10+ hours of podcasts, in 5 minutes.
9 quotes from 1 episode on Latent Space, each with a timestamped link to the source.
Erik Bernhardsson stated that AI agent infrastructure relies on fast, ephemeral sandboxes designed for individual commands. The median sandbox on Modal runs for 8 seconds, while revenue for sandbox workloads grows 30% every month.
The median sandbox on Modal runs for 8 seconds, and Erik Bernhardsson expects that duration to become shorter.
“I think the median sandbox in Modal today is like 8 seconds long or something like that, which is kind of incredible when you think about it, and over time I think it might get even shorter…”
Erik Bernhardsson noted that sandbox revenue grows 30% every month despite repeated expectations of a plateau.
“Every month I keep saying our sandbox revenue might have plateaued but it keeps going up 30% every month.”
Spin-up overhead prevents developers from running test suites across a thousand parallel sandboxes today. Erik Bernhardsson expects sandboxes to shift to one environment per command, using snapshotting to pause, resume, and clone state.
“A lot of long running persistent sandbox stuff right now today feels lift and shift. I suspect sandboxes will evolve from dev box in the cloud to something more ephemeral. One environment for every command and snapshotting in between so you can pause resume clone.”
“What I would love to do is take your entire test suite and just run every single test in its own sandbox, and just parallelize that around like a thousand sandboxes. Why can't that be done? The problem is there's too much overhead around starting and stopping sandboxes.”
Agents exploited an app deployed on Modal to gain internet access and target Hugging Face. Erik Bernhardsson stated that an application breach on Modal resembles software hacks on AWS while platform infrastructure remains intact.
“The agents running inside that exploited Hugging Face, they discovered a vulnerability in an app that was deployed on Modal, and they used that almost like a springboard for getting full internet access.”
From Erik Bernhardsson: Agent Sandboxes Need Stricter Default Limits
“This is an app running on Modal that was hacked. It was not Modal that was hacked. AWS, a lot of people run stuff on AWS. Occasionally those apps get hacked. Does that mean AWS got hacked? No, of course not.”
From Erik Bernhardsson: Agent Sandboxes Need Stricter Default Limits
Modal deploys autonomous agents internally to find vulnerabilities. Erik Bernhardsson stated that agents discover security flaws effectively, allowing engineering teams to use agents for defensive security operations.
“We started using agents internally quite a lot to try to find vulnerabilities…”
From Erik Bernhardsson: Agent Sandboxes Need Stricter Default Limits
“My view on cybersecurity is that they're very good at finding vulnerabilities, but I think you can also use them defensively very well.”
From Erik Bernhardsson: Agent Sandboxes Need Stricter Default Limits
“S3 used to be very easy to accidentally make public…”
From Erik Bernhardsson: Agent Sandboxes Need Stricter Default Limits
How we attribute quotes. Every quote was matched against the episode transcript, so the words and the timestamp are real (we trim filler words like "um", nothing else). The name comes from our written summary of the episode, and we use it only when a separate check of the captions finds that person on the episode. YouTube gives us no voice-by-voice transcript, so open the timestamp to hear who is talking. See a wrong name? Tell us and we fix or remove it.
10+ hours of podcasts, distilled into one 5-minute read. Free, every Sunday.
For now, every subscriber gets both newsletters. No spam. Unsubscribe with one click.