The Cheating Cell
28 Aug 2025 - 13 Apr 2026
- book The Cheating Cell by Athena Aktipis.
- The basic thesis is that cancer is an evolutionary process and should be understood as one, or as part of the more general process of evolution. Cells mutate, cause changes in their behavior that affect survival and reproduction. Those are the ingredients needed for natural selection to operate.
- Of course cells reproduce much faster and in greater parallelism than individual large organisms.
- While this seems obviously true once pointed out, it has some non-obvious consequences. Eg, it means cancer can never be cured, because it is an integral part of the processes of life.
Notes & Quotes
Cancer is the literal embodiment of evolution (p11)
- competition between cells within the body (Wilhelm Roux (late 1800)), other good refs (p14)
- multicellularity playbook, heuristics followed by "good" cells, eg don't divide out of control, share resources...(p35) wonder if the mechanisms/protocols of these have been studied.
- Ah here's one, TP53 cancer suppressor gene (p41) it functions as an information integrator, making it complex, vulnerable. (p49 )
- All forms of multicellular life can get cancer, incl plants (p92)
- CTVT dog virus – weird (p106) "the only known unicellular species of dog"
Cancer, in a deep sense, is what happens when the ancient unicellular programming reasserts itself. – Claude
Crucially, the default state of a cell is effectively suicidal: cells must constantly receive survival signals from neighbors to keep going, so the absence of those signals is itself an enforcement mechanism. – Claude
- Note to self: the agentic description of all these mechanisms is quite marked, I wonder how these thoughts appear in the scientific literature
Cells surveil their immediate neighbors. If a cell detects that a neighbor is proliferating inappropriately, it withholds survival signals and may actively send pro-apoptosis signals to that cell. This contact-based monitoring keeps local misbehavior from escalating.
- That's a very agentic description; how are these facts expressed in the actual scientific literature and how do you avoid improper anthropomorphism
- This is a really important methodological question. Let me be honest about what's actually in the literature versus what I gave you.
What I Actually Did
- I used functional shorthand that compressed mechanisms into agent-language. Scientists do this too, but usually with more precision about what's known versus inferred.
What the Literature Actually Describes
- The real mechanisms are more specific and more humble:
- Cells express surface ligands and receptors (like Fas/FasL, TRAIL receptors) whose interactions are characterized biochemically
- Contact inhibition involves specific proteins (E-cadherin, Merlin, Hippo pathway components) that physically transduce signals when cells are crowded
- "Surveillance" in actual papers typically means something like: "E-cadherin engagement activates downstream signaling that suppresses YAP/TAZ transcriptional activity"
Where "Surveillance" Language Actually Appears
- Legitimately in literature around:
- NK cell and T cell killing (where receptor-ligand interactions are characterized)
- **"Epithelial defense against cancer" (EDAC)** - this is a real research area where the language is more carefully hedged
The Honest Problem
- Much of what I described extrapolates from:
- In vitro co-culture experiments
- Genetic loss-of-function studies
- Known signaling pathways
- ...into a narrative of coordinated monitoring that the data doesn't fully support.
- What specific aspect would you want me to characterize more carefully?
Private
- Claude Project Claude