Monday, 9 November 2026

Trusted News Discovery – Powered by AI and Verified by Humans

Developer’s Ram-Altman script can email Sam Altman daily about unused RAM on Linux PCs

The Bash project from Iain Kay checks memory every five minutes, sends at most one message a day and, by design, ships without anyone's email address.

Produced with AI assistance and reviewed by the Badiyaa Editorial Desk. How we work.

Verified by humans. Checked by machines. Open to you.

Grok Fact CheckCheck it yourself in one tapVerify with Grok ↗Ask @grok on X ↗
OpenAI chief Sam Altman speaking on stage at a TED event
Image: Steve Jurvetson, CC BY 2.0 via Wikimedia Commons

A Linux script called Ram-Altman will write to OpenAI chief Sam Altman once a day to tell him how much memory is sitting idle in your computer. Its sample message opens with the subject line “You missed a spot” and reports 8.6 GiB of RAM “on my desk that isn’t in a data centre.”

The project landed on GitHub under the account of developer Iain Kay and had collected 906 stars and 21 forks when we checked it on November 9. The code itself is two commits old. The joke needed little more than that.

The remark that started it

The target is something Altman said on the Cheeky Pint podcast during a discussion of running AI models locally. The repository quotes him directly: “From a global allocation perspective, actually, it’s quite terrible to have everybody have a bunch of the world’s limited supply of DRAM sitting in their own computer.” That memory, he added, is “being used maybe 5% of the time.”

The line did not go over well with people who own computers. Altman then wrote on X that the remarks had been stripped of context: “i actually do want the world to have tons of memory and people to put as much as they want in their computers.” He also swapped his display name to Ram Altman for a while, and that is where the script gets its title.

How the script actually works

Under the gags sits a working memory monitor. It is a Bash program that reads the Linux kernel’s own figures from /proc/meminfo and is meant to be launched by cron, the decades-old Unix job scheduler. It treats “used” memory as the total minus what the kernel says programs could still claim, so reclaimable cache counts as spare.

If more than 10% of the machine’s RAM is available, it drafts a message. The FAQ explains the cutoff: “Below that, your machine has suffered enough.” Every sample also goes into a CSV log, “for when the auditors arrive.”

The arithmetic shows how little of this ever reaches an inbox. A check every five minutes means 12 runs an hour, or 288 a day, and only one of those 288 runs can send anything. The bank of text holds 8 subject lines, 10 messages and 6 sign-offs; if every combination can occur, that is 8 x 10 x 6 = 480 possible emails, enough for about 16 months of one-a-day messages without a repeat.

The part with the brakes on

The author clearly anticipated what the internet might do with this. The repository ships without anyone’s email address and asks users not to add one. It tells them to send from their own real address, not to spoof, and not to run it across a fleet of servers. Every email carries a footer offering to stop on request.

Share this story

The warning is the best line in the document: “One email a day from one person is a joke. Ten thousand of them is a denial-of-service attack with better material.”

A –dry-run option previews a message without sending it, and the README calls that the best way to enjoy the project. There is also a website that cycles through the emails with sample figures, so nobody has to install anything at all.

The README even gives Altman a fair hearing, conceding that servers really are busier than desktops. Then it closes the argument with its own verdict on the shortage question. Does this help with the DRAM shortage? “No.”

Further reading

Image: Steve Jurvetson, CC BY 2.0 via Wikimedia Commons

This article was produced with AI assistance and reviewed by our Editorial Desk. How we work.

Spot an error? Help verify this story

Join our Reader Verifiers. Every report is read by an editor.

Grok Fact CheckCheck it yourself in one tapVerify with Grok ↗Ask @grok on X ↗

Verified by humans. Checked by machines. Open to you.