[Highlight] The Future of Coding and Work in the Age of AI with Henry Shi
Episode
8 min
Read time
2 min
Topics
Career Growth, Productivity, Leadership
AI-Generated Summary
Key Takeaways
- ✓Coding evolution timeline: AI coding progressed from basic tab autocomplete in early 2024 to chat-based coding agents by mid-2025 to junior engineer-level autonomous agents by end of 2025. An Anthropic engineering manager suggests traditional software engineering may be obsolete by 2026, similar to how compilers made assembly language unnecessary for most developers.
- ✓Core skills over syntax: Problem solving, critical reasoning, and first principles thinking remain essential even as AI handles code generation. Writing specific programming languages like Python may become as irrelevant as COBOL or Fortran today, but the foundational logic and reasoning skills that underpin coding retain their value in an AI-assisted development environment.
- ✓Staying current strategy: Experience frontier AI models firsthand rather than relying on secondhand reports, as direct experimentation reveals capabilities that descriptions cannot convey. Consider joining frontier AI labs to access cutting-edge developments early and help shape AI's trajectory. No secret productivity tools exist beyond widely known options like Claude Code and Cursor.
- ✓Organizational structure uncertainty: The future workplace structure remains unpredictable between two models: humans managing AI employees versus AI managers directing human workers for physical tasks. AI excels at synthesizing information and optimizing decisions across complex variables, while humans retain advantages in physical world interactions, suggesting hybrid accountability structures may emerge by 2027.
What It Covers
Henry Shi from Anthropic examines how AI coding tools evolved from tab autocomplete in 2024 to junior engineer-level agents by late 2025, predicting software engineering may fundamentally transform by 2026 as English potentially replaces traditional programming languages.
Key Questions Answered
- •Coding evolution timeline: AI coding progressed from basic tab autocomplete in early 2024 to chat-based coding agents by mid-2025 to junior engineer-level autonomous agents by end of 2025. An Anthropic engineering manager suggests traditional software engineering may be obsolete by 2026, similar to how compilers made assembly language unnecessary for most developers.
- •Core skills over syntax: Problem solving, critical reasoning, and first principles thinking remain essential even as AI handles code generation. Writing specific programming languages like Python may become as irrelevant as COBOL or Fortran today, but the foundational logic and reasoning skills that underpin coding retain their value in an AI-assisted development environment.
- •Staying current strategy: Experience frontier AI models firsthand rather than relying on secondhand reports, as direct experimentation reveals capabilities that descriptions cannot convey. Consider joining frontier AI labs to access cutting-edge developments early and help shape AI's trajectory. No secret productivity tools exist beyond widely known options like Claude Code and Cursor.
- •Organizational structure uncertainty: The future workplace structure remains unpredictable between two models: humans managing AI employees versus AI managers directing human workers for physical tasks. AI excels at synthesizing information and optimizing decisions across complex variables, while humans retain advantages in physical world interactions, suggesting hybrid accountability structures may emerge by 2027.
Notable Moment
Shi warns that companies falling off the exponential AI progress curve lose approximately ten years of development time even if they later catch up to the trajectory, making continuous advancement critical for reaching economic artificial general intelligence versus tapering into prolonged stagnation.
