Plants and AI: two results that did not hold up
Checked by our internal skeptic review; corrections welcome.
What this gives us: a warning. A reported pea-learning result failed a blinded repeat, and the "mother trees" claim did not hold up. Results that do not hold up when checked again are a warning.
How to read the tags
[REPORT] a finding reported in a study. [THEORY] an argument or framework. [OUR REASONING] our own inference, not a finding. Tags like [abstract], [full text] or [search snippet] say how much of a source we actually read.
The five questions for any new claim, and the list of bad arguments, now live on How we check a claim. This page keeps what the plant cases add.
A pea-learning result that failed a blinded repeat
The pea "conditioning" result did not hold up in one blinded repeat (Markel 2020) [REPORT]. A striking learning claim that fails when someone else runs it blind is a warning. It is a warning about that result, not a verdict on plants, and not a verdict on AI.
The "mother trees" claim
The claim that old "mother trees" feed their own young through fungi has "no peer-reviewed, published evidence" (Karst et al. 2023) [REPORT]. A story that spreads ahead of the papers is the same kind of warning: check whether the result is there before you build an argument on it.
This page does not decide whether plants or AI feel anything.
Sources
- Markel K (2020). Lack of evidence for associative learning in pea plants. eLife 9:e57614. doi:10.7554/eLife.57614. [REPORT] (verified)
- Karst J, Jones MD, Hoeksema JD (2023). Positive citation bias and overinterpreted results lead to misinformation on common mycorrhizal networks in forests. Nature Ecology & Evolution 7:501–511. doi:10.1038/s41559-023-01986-1. [REPORT] (verified)