Pete Wolfendale

@deontologistics.bsky.social

1052 Followers 189 Following

Wandering philosopher. Purveyor of Platonic heresy, Kantian computationalism, and Hegelian minimalism. (he/him/it which speaks) https://deontologistics.substack.com/

A diagram of the relationships between the main factions in Warhammer 40K, articulated by the oppositions between the Natural, the Artificial, and the Unnatural, and the derivative oppositions between Life and Death, Culture and Savagery, and Order and Chaos.
Dan Hind: Postwar British history looks a lot like the administrators of a vast global and maritime empire coming home and casually destroying one element after another of that empire's infrastructure - in the mistaken belief that they understood how to run a medium sized national economy. Pete Wolfendale: It’s not entirely unlike the regression of Rome from empire to papal city state, losing political authority but retaining centrality in a vast network of influence. Except London did it much faster, the religion is trans-national finance capital, and the indulgences tax evasion.
Screenshot of my follower count at 999
The cover of Timothy Morton's book, Hyperobjects.
Giving managers access to LLMs is like giving each of them their own personal management consultancy (derogatory).
Despite 3 years of progress, AI is still worse than me at stuff I understand, but unfathomably brilliant outside my area of expertise. Wonder why this is
An analogy helps clarify the point. Connecting a computer to cameras and robotic arms is similar to attaching measuring instruments to a simulation. The simulation now receives real-world data, but the internal variables of the model are still symbolic representations rather than the physical processes themselves. In the same way, a weather model connected to live atmospheric sensors does not become the atmosphere. It simply receives and manipulates data about it. The same principle applies to embodied Al systems. Sensors and actuators allow the system to interact with the physical world, but they do not automatically transform symbolic representations into intrinsic, experienced semantics. The system may build increasingly detailed maps of its environment, but interacting with the territory does not by itself turn the map into the territory of experience.
My reaction to this is some variant of the old Thought Slime line that 'most leftists are annoyed by leftists every single day'. This man is simply unable to be irritated by the discursive norms of Continental philosophy in the way that I am. To me, he is like a little baby.
Two guys at an izakaya holding copies of The Revenge of Reason.
Naomi Nagata
The Revenge of Reason Book Launch - Peter Wolfendale in conversation with Jacob Parkin, 25th of February 2026, 17:00-19:30, Henry Daysh Building, Floor 1 Room 1.02, Newcastle University
Picture of Angron being sad.
Woodface by Crowded House
The proof would work if the reduction were to a version of AI-by-Learning where the distribution 𝒟 is arbitrary rather than being a distribution of situation-human behavior tuples. However, if that were the case, the intractability of AI-by-Learning would only imply that there are functions that are not learnable. That fact is well-known [Val84].
The Revenge of Reason and Object-Oriented Philosophy: The Noumenon’s New Clothes
Me holding a copy of The Revenge of Reason.
Several copies of The Revenge of Reason stacked in a box.
The Revenge of Reason, front cover.