AI Security Table
AI Security Table
✓

Is Spec-Driven Development Already Dead

00:41:49
Podcasts
RSS
Link

SOBRE ESTE EPISÓDIO

In this episode, we take on spec-driven development, the resurgent idea that writing a detailed spec and letting AI implement it will finally give us the precision engineering promised us since the 1950s. We push back on the notion that this fixes what actually sank spec-driven approaches decades ago: specs were never the problem; following them was. Along the way, we debate whether AI-generated code is any more correct than what a room of human developers would produce, dig into the "Phoenix" analogy of regenerating disposable code from spec instead of patching it, and raise the uncomfortable question of what happens when an integration quietly regenerates itself around a security flaw baked into the system it connects to. 

🚀 Join the Conversation
 If the AI can regenerate your code from spec whenever a bug shows up, who's actually responsible when it regenerates the same vulnerability?

FOLLOW OUR SOCIAL MEDIA:

➜Twitter: @SecTablePodcast
➜LinkedIn: The Security Table Podcast
➜YouTube: The Security Table YouTube Channel

Thanks for Listening!

Tecnologia
Inglês
Estados Unidos da América
UseAPIpara buscar este episódio emJSON

TRANSCRIÇÃO 🔗

Are you the producer of this podcast? Add a podcast transcript
Need Audio-to-Text? Transcribe with Listen411 in Just 60 Seconds

PESQUISAR EPISÓDIOS PASSADOS

Pesquisar episódios anteriores de AI Security Table.
Incorpore esta barra de pesquisa ao seu site

OUTROS EPISÓDIOS NESTE PODCAST

Mostly Dead or Mostly Back: The Zombie Resurrection of DAST in an AI World

Mostly Dead or Mostly Back: The Zombie Resurrection of DAST in an AI World

In this episode, we dig into whether DAST is dead, mostly dead, or quietly making a comeback dressed in an AI trench coat. The conversation traces the origins of dynamic application security testing from nmap scans and open source hacker tools to a market now valued at nearly four billion dollars a…
00:42:17
PARTILHAR
GUARDAR
The End of Bug Bounty As We Know It

The End of Bug Bounty As We Know It

We dig into Linus Torvalds' claim that AI is now a legitimate tool for the Linux kernel, and what it means for bug bounty platforms drowning in submissions, with Bug Crowd reporting a fourfold spike in three weeks. We debate the game theory of vulnerability disclosure, whether AI validation is…
00:42:07
PARTILHAR
GUARDAR
Don't Bury the Model T: Why STRIDE Still Drives in an AI World

Don't Bury the Model T: Why STRIDE Still Drives in an AI World

In this episode, we dig into two things the security community loves to argue about: npm finally doing the right thing and whether STRIDE has any business being called dead. The npm v12 changes gate dangerous install script behavior by default, which is a good step forward and also about a decade o…
00:59:03
PARTILHAR
GUARDAR
Make No Mistakes: Inside the First "Agentic Ransomware"

Make No Mistakes: Inside the First "Agentic Ransomware"

We dig into Sysdig's Jade Puffer report, the so-called first agentic ransomware, and argue about whether the evidence actually proves an LLM was driving the attack or if it's just a well-trained script wearing an agent costume. We walk through the four signals Sysdig points to, including …
00:43:53
PARTILHAR
GUARDAR
Aviso: O podcast e a arte incorporada nesta página são de Izar Tarandach, Matt Coles, and Chris Romeo, que é propriedade de seu proprietário e não é afiliado ou endossado por Listen Notes, Inc.
PROMOTED
Listen411.com: Lightning-fast Podcast Transcription and Summarization. Transcribe 1-hour audio in under 1 minute. Pay-as-you-go.
Listen411.com: Lightning-fast Podcast Transcription and Summarization. Transcribe 1-hour audio in under 1 minute. Pay-as-you-go.
Audio to Text

RECOMENDAÇÕES
Use a API para ir buscar este episódio