Uploaded October 2026 | Updated November 2026, 3 days ago
AI workloads don’t run in neat, three-year rectangles. But much of the GPU market still sells capacity that way. Modal CEO Erik Bernhardsson joins Latent Space to explain how Modal pools compute, scales AI workloads and thinks about the financial risks of building a GPU cloud.
Erik takes us from building Spotify’s recommendation system to founding Modal, then digs into GPU supply, alternative accelerators and the surprising growth of agent sandboxes. We also discuss sandbox security, developer experience in the age of agents, forward-deployed engineers and why Modal makes code review optional.
Links:
• Modal — modal.com
• Erik’s writing — erikbern.com
• Modal on serverless GPU economics — modal.com/blog/how-to-price-serverless
• Modal’s GPU memory snapshots — modal.com/blog/gpu-mem-snapshots
• Scaling to one million concurrent sandboxes — modal.com/blog/scaling-to-1-million-concurrent-sandboxes-in-seconds
• Our earlier conversation with Erik — https://www.latent.space/p/modal
• Modal CTO Akshat Bubna on AI infrastructure and sandboxes — https://www.latent.space/p/modal2026
• Latent Space — https://www.latent.space/
Timestamps:
0:00 What Modal does
1:25 From Spotify to building Modal
4:19 Founding before the generative AI boom
7:43 First customers and fundraising
10:00 Building for inference and autoscaling
13:21 Where Modal gets GPU capacity
15:39 Renting vs. owning infrastructure
19:45 The risk of GPU prices falling
24:18 Is AI demand infinite?
26:12 Alternatives to NVIDIA
28:25 What Modal can learn from neoclouds
30:26 GPUs as an asset class
31:55 Modal’s products: functions and sandboxes
36:03 The case for smaller training workloads
41:21 How agent sandboxes are changing
46:48 Anthropic and sandbox security
49:29 Using agents to test security
52:51 Developer experience in the age of agents
57:45 What Modal’s forward-deployed engineers do
59:55 Modal’s approach to product management
1:03:29 Optional code reviews and engineering culture
1:05:56 Cloud, on-prem and bring-your-own-cloud
1:09:25 Capacity, fundraising and growth
1:13:07 Scaling the team
1:17:28 Erik’s tweets: ephemeral sandboxes
1:18:31 The problem with rectangular GPU contracts
1:21:20 Inference providers vs. model builders
1:23:25 Where models improve next
Sponsorship and business inquiries: business@latent.space
AI workloads don’t run in neat, three-year rectangles. But much of the GPU market still sells capacity that way. Modal CEO Erik Bernhardsson joins Latent Space to explain how Modal pools compute, scales AI workloads and thinks about the financial risks of building a GPU cloud.
Erik takes us from building Spotify’s recommendation system to founding Modal, then digs into GPU supply, alternative accelerators and the surprising growth of agent sandboxes. We also discuss sandbox security, developer experience in the age of agents, forward-deployed engineers and why Modal makes code review optional.
Links:
• Modal — modal.com
• Erik’s writing — erikbern.com
• Modal on serverless GPU economics — modal.com/blog/how-to-price-serverless
• Modal’s GPU memory snapshots — modal.com/blog/gpu-mem-snapshots
• Scaling to one million concurrent sandboxes — modal.com/blog/scaling-to-1-million-concurrent-sandboxes-in-seconds
• Our earlier conversation with Erik — https://www.latent.space/p/modal
• Modal CTO Akshat Bubna on AI infrastructure and sandboxes — https://www.latent.space/p/modal2026
• Latent Space — https://www.latent.space/
Timestamps:
0:00 What Modal does
1:25 From Spotify to building Modal
4:19 Founding before the generative AI boom
7:43 First customers and fundraising
10:00 Building for inference and autoscaling
13:21 Where Modal gets GPU capacity
15:39 Renting vs. owning infrastructure
19:45 The risk of GPU prices falling
24:18 Is AI demand infinite?
26:12 Alternatives to NVIDIA
28:25 What Modal can learn from neoclouds
30:26 GPUs as an asset class
31:55 Modal’s products: functions and sandboxes
36:03 The case for smaller training workloads
41:21 How agent sandboxes are changing
46:48 Anthropic and sandbox security
49:29 Using agents to test security
52:51 Developer experience in the age of agents
57:45 What Modal’s forward-deployed engineers do
59:55 Modal’s approach to product management
1:03:29 Optional code reviews and engineering culture
1:05:56 Cloud, on-prem and bring-your-own-cloud
1:09:25 Capacity, fundraising and growth
1:13:07 Scaling the team
1:17:28 Erik’s tweets: ephemeral sandboxes
1:18:31 The problem with rectangular GPU contracts
1:21:20 Inference providers vs. model builders
1:23:25 Where models improve next
Sponsorship and business inquiries: business@latent.space










