What are the key takeaways from this a16z Podcast episode?

Key insights include: **The Coding Agent Limitation:** Coding agents like Claude Code, Codex, and Cursor produce the same type of code humans wrote a decade ago — faster, but not more capable. Developers should distinguish between tools that accelerate code production and tools that expand software's expressive power. Only the latter enables genuine automation of tasks software couldn't previously handle.; **JEV as a New Primitive:** JEV functions as a library developers embed directly in code, accepting natural language intent and a state machine, then returning decisions with confidence levels. This is architecturally distinct from calling an external LLM. Developers can program against it without constructing example queries once reliability reaches sufficient thresholds, enabling continuous flow-state development.; **Reliability Over Demo Performance:** TypeSafe prioritizes three distinct reliability properties: uptime, robustness (consistent intelligence level across calls), and functional equivalence under semantically identical inputs. Almeida argues each additional "nine" of reliability unlocks entirely new application categories, and that benchmark-optimized models have systematically underdelivered on real automation precisely because human evaluators judge style, not task completion.

What did Diogo Almeida discuss on a16z Podcast?

a16z partners Ben Horowitz and Martin Casado interview TypeSafe AI founder Diogo Almeida about JEV, a new programming primitive that embeds AI decision-making directly into software code. The conversation contrasts coding agents that accelerate existing software production against tools that expand what software itself can fundamentally do. Key topics include: **The Coding Agent Limitation:** Coding agents like Claude Code, Codex, and Cursor produce the same type of code humans wrote a decade ago — faster, but not more capable. Developers should distinguish between tools that accelerate code production and tools that expand software's expressive power. Only the latter enables genuine automation of tasks software couldn't previously handle.; **JEV as a New Primitive:** JEV functions as a library developers embed directly in code, accepting natural language intent and a state machine, then returning decisions with confidence levels. This is architecturally distinct from calling an external LLM. Developers can program against it without constructing example queries once reliability reaches sufficient thresholds, enabling continuous flow-state development..

How long is this episode of a16z Podcast?

This episode is 43 minutes long. SignalCast provides an AI-generated summary so you can get the key insights in about 3 minutes.

a16z Podcast

AI Can Write Code. Why Isn’t Software Better?

September 28, 2026

43 min episode · 2 min read

Diogo Almeida

Episode

43 min

Read time

2 min

Topics

Productivity, Relationships, Investing

AI-Generated Summary

Published Sep 28, 2026

Key Takeaways

✓The Coding Agent Limitation: Coding agents like Claude Code, Codex, and Cursor produce the same type of code humans wrote a decade ago — faster, but not more capable. Developers should distinguish between tools that accelerate code production and tools that expand software's expressive power. Only the latter enables genuine automation of tasks software couldn't previously handle.
✓JEV as a New Primitive: JEV functions as a library developers embed directly in code, accepting natural language intent and a state machine, then returning decisions with confidence levels. This is architecturally distinct from calling an external LLM. Developers can program against it without constructing example queries once reliability reaches sufficient thresholds, enabling continuous flow-state development.
✓Reliability Over Demo Performance: TypeSafe prioritizes three distinct reliability properties: uptime, robustness (consistent intelligence level across calls), and functional equivalence under semantically identical inputs. Almeida argues each additional "nine" of reliability unlocks entirely new application categories, and that benchmark-optimized models have systematically underdelivered on real automation precisely because human evaluators judge style, not task completion.
✓SaaS as AI's Biggest Winner: Contrary to the "SaaSpocalypse" narrative that coding agents commoditize software, Almeida argues established SaaS companies hold the strongest position in the AI era. They already possess distribution, deep workflow knowledge, and user trust. Adding genuine intelligence as a software primitive — not a chatbot layer — compounds their existing capital investment rather than eroding it.
✓Intelligence Per Dollar as North Star: When designing AI systems, optimize for intelligence per dollar rather than intelligence per second, especially for automation targets deep inside software stacks. Almeida estimates that the vast majority of future AI function calls will be machine-to-machine, embedded in system internals, not human-facing. Building toward that architecture now determines whether AI reaches genuine economic automation.

What It Covers

a16z partners Ben Horowitz and Martin Casado interview TypeSafe AI founder Diogo Almeida about JEV, a new programming primitive that embeds AI decision-making directly into software code. The conversation contrasts coding agents that accelerate existing software production against tools that expand what software itself can fundamentally do.

