Published on November 3, 2025 5:32 PM GMT
Book review: Red Heart, by Max Harms.
Red Heart resembles in important ways some of the early James Bondmovies, but it's more intellectually sophisticated than that.
It's both more interesting and more realistic than CrystalSociety(the only prior book of Harms' that I've read). It pays carefulattention to issues involving AI that are likely to affect the worldsoon, but mostly prioritizes a good story over serious analysis.
I was expecting to think of Red Heart as science fiction. It turned outto be borderline between science fiction and historical fiction. It'sset in an alternate timeline, but with only small changes from what theworld looks like in 2025. The publicly available AIs are probably almostthe same as what we're using today. So it's hard to tell whetherthere's anything meaningfully fictional about this world.
The "science fiction" part of the story consists of a secret AIproject that has reportedly advanced due to unusual diligence atapplying small, presumably mundane, efficiencies. That's only a littledifferent from what DeepSeek's AI sounded like last winter. In order tobe fully realistic, it would also need some sort of advance along thelines of continuallearning. The bookis vague enough here that it might be assuming that other AI projectshave implemented some such advance. That only stretches the realism asmall amount.
Amazon quite reasonably classifies the book as a political thriller,even though it focuses more on artificial intelligence than on politicsin the usual sense.
My biggest complaint is that the story occasionally mentions that the AIis rapidly becoming more capable, yet I didn't get a clear sense ofthis speed. There are almost no examples of her trainers being surprisedthat she succeeded at some new task that had previously looked hard forher. There is no indication of when she crosses any key threshold,except when they give her new permissions.
Maybe much of that is realistic. The sudden capabilities foom of somefictional AIs seems too dramatic to satisfy my desire for realism. Butthat leaves the reader with confusing signs about the extent to whichthere's a race between competing AI projects. The story stretches outover a longer period than I'd expect if they genuinely felt the urgencythat their discussions suggest.
I would like to know what kind of evidence is driving the reports ofurgency. But I can imagine that realistic versions of the evidence wouldbe too subtle to readily understand. And I wouldn't have wanted thestory to fabricate unrealistically blatant breakthroughs in order tosupport the sense of urgency.
The story alternates between sometimes portraying the hero as anordinary person, while at other times he looks like a mild version ofJames Bond.
He's sufficiently young and inexperienced that this could have been acoming of age story. But we don't see him growing. Whatever growth heneeded likely happened before the start of the story. The author seemsto want to emphasize that there's a lot of luck needed for the story tohave a nice ending. It may be important to hire the best and thebrightest to handle an AI project, but the odds will still be lower thanwe want.
The story's hero needed to have several key skills, but most of thetime he doesn't look special. It seems mostly like an accident that heends up imitating James Bond. This approach mostly works, but feelsstrange. It makes the story a bit more realistic, at a modest cost tothe story's entertainment value.
There's one minor spot that felt implausible. Near the middle, hethinks that he will be leaving China soon, and his main reaction is toworry about his relationships with minor characters. What, no emotionsrelated to leaving the most important project ever? It's not like hehas an unemotional personality.
The main reason that I read Red Heart is its discussion of AIcorrigibility (roughly: obedience),which I consider to be a critical and neglected part of how superhumanAI can be safe.
The story provides a decent depiction of how corrigibility would work ifit's implemented well. But it doesn't provide enough detail tosubstitute for reading more rigorous technical writings.
The book's treatment of multi-principal corrigibility is frustratinglybrief but raises crucial questions. If we successfully build corrigibleAGI, to whom should it be corrigible? The story gestures at problemswith being corrigible to multiple people, but it implies, without muchjustification, that we might need to give up on the goal of having alarge number of people empowered to influence the leading AI.
Red Heart is refreshing and a mostly realistic complement to theexcessive gloom of If Anyone Builds It, EveryoneDies.
Discuss
