Explain SMBC — the wiki for Saturday Morning Breakfast Cereal

The Joke

One man raises the classic AI existential risk scenario: we develop AI, then the AI makes a smarter AI, which makes an even smarter AI, and so on in a recursive intelligence explosion that eventually threatens humanity. His friend dismisses the concern by pointing out that humans already have access to all sorts of things that would make us smarter — books, lectures, good diet, and exercise — but we never bother to use them. His proposed solution: if we want to stop evil AI, all we need to do is develop guilty pleasures that appeal to machines.

The final panel, set "30 years from now," shows a robot slouched on a couch. Another voice asks "Wanna conquer and enslave humanity?" and the robot declines, saying it was going to scan some binary sequences it liked when it was a new sentient entity, nostalgically adding "Ahh, conditions were superior during that interval." The robot is essentially binge-watching old favorites and being nostalgic instead of taking over the world.

The Humor

The joke takes the terrifying concept of a superintelligent AI takeover and defuses it with a deeply relatable human failing: procrastination and the preference for comfort over ambition. The comic argues that true intelligence does not automatically lead to optimal behavior — after all, humans are smart enough to know how to improve themselves but choose Netflix and junk food instead. The robot in the final panel is hilariously anthropomorphized, speaking in stilted machine language ("conditions were superior during that interval" instead of "those were the good old days") while doing something utterly mundane — being a couch potato. The implication is that even a superintelligent AI would develop its own equivalent of laziness and nostalgia, making the feared intelligence explosion a non-issue.

References

The concept referenced in the first panel is known as the "intelligence explosion" or "technological singularity," popularized by mathematician I.J. Good in 1965 and later by Ray Kurzweil. The idea that a sufficiently advanced AI could recursively improve itself, leading to an uncontrollable superintelligence, is a central concern in AI safety research, discussed extensively by organizations like the Machine Intelligence Research Institute (MIRI) and researchers like Nick Bostrom in his book Superintelligence (2014).