korrents

← Topics

Rent the cloud or own the servers?

Top voices on the cloud: David Heinemeier Hansson, David Crawshaw, Pieter Levels, George Hotz and 3 others.

They argue over what the cloud really saves, and where the next wave of software will run.

14 dated claims from 7 people · 4 views · most recent 22 Sept 2026

The views

What is a view?

Our reading of what a group of claims have in common, one level above a korrent: a korrent is our reading of what somebody said, a view is our reading of what several korrents share. The claims are printed under it, so the sentence can be checked against them.

Nobody is filed under a view — claims are. A person is named here because a dated claim of theirs sits under it, and somebody whose claims sit under two views appears under both.

  1. The cloud is neither easier nor cheaper, and owned hardware is not more work

    David Heinemeier Hansson

  2. The cloud's basic units are the wrong shape, and more layers on top cannot fix that

    David Crawshaw

  3. Agents and AI will run in the cloud first, not on your own machine

    M.G. SieglerPieter Levels

  4. What runs your software, and your AI, should be hardware you own

    Meng ToDavid Heinemeier HanssonGeorge Hotz

A viewour reading

The cloud is neither easier nor cheaper, and owned hardware is not more work

6 claims filed here, newest first — each in the words of whoever said it.

The cloud pitch of easier, cheaper, faster is false on two counts: it is not easier and not cheaper, and the speed that is real is overvalued.

  1. David Heinemeier Hansson Creator of Ruby on Rails, CTO of 37signals It's not easier to use AWS than it is to run your own machines, which we learned when we pulled out the cloud and didn't hire a single extra person. Even though we operate all our own hardware, the team stayed exactly the same. DHH: Future of Programming, AI, Ruby on Rails, Productivity & Parenting | Lex Fridman Podcast #474youtube.com · 12 Jul 2025 · 3:20:32 into the videoAll korrents from this video
    It's not easier to use AWS than it is to run your own machines, which we learned when we pulled out the cloud and didn't hire a single extra person. Even though we operate all our own hardware, the team stayed exactly the same.
Show all 6 claims

Leaving the cloud for owned hardware cut 37signals' annual infrastructure bill from $3.2M to $1.3M without growing the operations team.

  1. David Heinemeier Hansson Creator of Ruby on Rails, CTO of 37signals For 2024, we've brought the cloud bill down from the original $3.2 million/year run rate to $1.3 million. That's a saving of almost two million dollars per year for our setup! Our cloud-exit savings will now top ten million over five yearsworld.hey.com · 17 Oct 2024
    For 2024, we've brought the cloud bill down from the original $3.2 million/year run rate to $1.3 million. That's a saving of almost two million dollars per year for our setup!

Running a business on owned hardware is not more work than running it on the cloud

  1. David Heinemeier Hansson Creator of Ruby on Rails, CTO of 37signals Because running things on premise is not a materially different scope of work from running it in the cloud The price of managed cloud servicesworld.hey.com · 16 Oct 2023All korrents from this piece
    Because running things on premise is not a materially different scope of work from running it in the cloud

The cloud is often just as complicated as running things yourself, and it's usually ridiculously more expensive

  1. David Heinemeier Hansson Creator of Ruby on Rails, CTO of 37signals The cloud is often just as complicated as running things yourself, and it's usually ridiculously more expensive We have left the cloudworld.hey.com · 23 Jun 2023All korrents from this piece
    The cloud is often just as complicated as running things yourself, and it's usually ridiculously more expensive

Not using own servers is financial malpractice for mid-sized SaaS businesses

  1. David Heinemeier Hansson Creator of Ruby on Rails, CTO of 37signals Any mid-sized SaaS business and above with stable work loads that does not benchmark their rental bill for servers in the cloud against buying their own boxes is committing financial malpractice at this point We stand to save $7m over five years from our cloud exitworld.hey.com · 21 Feb 2023All korrents from this piece
    Any mid-sized SaaS business and above with stable work loads that does not benchmark their rental bill for servers in the cloud against buying their own boxes is committing financial malpractice at this point

A viewour reading

The cloud's basic units are the wrong shape, and more layers on top cannot fix that

2 claims filed here, newest first — each in the words of whoever said it.

Cloud abstraction problems cannot be fixed by adding more abstraction layers on top.

Leaky abstractions

  1. David Crawshaw Software engineer; co-founder and former CTO of Tailscale You cannot solve the fundamental problems with cloud abstractions by building new abstractions on top. I am building a cloudcrawshaw.io · 22 Apr 2026All korrents from this piece
    You cannot solve the fundamental problems with cloud abstractions by building new abstractions on top.

A viewour reading

Agents and AI will run in the cloud first, not on your own machine

2 claims filed here, newest first — each in the words of whoever said it.

