Gemini 4 Argon Really Surprised Me WebTwo Minute PapersOct 19 Google Gemini 4 agents just beat Minecraft with no human intervention WebEngadget AIOct 19 Who could it be? 🧐 OfficialGoogle on YouTubeOct 18 Customize your Gemini chats experience Learn & DoGemini Help CenterOct 17 Gemini 4 Argon Made a Fall Guys Clone WebMatthew BermanOct 17
Google OfficialWebLearn & DoResources People Podcasts
All NewsTradeVideosVoicesCommunityResearchPodcasts
Subvolts take AI can automate AI verification, speeding up safety checks for complex systems. Feb 5, 2026
For you
Developers
Learn how AI can automate complex verification tasks, improving reliability and reducing development time for your AI systems.
Business
Understand how AI-driven verification can accelerate the deployment of safe and trustworthy AI in critical applications.
Hype note
3/5Notable

This interview explores a critical, practical challenge in AI development: verification. It offers insights into how AI itself can solve these complex problems.

Gradient Dissent
Gradient Dissent · S1 · Ep 136 Episode

The $64M Bet on an AI That Has to Be Right | Carina Hong, CEO of Axiom

51 min Interview Hosts: Lukas Biewald
0:0050:40

Audio streams from Lukas Biewald. Your place is remembered on this device.

Go to the podcast episode page
Open Gradient Dissent in
Apple PodcastsOvercastPocket Casts
Follow the show
Apple PodcastsSpotifyOvercastPocket CastsAny Android app

From Subvolts

Summary and analysis by Subvolts editorial, AI-assisted.

Summary

Lukas Biewald interviews Carina Hong, CEO of Axiom, on the Gradient Dissent podcast. They discuss how AI can automate formal verification, a process currently consuming years of human effort. Hong explains Axiom's approach to auto-formalization, which mirrors spec-driven models and addresses the growing bottleneck in high-stakes AI systems, from mathematics to hardware and software.

Explained simply

Imagine checking if a complex machine works perfectly. This takes a long time. This episode discusses how AI can help do that checking much faster, which is important for making sure AI systems are safe and reliable.

Why listen

This episode explores the critical challenge of verifying complex AI systems. It's for AI engineers, researchers, and product leaders concerned with reliability and safety in AI. Axiom's work on automated formal verification directly addresses a growing concern for frontier labs like Google Gemini, Anthropic, and OpenAI, who are developing increasingly sophisticated models that require rigorous testing to ensure they operate correctly and safely, especially in high-stakes applications.

On the mic
  • HostLukas Biewald
  • GuestCarina HongFounder & CEO, Axiom
About
GoogleAnthropicOpenAIResearchDeveloper ToolsSafety
Key facts
  • Guest: Carina Hong, CEO of Axiom
  • Show: Gradient Dissent
  • Length: 51 minutes
  • Topic: AI for formal verification
  • Current State: Formal verification consumes years of human effort.
Show notes from Lukas Biewald

Formal verification already consumes years of human effort.

In this episode, Lukas Biewald talks with Carina Hong, Founder & CEO of Axiom, about why verification is becoming the real bottleneck in high stakes AI systems.

They discuss how Axiom uses AI to take on the tedious checking that stretches verification cycles across years, starting with formal mathematics and extending to hardware and software.

Carina also explains why Axiom’s approach to auto-formalization mirrors spec driven models like Kiro from AWS.

Connect with us here:

Carina Hong: https://www.linkedin.com/in/carina-hong/

Axiom: https://www.linkedin.com/company/axiommath/

Lukas Biewald: https://www.linkedin.com/in/lbiewald/

Weights & Biases: https://www.linkedin.com/company/wandb/

Jargon decoded
Formal verification
Carina Hong
Using math to prove a system works exactly as intended.
Auto-formalization
Carina Hong
Automatically creating the mathematical proofs needed for verification.
Spec driven models
Carina Hong
AI models built based on detailed requirements or specifications.
Published
Show
Gradient Dissent · Lukas Biewald
Length
51 min
Companies
Google, Anthropic, OpenAI
Summary by
Subvolts, using an AI model (how we work). Spotted a mistake? Tell us.
#Research#Developer Tools#Safety

Best of Gradient Dissent

Same guests, other shows

More on Google this week

Questions people ask

What is formal verification?
Formal verification is a process that uses mathematical methods to check if a system meets its specifications. It currently takes years of human effort.
How does Axiom use AI for verification?
Axiom uses AI to automate the tedious checking involved in formal verification, speeding up cycles for mathematics, hardware, and software.

Page generated Oct 19, 2026. Summaries are Subvolts' own; the audio, artwork and notes belong to Lukas Biewald.