Key Questions Answered

•The Coding Agent Limitation: Coding agents like Claude Code, Codex, and Cursor produce the same type of code humans wrote a decade ago — faster, but not more capable. Developers should distinguish between tools that accelerate code production and tools that expand software's expressive power. Only the latter enables genuine automation of tasks software couldn't previously handle.
•JEV as a New Primitive: JEV functions as a library developers embed directly in code, accepting natural language intent and a state machine, then returning decisions with confidence levels. This is architecturally distinct from calling an external LLM. Developers can program against it without constructing example queries once reliability reaches sufficient thresholds, enabling continuous flow-state development.
•Reliability Over Demo Performance: TypeSafe prioritizes three distinct reliability properties: uptime, robustness (consistent intelligence level across calls), and functional equivalence under semantically identical inputs. Almeida argues each additional "nine" of reliability unlocks entirely new application categories, and that benchmark-optimized models have systematically underdelivered on real automation precisely because human evaluators judge style, not task completion.
•SaaS as AI's Biggest Winner: Contrary to the "SaaSpocalypse" narrative that coding agents commoditize software, Almeida argues established SaaS companies hold the strongest position in the AI era. They already possess distribution, deep workflow knowledge, and user trust. Adding genuine intelligence as a software primitive — not a chatbot layer — compounds their existing capital investment rather than eroding it.
•Intelligence Per Dollar as North Star: When designing AI systems, optimize for intelligence per dollar rather than intelligence per second, especially for automation targets deep inside software stacks. Almeida estimates that the vast majority of future AI function calls will be machine-to-machine, embedded in system internals, not human-facing. Building toward that architecture now determines whether AI reaches genuine economic automation.

Notable Moment

Almeida reveals that OpenAI internally described AGI circa 2020 as placing a leading AI researcher inside every conditional statement in code — a framing that reframes AGI not as a sentient superintelligence but as pervasive, embedded decision-making throughout software systems, which is precisely what JEV attempts to operationalize today.

Know someone who'd find this useful?

Episode Transcript

Where the fuck is all the automation? AI is so unbelievably smart, and yet it's so useless at all other stuff. It doesn't matter how much AI coding agents you use. The software actually isn't getting better. Maybe you're writing it faster. It's like arguably getting worse. OpenAI has been trying to automate customer service since 2020. What I want instead is smart software. I want to expand what software itself can do such that things that should be automatable can then be automatable. My favorite thing that you guys say is we built prod, not god. That's so good. Because if we had any other kind of, like, big lab leader, even if they had joy, they would cover that. And then your view is so different. You're like, no. We're gonna create a way better world. For nuanced reasons, I don't think we are on the path of RSI. In the SaaSpocalypse story AI can write software, but what if the bigger opportunity is putting intelligence inside the software itself? In this episode, Ben Horowitz and Martin Casado sit down with TypeSafe AI founder Diogo Almeida to talk about JEV and a different vision for how AI changes computing. Diogo argues that coding agents make it faster to produce the same kind of software we already have. JEV is aimed at expanding what software itself can do, giving developers a new primitive for turning natural language intent into decisions that programs can actually use. They get into why reliability matters more than impressive demos, what a new era of programming could look like, and why AI might make existing software dramatically more useful rather than simply replacing it. And beneath all of this is the question that drove Diogo to build TypeSave in the first place. If AI is already this smart, where is all the automation? Today, we have the founder and leader of TypeSafe, Diogo, with us, who is a bit of a hero to both Martine and me. He is not only building like a really interesting product, but creating what we think is a very important movement. So we're super excited about today. Welcome to Thank you. Maybe you can give us a brief on what is JEV, what is TypeSafe, why is it important? Is this a curse friendly or no? Yeah. Oh, yeah. Yeah. Oh, okay. Cool. Parker, you talk. Okay. Okay. Okay. Cool. So I was actually asked sounds when you're ready. For, like, a elevator pitch, which I tend to ramble on, and I don't do well. But, like, I realized my favorite elevator pitch for Jeff is Yeah. Where the fuck is all the automation? Like, this is, like, so unbelievably tragic. Yes. So much intelligence. AI is so unbelievably smart, and yet so not that I hate on chatbots or coding agents. I love them myself. But it's, like, so useless at all other stuff, and it's tragic. It's tragic so that we have so much, like, diamond …

Get the full transcript (8,455 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.

Browse all a16z Podcast transcripts →