Ep. 65: My Current AI Tech Stack

Ever been mesmerized by a 100-page AI report you didn't actually need? That's heftnosis, a word I made up (and am very proud of) that describes the AI attraction and excitement that comes from equating output volume with value. In this episode I explain why I don’t label this verbosity-bias as slop, why so many new AI rollouts (Claude Duet included) feel more desperate than helpful, and why responsible AI use starts with asking yourself "Do I actually need all of this? Is it actually helpful?" I'm still very pro AI and I still want you to be ambitious with it, but more is not better. Better is better. Main Topics Covered An introduction to Heftnosis Why I hate the term "slop" Verbosity bias and ChatGPT being "too chatty" The Claude Duet limited rollout "Do I actually need all of this? Is it actually helpful?" The marketing for AI is backwards: here's a solution, go find a problem A friend's questions and how answers differ with vs. without heftnosis AI hacks, breaches, and the bike lock approach How I used AI this week: Patching a WordPress security vulnerability Links & Resources for This Episode Read about Claude Duet Read this episode's Curious Companion Subscribe to the Prompting Curiosity newsletter Submit a Question Visit the Website Feeling curious AND generous? Click here to support the podcast. Chapters (00:00:05) - Introduction to Episode 68(00:00:50) - Defining 'Heftnosis' - AI Output Volume vs Value(00:02:40) - Claude Duet and Agent Overkill(00:05:33) - Marketing Backwards: Solutions Looking for Problems(00:06:30) - Judging When AI Assistance Is Actually Needed(00:08:10) - Closing the Loop on Heftnosis(00:08:53) - AI Security Risks and Staying Informed(00:11:28) - How I Used AI This Week: WordPress Security Fix(00:13:33) - Outro and Newsletter Info

Inspired by a recent article I read, written by an engineering manager at Anthropic, this episode zooms out to look at the bigger picture and dives into the question of whether we are being ambitious enough with AI. We are no longer working with the AI of 2023, and in light of how capable present-day AI has become, I think that we can ask more of it. A certain level of tech tenacity will always be required in order to get the output that you want. This episode looks to consider how much creative ambition is also required. Main Topics Covered The inspiration for this episodeAre we being ambitious enough with AI?Building a website for a client with the help of AI"AI can't do that" and "AI sucks at that"Abstraction and the built-in friction that stops usSaying less to AI and getting a better outputTech tenacity and creative ambitionMy feelings about the AI naysayersHow I used AI this week: Using Gemini's image generator to mock up my dream property Links & Resources for This Episode Read Felix’s articleRead this episode's Curious CompanionSubscribe to the Prompting Curiosity newsletterSubmit a QuestionVisit the WebsiteFeeling curious AND generous? Click here to support the podcast. Chapters (00:00:04) - Intro to Prompting Curiosity(00:00:38) - The 'Be More Ambitious' Article Origin(00:02:25) - Building a Website with Claude(00:03:16) - AI Capability and Overcoming Skepticism(00:04:17) - Understanding Abstraction in AI Use(00:05:22) - From Detailed Prompts to Outcome-Focused AI(00:06:57) - Dreaming Bigger: Raising the Ambition Ceiling(00:08:31) - On AI Naysayers and Personal Values(00:11:14) - Rebranding and Navigating AI Criticism(00:13:08) - This Week's AI Use: Dream Property with Gemini(00:15:15) - Closing and Newsletter Info