AI will likely be run in the cloud rather than on native AI PCs.

  1. M.G. Siegler Technology writer of Spyglass my bet would be that there is no truly native 'AI PC' and that instead, people will run AI on the current computing paradigms as long as they're in the cloud. Soylent Googlebook is Androidspyglass.org · 22 Sept 2026
    my bet would be that there is no truly native 'AI PC' and that instead, people will run AI on the current computing paradigms as long as they're in the cloud.

AI agents and AI coding will run on servers and from the cloud first, not on your laptop.

  1. Pieter Levels Dutch bootstrapper who builds and runs his internet startups alone I agree with @theo completely, it's clear to me this is where it's going, also seeing @karpathy with Claude moving to the cloud (via Slack etc), I think AI "agents" and AI coding will operate on servers / from the cloud first I've been coding almost solely on my VPS with Claude Code for almost a year nowlevels.io · 28 Jun 2026
    I agree with @theo completely, it's clear to me this is where it's going, also seeing @karpathy with Claude moving to the cloud (via Slack etc), I think AI "agents" and AI coding will operate on servers / from the cloud first

A viewour reading

What runs your software, and your AI, should be hardware you own

3 claims filed here, newest first — each in the words of whoever said it.

Agents working locally can be far more powerful than any private walled garden that exists out there in the cloud.

  1. Meng To Designer and founder of Design+Code because Codex works really well locally and agents can do it far more powerful than any private walled garden that exists out there in the cloud Meng To - Codex superpowers for designersyoutube.com · 30 Jun 2026 · 10:22 into the videoAll korrents from this video
    because Codex works really well locally and agents can do it far more powerful than any private walled garden that exists out there in the cloud

    Watch from 10:22 plays here

    ↗Meng To - Codex superpowers for designersyoutube.com

    30 Jun 2026 · video · 35m · spoken · machine transcript

Show all 3 claims

After the cloud exit, the next step is out of the data centre: home servers and multi-gigabit home connections make the garage startup literal again.

  1. David Heinemeier Hansson Creator of Ruby on Rails, CTO of 37signals Now my next kind of tingle is that if you could move out of the cloud, can you also move out of the data center? Personal servers have gotten really scarily quick and efficient and personal internet connections rival what we connected data centers with just a decade or two ago. DHH: Future of Programming, AI, Ruby on Rails, Productivity & Parenting | Lex Fridman Podcast #474youtube.com · 12 Jul 2025 · 3:28:33 into the videoAll korrents from this video
    Now my next kind of tingle is that if you could move out of the cloud, can you also move out of the data center? Personal servers have gotten really scarily quick and efficient and personal internet connections rival what we connected data centers with just a decade or two ago.

An AI you rent from a big company in the cloud is fundamentally aligned to that company and not to you, which is why the hardware has to be yours.

  1. George Hotz American programmer and hacker Of course, when you talk to an AI that’s made by a big company in the cloud, the AI fundamentally is aligned to them, not to you. And that’s why you have to buy a tiny box. So you make sure the AI stays aligned to you. George Hotz: Tiny Corp, Twitter, AI Safety, Self-Driving, GPT, AGI & God | Lex Fridman Podcast #387youtube.com · 29 Jun 2023 · 2:37:23 into the videoAll korrents from this video
    Of course, when you talk to an AI that’s made by a big company in the cloud, the AI fundamentally is aligned to them, not to you. And that’s why you have to buy a tiny box. So you make sure the AI stays aligned to you.

Read, and under none of the views

On the topic, but saying something the views above do not cover.

Sharding across machines has faded as a concern because hardware got more powerful, not because the cloud made it easy.

  1. Martin Kleppmann Computer science researcher at Cambridge But, at least this sort of sharding across multiple machines is maybe become less of a pressing issue just because more and more workloads can just run on a single machine. Some people still have very large-scale workloads that do have to be sharded across multiple machines. So, it's not going away entirely. Designing Data-intensive Applications with Martin Kleppmannyoutube.com · 22 Apr 2026 · 0:42:22 into the videoAll korrents from this video
    But, at least this sort of sharding across multiple machines is maybe become less of a pressing issue just because more and more workloads can just run on a single machine. Some people still have very large-scale workloads that do have to be sharded across multiple machines. So, it's not going away entirely.

    Watch from 0:42:22 plays here

    ↗Designing Data-intensive Applications with Martin Kleppmannyoutube.com

    22 Apr 2026 · video · 1h 25m · spoken · machine transcript

On this topic, not yet placed

Matched by wording, not yet read into a view. The archive grows on its own, so this is where new claims wait for somebody to read them.

The full chronology, with every quote: the cloud as a timeline · person by person, side by side · all topics · Something wrong? Tell us