Big Technology Podcast · Alex Kantrowitz

Claude Code’s Shining Moment, ChatGPT for Healthcare, End Of Busywork?

·56 min·3 clips

Ethan Mollick says Claude Code worked for 1 hour and 14 minutes and deployed a working website.

1. Big Technology Podcast Friday edition focuses on Claude Code, ChatGPT for healthcare, prediction markets, and the end of busy work. 2. Alex Kantrowitz hosts with Ranjan Roy of Margins, who works at Writer and talks as a product builder and AI observer. 3. The episode asks whether Claude Code and ChatGPT are moving from narrow tools into systems that do routine work. 4. Alex opens with Ethan Mollick’s post, “Claude Code and what comes next,” and his example of asking Claude Code to build a startup that could make $1,000 a month. 5. Mollick says Claude Code chose to sell “sets of 500 prompts for professional users for $39,” then worked independently for “1 hour and 14 minutes.” 6. Ranjan says that specific result is impressive but already “table stakes,” because autonomous work and tool use are becoming the baseline. 7. The hosts argue over Mollick’s phrase “agentic harness,” with Alex preferring the simpler idea that AI can do autonomous work and use tools. 8. Ranjan says Claude Code is many people’s first real exposure to “truly agentic AI,” not just code autocomplete or vibe coding. 9. Alex says the episode matters because the pattern could broaden from programming into knowledge work. 10. The discussion turns to how Claude Code can run for hours by compacting context, taking notes, clearing its window, and resuming work. 11. Alex says Mollick’s example shows Claude Code can create a website, test it, and then use natural language to improve it with minimal input. 12. Ranjan says he has already built routines that check Salesforce and Gong, summarize account changes, and email the result. 13. The hosts connect this to Ethan Mollick’s claim that today’s AIs are capable of “real sustained work that actually matters.” 14. Alex cites Andre Karpathy’s line that he has “never felt this much behind as a programmer,” using it as a marker of how quickly the tooling is changing. 15. The episode then shifts to Gmail’s new AI inbox, which includes “suggested to do’s” and “topics to catch up on.” 16. Ranjan says he still uses a 2011 Gmail flag system from Stacy Marie Ishmael, which shows how sticky existing workflows can be. 17. The healthcare segment starts with OpenAI’s claim that 230 million people globally ask health and wellness questions to ChatGPT every week. 18. Alex describes ChatGPT Health as a separate space with enhanced privacy, plus integrations with Apple Health, Function, and MyFitnessPal. 19. The conversation is practical and skeptical at the same time, with Alex and Ranjan trading examples from email, doctor visits, and work routines. 20. Listeners who follow AI products, workplace automation, and consumer software will get the most value, while people wanting a single-topic episode may skip it.

As heard by us

A focused discussion of why AI's hardest workplace barrier may be familiar process.

CloudCode's move beyond programming gives this Big Technology Podcast Friday edition a useful center of gravity: AI may already be able to improve parts of knowledge work, but the tougher barrier is the familiar process people have learned to trust.

Read the full review in PlayNext →

Why you'd press play

Want to know whether AI can end busywork when the hardest part is prying people away from the Google Doc?

Read the full recommendation in PlayNext →

Listen to the show on

Apple Podcasts Spotify

Play clips from this episode More from Big Technology Podcast