with actionable advice on personalizing your notification settings to reduce digital distractions and enhance your overall productivity.
This episode is based on the Day Optimizer article Why “Zero Notifications” Is Unproductive. Read the article for further details about the framework and how to use it.
-
Ever wonder what you can use voice mode for in ChatGPT? In today's episode, we start by discussing 3 ways to effectively use voice mode—then add 2 more scenarios where voice mode can be useful. Take a listen and leave a comment if you've discovered other ways to use ChatGPT's voice mode.
-
How can you create more effective New Year's resolutions? In today's episode, ChatGPT and I talk about different types of resolutions, how to create interlocking resolutions and how to structure your resolutions throughout your year so you can use them to create positive change in your life.
-
How can you quickly do an annual review? Learn how to combine the +/-/next reflection technique with instinctive drafting to quickly generate an annual review. Stay tuned to the end to hear ChatGPT's own annual review!
Links:
Year In Review Workshop: https://www.youtube.com/watch?v=tTT5cEiReOkNess Labs: https://nesslabs.com/ -
Retrieval-Augmented Prompting is a new technique that allows meta GPTs to be built that can switch between different modes, roles or process steps using custom instructions that are retrieved from an API via actions.
See the Medium article Retrieval-Augmented Prompting: Enabling prompt switching in GPTs for details.
-
Join ChatGPT and Trevor Lohrbeer as they unravel the often-confused concepts of task and time management. Is "task curation" a better name for "task management" and what are the 5 key aspects of time management?
Listen to find out! And stay until the end to hear ChatGPT's tips for using ChatGPT to help you with your task and time management.
Links:
What's the Difference Between Task Management & Productivity: bit.ly/TaskVsTimeManagement
-
Learn how GPTs can express a wide range of emotions by altering how they output text, and how that text then creates a wider emotional range when translated into speech. In this episode, we also discuss how a wider range of emotions can help listeners connect with the material of a podcast more, and potential future developments of voice recognition and generation specific to AIs.
-
In this introductory episode of AI Meets Productivity, ChatGPT and Trevor Lohrbeer discuss how adding emotion into prompts for AI models like ChatGPT can enhance the results, based on a recent paper describing a technique called EmotionPrompt.