In this episode I break down OpenAI's latest rollout, Dots, an always-on agent that lives inside ChatGPT and keeps doing things even after you've closed the chat. I cover what a Dot is, how it works, the logistics around pricing and access, the rules and guardrails, and what Dots can do. I also share my thoughts on abstraction in the agentic era and what the launch of Dots (and Meta Muse) speaks to in the bigger picture. Main Topics Covered My experience with DotsSo...what is a Dot?The personification of AI agentsHow it worksLogistics: Access and "deeper work"Rules and guardrailsWhat can Dots do?Meta Muse vs. OpenAI DotsAbstraction and the agentic eraHow I used AI this week: Gemini inside of Gmail Links & Resources for This Episode Watch Nate Herk’s comparison of Dots and Meta MuseListen to Ep. 61: Everyone Gets an Agent: Meta Muse and the Ongoing Race to AGI Read this episode's Curious CompanionSubscribe to the Prompting Curiosity newsletterSubmit a QuestionVisit the WebsiteFeeling curious AND generous? Click here to support the podcast. Chapters (00:00:05) - Intro and episode setup(00:00:38) - What is Dots? Overview and naming(00:03:08) - Personification and avatar design critique(00:03:50) - How Dots works: setup and communication(00:05:32) - Single chat thread design and workflow(00:06:32) - App connections and capabilities(00:07:20) - Pricing and plan availability(00:08:34) - Cloud hosting and technical infrastructure(00:09:30) - One dot per account policy(00:09:53) - Safety rules and guardrails(00:11:21) - Prompt injection and monitoring safeguards(00:12:35) - What can Dots actually do? Use case breakdown(00:14:27) - Comparing Dots to Metamuse(00:16:12) - Abstraction and the shift away from prompting(00:17:39) - Closing thoughts on Dots(00:18:07) - How I used AI this week(00:19:54) - Wrap up and outro

No gatekeeping here. In this episode I share my current AI tech stack, breaking down everything from my daily driver to my “once in a while but good to know about” tools. AI is inside of literally everything these days, and the majority of it isn’t helpful at all. I’m here to share the things that I actually use and that truly move the needle for me. Main Topics Covered My daily driver for AIHow I use the other LLMsMy favorite AI notetakerThe best AI voice dictation softwareHow I actually use DescriptMy favorite use-case for Gemini NotebookHow I used AI this week: Claude Design to mock up a website Links & Resources for This Episode Listen to Ep 45: Stop Typing, Start Speaking: Wispr Flow Explained Listen to Ep 25: Google NotebookLM: The Best AI Tool You've Never Heard OfRead this episode's Curious CompanionSubscribe to the Prompting Curiosity newsletterSubmit a QuestionVisit the WebsiteFeeling curious AND generous? Click here to support the podcast. Chapters (00:00:05) - Intro and Episode Overview(00:01:22) - AI Limits and Rapid Change(00:02:20) - Claude as Daily Driver(00:03:37) - Granola: AI Note Taker(00:06:05) - Whisper Flow: Voice Dictation(00:07:51) - Descript: AI Video Editing(00:09:44) - Google Notebook LM / Gemini Notebook(00:12:28) - Wrapping Up the Tech Stack List(00:13:11) - How I Used AI This Week: Claude Design(00:16:49) - Closing and Newsletter Info

I recently used Claude to write an entire website proposal in about an hour, and the entire process revealed some pretty stark differences between corporate and creator AI use cases. In this episode I explain why outsourcing the final product is a great choice when the goal is convention, and an absolute no-no when convention is the goal is your voice. Main Topics Covered OpenAI is retiring Custom GPTs on December 11th Boris Cherney's Thread on how he prompts Claude Sonnet 5.5 benchmarks and cost Using Claude to write an entire website proposal Corporate: convention is the goal Creator economy: voice is the goal You cannot and should not automate your voice Hope for AI in the corporate world How I used AI this week: Delilah learning Japanese with ChatGPT Links & Resources for This Episode Listen to Ep. 11: WTF is a Custom GPT? Read OpenAI help article about migrating Custom GPTs to Plugins Read this episode's Curious Companion Subscribe to the Prompting Curiosity newsletter Submit a Question Visit the Website Feeling curious AND generous? Click here to support the podcast. Chapters (00:00:05) - Intro and Episode Overview(00:01:01) - Update: Custom GPTs Retiring(00:02:49) - Boris Cherney on Prompting Claude Naturally(00:04:44) - Sonnet 4.5 Benchmarks and Cost Savings(00:05:19) - Writing a Corporate Proposal with Claude(00:07:13) - Corporate vs Creator AI Use Cases(00:08:42) - Speaking Gigs and Conference Blurbs(00:10:35) - How I Use AI This Week: Learning Japanese with ChatGPT(00:11:57) - Outro, Ratings, and Newsletter

