A developer based in Edinburgh has published an open-source Linux Bash script that monitors a computer’s available memory and can send OpenAI CEO Sam Altman a single daily email reporting any unused RAM, complete with rotating humorous messages.
The project, called Ram-Altman, appeared on GitHub under the handle IainKay in early November 2026 and quickly drew attention after coverage on social media and tech sites. Its creator, Iain Kay, lists himself as a full-stack web developer, DevOps lead and network operations manager. The script checks memory usage every five minutes through cron. If more than 10 percent of the system’s RAM is available, it prepares one carefully worded email per day.
Kay designed the tool as a direct response to comments Altman made in late October. On the Oct. 29 episode of the Cheeky Pint podcast, hosted by Stripe co-founder John Collison, Altman discussed the global DRAM shortage driven by AI data-center demand. Speaking about people who want to run open-source models locally, he said: “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 and being used 5% of the time. You want maximum utilization in the cloud if you’re trying to deal with this very supply constrained environment.”
The 5 percent figure was Altman’s own estimate. The remarks triggered swift online backlash, including memes and criticism that OpenAI itself is a major buyer of memory. On Nov. 4, Altman addressed the reaction on X. “I actually do want the world to have tons of memory and people to put as much as they want in their computers, and it is fucking annoying that this was taken out of context,” he wrote. “I have been arguing for the world to ramp up memory production for a long time.” He briefly changed his display name to Ram Altman, a move that directly inspired the script’s title.
The script uses the Linux kernel’s MemAvailable figure, which counts free memory plus reclaimable page cache. It logs samples to a CSV file, rotates eight subject lines, ten message bodies and six sign-offs, and includes a dry-run mode plus a test suite. No email address ships with the repository. Users must configure their own sender and recipient settings. Kay warned that one message a day is a joke; thousands of them would amount to a denial-of-service attack.
One sample email opens with the subject “You missed a spot” and reads: “Please don’t be alarmed, but there is 8.6 GiB of RAM on my desk that isn’t in a data centre. That’s 56.2% of what’s installed here, and as far as I can tell it has no plans today. Hoarding responsibly, A concerned RAM owner.” Another subject line used in coverage was “Daily RAM census: some survived.” The project site at ram-altman.lol cycles through sample messages without requiring installation.
Kay’s README is packed with dry sysadmin humor. It notes that every systems administrator has spent years explaining that free RAM is not wasted and that Linux does not “eat” memory. “It took decades for someone important to agree with the second part, and he’d like it shipped to him,” the file states. Counting reclaimable cache as available is described as “a generous concession to the other side. Think of it as a plea bargain.” The 10 percent threshold exists “because below that, your machine has suffered enough.” An Electron version, it adds, “would have solved the problem by using all the RAM.”
The script arrives against a backdrop of tight DRAM supply. AI data-center demand has pushed consumer memory prices sharply higher since mid-2025, with some kits rising several-fold. Manufacturers have shifted capacity toward high-bandwidth memory and server DRAM. Micron left the consumer Crucial brand to focus on data-center customers. Analysts have warned that tight conditions could last into 2027 or beyond. OpenAI’s Stargate plans have been linked to projected demand of up to 900,000 DRAM wafers a month under earlier letters of intent with Samsung and SK hynix, a volume some analyses equated to roughly 40 percent of global output based on 2025 estimates. Those figures remain projections rather than locked volumes.
Kay’s project sits at the intersection of precise kernel accounting and Silicon Valley resource talk. The original podcast discussion centered on local inference competing with cloud utilization under scarcity, not a call to remove RAM from ordinary personal computers. By shipping no address, limiting messages to one per day and promising to stop if asked, the developer turned a potential harassment tool into controlled satire. The code is available at github.com/IainKay/ram-altman.