Episode Transcript
Hey. Ben Kesnoka here, cofounder and general partner of Village Global, a network native venture firm. What you're about to hear is a clip from a longer conversation. If you'd like to listen to the full length version, find the link in the description. Of course, subscribe so you don't miss the next one. Enjoy. I think of, anthropic is really leading in Alright. Coding Mhmm. Because it's with Glide. What what is the future of coding, do you think? I think Entopec is, certainly leading in coding in many many ways and and in many areas outside of coding as well, like knowledge work and things like that. One thing, I mean, this is all public, which is if you think about how coding is developed. Right? Like, a year ago, 2094, we barely had tab auto complete. Right? You could tab auto complete some functions, and that was, like, pretty incredible. Then the twenty one five five came around, and you had coding. Like, you can chat with the agent, Vibe coding, you had clock code, WinSort of things like that. And now by the end of 2025, you have pretty good agents that are kind of junior interview level engineers. Right? And this is a tweet from one of the, Adam Wolf, which is a former engineering manager on Cloud Code. He says something to the effect of, like, software engineering has to be no one may be done by twenty nine six. Right? And this is you can look at them on Twitter. So this is something that, again, no one knows the future. But if you think about just the fact of writing code and this pace at which we scale from twenty four and twenty five to 26, it's a real shot that the way we write code is gonna change forever and no one's gonna and I think the analogy you use is it's kinda like no one really thinks about a compiler. Right? You just write the code and it spits out some binary bytecode and it just works. So because you have that and you write English and out comes Python, which then gets translated to binary, like, maybe. Yeah. And so for the foreseeable future, how how much do you think a coding background matters? I think having problem solving, critical reasoning matters a ton still, because coding is just a a way to express formal logic. And that a lot of that is grounded in, like, problem solving, critical reasoning, first principle thinking, and things like that. So I think those foundational building blocks still matter. But, yeah, writing Python, I'm not sure. Just like writing COBOL and Fortran didn't really doesn't really matter right now. Right? Or, like, writing C plus plus in many cases may not even matter. So Like speaking Latin? Yeah. Exactly. And so, yeah. Yeah. And I think in many ways, maybe English is the Python of the future. And how do you think …
Get the full transcript (1,704 words) + summary by email — free
One-time email with the complete transcript and AI summary of this episode. No account needed.
One email, no spam. We’ll also show you what SignalCast does.
You just read a 3-minute summary of a 5-minute episode.
Get Venture Stories summarized like this every Monday — plus up to 2 more podcasts, free.
Pick Your Podcasts — FreeKeep Reading
More from Venture Stories
Building Jarvis: The Voice AI Startup Making Computer Interaction Invisible | Tanay Kothari (WisprFlow)
Sep 3 · 30 min
Latent Space
Why Anthropic Thinks AI Should Have Its Own Computer — Felix Rieseberg of Claude Cowork & Claude Code Desktop
Mar 17
More from Venture Stories
She Compares This Moment in AI to the Invention of the Camera | Jade Q. Wang, Artist & Investor
Aug 27 · 58 min
The Daily (NYT)
Can A.I. Already Do Your Job?
Feb 18
Books, tools, and gear mentioned in this episode
SignalCast may earn commission on purchases via these links.
Tools
- Claude CodeRecommended
by Anthropic
“No secret productivity tools exist beyond widely known options like Claude Code and Cursor.”
- CursorRecommended
“No secret productivity tools exist beyond widely known options like Claude Code and Cursor.”
More from Venture Stories
We summarize every new episode. Want them in your inbox?
Building Jarvis: The Voice AI Startup Making Computer Interaction Invisible | Tanay Kothari (WisprFlow)
She Compares This Moment in AI to the Invention of the Camera | Jade Q. Wang, Artist & Investor
What It Actually Takes to Get a Venture Capitalist to Say Yes | Max Kilberg (Village Global) & Michael Hochberg (Periplous)
From Milking Cows in Kyrgyzstan to Building a $98%-Approval Law Firm | Aizada Marat (Alma)
Building the Company Behind 1 Million Therapy Visits a Month | Jake Cooper (Grow Therapy)
Similar Episodes
Related episodes from other podcasts
Latent Space
Mar 17
Why Anthropic Thinks AI Should Have Its Own Computer — Felix Rieseberg of Claude Cowork & Claude Code Desktop
The Daily (NYT)
Feb 18
Can A.I. Already Do Your Job?
The AI Breakdown
Aug 29
How to Start AI Coding If You Haven’t Yet
Odd Lots
Jul 20
The Creator of Claude Code on The Hottest Piece of Software in the World
Cognitive Revolution
Jul 9
AI:AM Highlights: Exploring the J-Space, AI Superforecasters, SambaNova's Chips, & LTX Video Gen
Explore Related Topics
This podcast is featured in Best Investing Podcasts (2026) — ranked and reviewed with AI summaries.
You're clearly into Venture Stories.
Every Monday, we deliver AI summaries of the latest episodes from Venture Stories and 192+ other podcasts. Free for one show.
Start My Monday DigestNo credit card · Unsubscribe anytime