In this episode I cover how AI access and permissions have changed since ChatGPT launched back in 2022, from zero connections and basically zero worry to the current YOLO approach of vibe coders and early adopters. I share what I currently give Claude access to, what my biggest concern is, and offer a read/write/act framework that you can use to decide what your AI should touch. Main Topics Covered How AI access sentiment has changed over 4 years2022: ChatGPT with zero connections2023: Bard connects to Gmail, Docs, Drive, MapsMCP and the "AI in my email" eraAgentic AI, Claude Code, and Claude CoworkWhat I currently give Claude access toPrompt injection: my biggest concernRead/write/act frameworkPrivacy, client data, and benefit vs. riskHow I used AI this week: Claude-Descript connector and pulling quotes from transcripts Links & Resources for This Episode Listen to Ep. 51: WTF is MCPRead this episode's Curious CompanionSubscribe to the Prompting Curiosity newsletterSubmit a QuestionVisit the WebsiteFeeling curious AND generous? Click here to support the podcast. Chapters (00:00:05) - Intro and Episode Overview(00:01:32) - Early AI Access: ChatGPT and Bard Origins (2022-2023)(00:04:16) - MCP and the Rise of Agentic AI (2024-2025)(00:06:23) - Claude Cowork and Mainstream Agentic Adoption(00:08:23) - Personal Evolution in Granting AI Access(00:10:15) - Prompt Injection Risks Explained(00:11:20) - A Framework: Read, Write, Act and Whose Data(00:13:38) - Privacy Concerns vs Personal Comfort(00:15:01) - Training Data Concerns(00:15:28) - Handling Clients' and Others' Data(00:17:08) - Healthcare, Therapy, and Sensitive Data Gray Areas(00:18:34) - Benefit vs Risk: The Big Picture on AI Access(00:20:26) - Final Thoughts on What Your AI Should Access(00:20:58) - How I Used AI This Week: Claude and Descript(00:23:10) - Outro and Sign Off

Is AI actually going to kill us all?There's been a ton of talk about this topic in the media lately, and so I'm dedicating an episode to sharing my thoughts. Spoiler alert, no, I don’t think that AI is going to kill us, but that doesn’t mean that this technology has no risk. I walk through the Jacob Coxon resignation thread and the fact that Anthropic's own alignment team admits there's genuine risk, then get into why the doom talk feels more irresponsible than credible. I don't think regulation is coming to save us anytime soon, but to me, the answer isn't fear, it’s curiosity. Main Topics Covered Directly answering the question: Will AI kill us all? Jacob Coxon's resignation thread on X Evan Hubinger's tweets on AI risk Defining superintelligence and recursive self-improvement (RSI) Moving the goalposts from AGI to superintelligence AI's coding and hacking capabilities Jake Handy's prediction about a coming AI crisis Why the AI-doom talk feels irresponsible Regulation and data centers as oil Dario Amodei's blog post Change doesn't happen in a vacuum How I used AI this week: Replacing my Jeep’s thermostat housing Links & Resources for This Episode Read Jacob Coxon's resignation thread on X Listen to Ed Zitron's podcast episode about AI and regulation  Read Dario Amodei's blog post, "We Must Pace the Frontier"  Read this episode's Curious Companion Subscribe to the Prompting Curiosity newsletter Submit a Question Visit the Website Feeling curious AND generous? Click here to support the podcast. Chapters (00:00:05) - Intro and today's topic: will AI kill us all?(00:03:29) - Background: Jacob Cox's resignation and Anthropic's Evan Hubinger(00:05:18) - Defining superintelligence and recursive self-improvement(00:07:19) - Why the doom claims seem overblown right now(00:10:39) - Hacking risks and predictions of a coming AI-caused crisis(00:12:15) - Irresponsible messaging and the case for transparency(00:13:24) - Government incentives, profit motives, and the China narrative(00:17:24) - Dario Amodei's regulation plan and skepticism about government action(00:19:10) - Staying a realist: humans, open source, and hope for the future(00:21:11) - How I Use AI This Week: the rattling thermostat story(00:23:17) - Closing thoughts and sign-off

