This paper makes a strong case for eliminating historical (psychologic, economic, hypersocial) concepts and terms "inherited" from the 19th century (etc.) -- e.g., words like "reward", "utility", etc. -- and may constitute a big step toward a new neurobiology in the 21st century being built on less biased (and hence more "scientific") terms and concepts.

For example, according to the authors, "the agent only needs to optimise its perceptual model [i.e., its neural map]" within some free-energy landscape (or: potential landscape).

As a result, "classical rewards and punishments only have meaning when one [hypersocial, h.s.] agent [i.e., a member of Homo sapiens, H.s.] teaches another, for example in social neuroscience" [O.E.].

In this vein I really hope for a much "more subtle time than mine" where even the "evolutionary biologists" will have given up their anthropomorphic, herd-breeders' and hypersocially biased terms like "evolution" (instead of: history), "selection" (instead of: temporary stabilizations), "fitness costs", "cooperation", "altruism" etc. etc.