March 2025
Two Dialogs on Searle
03/31/25 05:45 PM Filed in: Dialogs | Philosophy | Natural Theology | Computing | Artificial Intelligence
If a computer can beat you at chess,
it can beat you at philosophy.
The first dialog explores the fundamental mistake Searle makes in his "Chinese Room Argument", namely that while in the general case syntax is insufficient for semantics, programming languages are semantics with syntax. I have written about Searle before (the latest here), but that's me making a case to an empty courtroom. Here, I carry on an adversarial conversation with Grok 3.0 and let it judge the result.
The second dialog considers where Searle can maintain his conclusion that computers cannot be conscious without his "syntax is insufficient for semantics" pillar except by denying the Church-Turing Hypothesis, which he affirms. My position is that he can't, Grok tried to show he can.
Grok is overly verbose, yet sometimes exhibits what appears to be keen insight and sometimes says things in a way that make me jealous that it can be a better wordsmith than me. Grok doesn't get mad, doesn't get frustrated, and doesn't get tired (at least as long as I maintain my $30/month subscription). It is a far more challenging opponent than I typically have access to.
In the future, these discussions will be multi-way with humans and chatbots pushing the boundaries, with the machines keeping everyone honest - demanding definitions, uncovering tacit assumptions, noticing loops, and keeping score. We will be able to save a transcript then ask ChatGPT to read it and comment on it. New insights can be added, repetitious material ignored. The possibility of positive transform is exciting.
Dialog 1 - Searle's mistake
Dialog 2 - Searle and the Church-Turing Hypothesis
Comments
Dialog on Reframing Eden
03/28/25 10:27 AM Filed in: Dialogs | Christianity
If a computer can beat you at chess,
it can beat you at theology.
I think that the traditional reading of Eden misses two very important aspects. First, that the image of God includes the perspective "my ways are right, my will be done" and, second, that Adam was initially created alone. I think these two points change the arc of the Augustinian reading. I test this reframing of the story with Grok supporting the orthodox view.
Notice how human Grok can appear at times - how it enters loops and has to be shown that it's in a loop; how it doubles down when it has already been shown to be wrong. Unlike many humans, however, it's easier to get it to see that it's doing this and it's more amenable to changing its mind.
Grok on reframing Eden.
Dialog on Dispensationalism
03/25/25 08:09 PM Filed in: Dialogs | Christianity
If a computer can beat you at chess,
it can beat you at theology.
I have written on dispensationalism before: here, here.
Having had so much fun interacting with Grok on Limited Atonement, and the Regulative and Normative principles of worship, I engaged it with a discussion on Dispensationalism. I asked Grok to take the "in favor" side, while I took "against." Grok conceded.
Grok on dispensationalism.
Dialog on Regulative vs. Normative
If a computer can beat you at chess,
it can beat you at theology.
Having had so much fun interacting with Grok on Limited Atonement, and because a friend recently asked me if I was familiar with Primitive Baptist churches, I decided to have it debate me on the Regulative and Normative principles of worship. Grok defended the regulative principle, I defended the normative principle. Grok conceded the argument.
Grok on the regulative principle.
Dialog on Limited Atonement
If a computer can beat you at chess,
it can beat you at theology.
Computers are better at chess than they are at theology... today. That won't last too much longer. Having written on the doctrine of limited atonement before, and being intrigued by the capabilities of today's chatbots, I engaged Grok 3 in a defense of limited atonement. The exchange runs to over 36,000 words. Grok has the advantage that it doesn't get tired or frustrated and will argue until one side concedes; it can also admit when an argument is stuck in a loop. It's also interactive. An author advocating for a side in a book can't be directly challenged on what was written.
Here, then, is a link to the debate where Grok concedes that "Christ died only for the elect" has no actual Biblical support. One next step would be to reverse the dialog where I try to defend limited atonement against Grok to see if Grok also concedes. But I'm not sure I could successfully advocate for what I don't think is true. Another step would be to pit two chatbots against each other and revise the arguments until nothing more can be said.
I do wish Grok was less verbose.
Grok on limited atonement