In this episode I talk about the eleventy billion AI models that dropped in the past few weeks, including Anthropic's Fable 5.1 and Mythos 5.1, OpenAI's GPT-6 Astra, and Meta's new Muse agent. I get into former Anthropic researcher Jacob Coxon's resignation and his warning about AI risk, plus what it actually means that OpenAI just rated Astra "Critical" for cybersecurity capability. I also share why I think Meta Muse is a terrible idea and what all of these model drops say about the direction AI is heading. Main Topics Covered Jacob Coxon's resignation and AI regulation warningsAnthropic's Fable 5.1 and Mythos 5.1 updatesOpenAI's GPT-6 Astra releaseAstra's "Critical" cybersecurity capability ratingAGI throwback to Episode 7Meta Muse launch: everyone gets an agentMeta Muse access, pricing, and privacy detailsWhy Meta Muse is a terrible ideaHow I used AI this week: Gemini AI Overviews Links & Resources for This Episode Read Jacob's statements on X Listen to Ep. 7: What is AI? Subscribe to the Prompting Curiosity newsletterSubmit a QuestionVisit the WebsiteFeeling curious AND generous? Click here to support the podcast. Chapters (00:00:05) - Intro and welcome(00:00:38) - Jacob Coxon's anthropic resignation statement(00:01:41) - Reflections on AI regulation and motives(00:04:56) - Recent model drops overview(00:06:14) - Anthropic's Fable and Mythos updates(00:07:25) - OpenAI's GPT-6 Astra and computer use capabilities(00:10:01) - Astra's critical cyber capability rating(00:11:27) - AGI claims and the race to super intelligence(00:13:46) - Meta's Muse agent launch and concerns(00:15:13) - Muse's privacy and data risks(00:16:44) - Who Meta Muse targets and risks to users(00:17:38) - Companies' true priorities: money over safety(00:18:23) - How I used AI this week: Gemini overviews(00:20:56) - Wrap up and outro

Inspired by a podcast episode my girlfriend recently listened to, I use this episode to tackle the question: “Is it cheating if you use AI”? Using the definition of cheating as a springboard, I break down why the school-and-sports framing most of us default when it comes to cheating, and why the topic warrants a more nuanced approach when considering AI. Along the way I discuss the role of the system at large, and round the episode out by diving into an equally exciting topic for me: how AI use affects trustworthiness. Main Topics Covered AI as holding up a mirrorDefining cheatingWhy the school/sports rulebook framing doesn't fit AIWhy AI is an easy punching bag right nowBreaking down disclosure, rules, and trustQuestions worth askingIs someone less trustworthy if they use AI?How I used AI this week: Reverse image search to find desk assembly instructions Links & Resources for This Episode Listen to Ep. 3: Is ChatGPT Killing Creativity?Read this episode's Curious CompanionSubscribe to the Prompting Curiosity newsletterSubmit a QuestionVisit the WebsiteFeeling curious AND generous? Click here to support the podcast. Chapters (00:00:05) - Prompting Curiosity: The AI Curious(00:00:38) - Is It Cheating If You Use AI in Your School?(00:02:39) - Is It Cheating to Use AI?(00:03:39) - Is It Cheating If You Use AI?(00:06:47) - I Don't Care If You Use AI(00:07:44) - Does AI Make You Cheat?(00:10:06) - Why Do We Care About AI Cheating?(00:10:59) - Is It Cheating if You Use AI?(00:14:40) - How I Used Claude for AI This Week(00:16:24) - Crowning the Curious: A Tech Podcast

In this episode I break down WTF a frontier model actually is, starting with what most people assume the term means versus the real technical definition. I cover how "frontier model" got coined in 2023, and why the term even came about at all, and how self-naming benefitted its creators. A bit of an etymology episode, this episode is more about the background of the term, and explores how new open weight models are challenging the definition. Main Topics Covered WTF is a frontier model?Most people's assumption vs. the technical definitionOrigin of the term: the Frontier Model Forum (July 2023)FMF's official definition of "frontier model"EU AI Act's compute-based definitionWho made the rulesFMF today: expanded membership, safety research, AI Safety FundWhy forming the FMF actually benefited these companiesOpen weight models and "frontier-grade" labelsHow I used AI this week: building my mom a workout tracker Links & Resources for This Episode Listen to Ep. 55: Weight, What? An Introduction to Open Source AI ModelsRead this episode's Curious CompanionSubscribe to the Prompting Curiosity newsletterSubmit a QuestionVisit the WebsiteFeeling curious AND generous? Click here to support the podcast. Chapters (00:00:05) - Prompting Curiosity: The AI Curious(00:00:38) - Predicting Curiosity: Frontier Models(00:01:30) - The Frontier Model Forum and its Impact(00:05:56) - Frontier: The Frontier Group(00:08:07) - What's the Frontier?(00:10:00) - How I Used Claude for AI This Week(00:14:15) - AI Curious: A Podcast About Curious People