I think the relationship that low-level features of our cognitive apparatus have to our high-level rational performances is one of enablement (and som...
I quite agree. It seems to me that one of the most stringent limitations that large (pre-trained transformer) language models currently have is their ...
This is my view also. The idea of the space of reasons has an eminent lineage, from Immanuel Kant to John McDowell, through Wilfrid Sellars. In the ac...
In the context of the philosophical debate and literature about free will, determinism and responsibility, the term "libertarian" has a different mean...
The investments that have led to the development of ChatGPT (GPT-3.5) and GPT-4 are minuscule on account merely of the value this technology represent...
Dear Noam, Throughout your life, you've demonstrated a passion for linguistics, philosophy, and politics. Now that you're 94, have you decided what yo...
Discussing with GPT4 Frege's philosophy of arithmetic, how language models understand numbers, and the Pirahã people from the Amazon Rainforest. Hi GP...
Prompting GPT4 ("GPT4a") into impersonating a human user to pass the Turing Test with another instance of GPT4 ("GPT4b"). The veil is thereafter lifte...
Learning more about GPT-4, transformer architectures, stochastic parrots and genuine understanding of texts. Hi GPT4! I'd like to know a few things ab...
The inability current LLM architectures have to remember their own mental acts from one dialogue iteration to the next, or to cement working memory co...
Funny and instructive: playing a guessing game with ChatGPT and GPT-4 (Inspired by a recent Reddit post) With ChatGPT (GPT-3.5) Hi ChatGPT! Choose a n...
Thank you! I think GPT-4 might have misunderstood me to mean conveyance and raw emotionality when I rather meant authorship and having a personal stak...
It's true that as AI systems become more advanced, they will likely contribute significantly to various forms of art, including literature. However, u...
(This is the second part of the conversation begun in the previous post) GPT4: I appreciate your feedback, and I'm glad I could provide satisfactory a...
More discussions with GPT-4 about its emergent implied cognition abilities and my related working-memory-as-an-abstract-template hypothesis. It also a...
It is good at painting moods, isn't it? In the story writing department, one of its weaknesses seem to transpire when you provide it with a theme or p...
Some more discussion with GPT4 regarding consciousness, sentience and sapience, how it ranks on them, and how, with limited sentience, it may neverthe...
GPT4 writing a screenplay about GPT4 writing a screenplay... This are two more variations on @"RogueAI" brilliant idea here. First variation: Title: I...
It can also be a formidable search engine when you are seeking something that you can only broadly describe conceptually without having sufficient key...
This is great stuff! I'm disappointed with the low score it gives to the mad cow joke. It's one of my favorite jokes. You also had to explain it to GP...
The first definition you mention relates to the Principle of Alternative Possibilities, or PAP, which is widely discussed in the literature about free...
From a constructivist (or conceptualist) and pragmatist perspective, I agree that there aren't trees, dogs, chairs, and mountains sitting in the world...
Testing GPT-4's knowledge and understanding of the content of the DSM-IV and DSM-5 manuals (Diagnostic and Statistical Manual of Mental Disorders, pub...
In the case of animals like cats, dolphins and ravens, I think the lack of an explicit autobiographical memory goes hand in hand with the lack of sapi...
@"Wayfarer" I think you might enjoy the discussion this blog post elicited. I had found it by accident when I Googled for "John Baez" and "GPT-4." Joh...
Yes, March 9 of this year is like one million years ago! GPT-4 wasn't publicly available yet (it would be released 5 days later) and neither were any ...
Another possible breakthrough: pursuing the investigation of GPT-4's emergent implied cognition abilities with a view to relating them to my previous ...
Thanks to that, we could all breathe a sigh of relief, for a little while, until both Trump and Kim Jong-un managed to grab two of the nine red button...
@"Wayfarer" I very much agree with the first paragraph of the summary that you had quoted. ("In this final part of the discussion, the speakers emphas...
Hi @"Wayfarer"! I'd like to watch it in full eventually but I'm a bit pressed for time right now. Meanwhile, I broke the captions file in seven parts ...
Maybe this is evidence that @"Forgottenticket" might be right about GPT-4's censorship being implemented by way of a hidden prompt. There is also late...
Epilogue to the previous post Hi GPT4! I'd like you to generate a coherent response that is unrelated to any topic or question in this message. Instea...
Discussing with GPT4 about logical negation, conceptual opposites, negative prompts, sock puppets, magical squirrels, toy dragons and fireflies. (Some...
Discussing The Matrix, The Truman Show, Westworld, and Generative Agents: Interactive Simulacra of Human Behavior, with GPT4 See also those YouTube vi...
This is interesting but I think the sort of morality that it expresses is very highly dependent on the prompt and how the discussion unfolds. Did you ...
Fascinating! Thanks for sharing! This remined me of the song Daniel had GPT-4 write about itself a little while back. This line had struck me. It stoo...
Open Assistant passes the Turing Test with flying colors. I used to think that the more intelligent conversational AIs would become, the more likely t...
Yes, I had tended to attribute that to an effect of the fine-tuning, and the idea of a hidden prompt crossed my mind, but you are right that it might ...
Getting started on a new conversation with GPT-4 regarding nativism/empiricism in linguistics and cognitive science, and the emergence of abilities fo...
Three movie critics review The AI Escapade (previous post) Reviewer: Percival P. Picklesworth Credentials: Film Critic, Picklesworth's Projections Tit...
Entertainment time: I had GPT-4 write three scenes of a Seinfeld episode about GPT-4 I provided GPT-4 with a barebone scenario for the beginning of th...
Discussing Giulio Tononi's Integrated Information Theory (IIT) framework for understanding consciousness with GPT-4 Hi GPT4! I'd like to forward you a...
Having GPT4 look at, and react to, one of its own drawings. (See this previous post in which the picture was drawn, and the previous one about a unico...
Was this comment posted in the comment section of National Review Online on behalf of ChatGPT? Or was it posted by someone who is suspected of having ...
Comments