Jon Awbrey

Inquiry Into Inquiry • On Initiative 1
• https://inquiryintoinquiry.com/2022/07/10/inquiry-into-inquiry-on-initiative-1/

Re: R.J. Lipton and K.W. Regan • Sorting and Proving
• https://rjlipton.wpcomstaging.com/2022/06/13/sorting-and-proving/

❝GPT‑3 works by playing a game of “guess the next word” in a phrase.
This is akin to “guess the next move” in chess and other games, and
we will have more to say about it.❞

My Observation —
• https://rjlipton.wpcomstaging.com/2022/06/13/sorting-and-proving/#comment-119056

As a person who struggles on a daily basis to rise to the level of sentience
I've learned it has more to do with beginning than ending this sentence.

Resources —

Survey of Inquiry Driven Systems
• https://inquiryintoinquiry.com/2023/04/23/survey-of-inquiry-driven-systems-5/

Survey of Theme One Program
• https://inquiryintoinquiry.com/2023/03/30/survey-of-theme-one-program-5/

#Peirce #Logic #Inquiry #InquiryDrivenSystems #InquiryIntoInquiry
#BooleanSatisfiability #ComputationalComplexity #GödelsLostLetter
#AbductionDeductionInduction #IntelligentSystems #GPT #LLM #SAT

Inquiry Into Inquiry • On Initiative 1

Inquiry Into Inquiry
May 19, 2023 at 1:16pmWeb
Show threadJon AwbreyMay 20, 2023

Inquiry Into Inquiry • On Initiative 2
• https://inquiryintoinquiry.com/2022/07/20/inquiry-into-inquiry-on-initiative-2/

Re: Scott Aaronson
• https://scottaaronson.blog/?p=6524
• https://scottaaronson.blog/?p=6534
• https://scottaaronson.blog/?p=6541

❝Personally, I'd give neither of them [Bohr or Einstein] perfect marks, in part because they not only both missed Bell's Theorem, but failed even to ask the requisite question (namely: what empirically verifiable tasks can Alice and Bob use entanglement to do, that they couldn't have done without entanglement?). But I'd give both of them very high marks for, y'know, still being Albert Einstein and Niels Bohr.❞

To Ask The Requisite Question —
• https://scottaaronson.blog/?p=6541#comment-1940887

This brings me to the question I was going to ask on the AI post, but was afraid to ask.

Does GPT‑3 ever ask an original question on its own?

Simply asking for clarification of an interlocutor's prompt is not insignificant but I'm really interested in something more spontaneous and “self‑starting” than that. Does it ever wake up one morning, as it were, and find itself in a “state of question”, a state of doubt or uncertainty so compelling as to bring it to ask on its own initiative what we might recognize as a novel question?

Resources —

Survey of Inquiry Driven Systems
• https://inquiryintoinquiry.com/2023/04/23/survey-of-inquiry-driven-systems-5/

Survey of Theme One Program
• https://inquiryintoinquiry.com/2023/03/30/survey-of-theme-one-program-5/

#Peirce #Inquiry #InquiryDrivenSystems #InquiryIntoInquiry
#ScottAaronson #AI #ChatGPT #GPT #LLM #LargeLanguageModels

Inquiry Into Inquiry • On Initiative 2

Inquiry Into Inquiry
Show threadJon AwbreyMay 20, 2023

Inquiry Into Inquiry • On Initiative 3
• https://inquiryintoinquiry.com/2023/05/01/inquiry-into-inquiry-on-initiative-3/

Re: Scott Aaronson • Should GPT Exist?
• https://scottaaronson.blog/?p=7042

My Comment —
• https://scottaaronson.blog/?p=7042#comment-1946961

The more fundamental problem I see here is the failure to grasp the nature of the task at hand, and this I attribute not to a program but to its developers.

Journalism, Research, and Scholarship are not matters of generating probable responses to prompts or other stimuli. What matters is producing evidentiary and logical supports for statements. That is the task requirement the developers of recent LLM‑Bots are failing to grasp.

There is nothing new about that failure. There is a long history of attempts to account for intelligence and indeed the workings of scientific inquiry based on the principles of associationism, behaviorism, connectionism, and theories of that order. But the relationship of empirical evidence, logical inference, and scientific information is more complex and intricate than is dreamt of in those reductive philosophies.

Note. The above comment was originally posted on March 1st but appears to have been deleted accidentally.

#Peirce #Inquiry #InquiryDrivenSystems #InquiryIntoInquiry
#ScottAaronson #AI #ChatGPT #GPT #LLM #LargeLanguageModels

Inquiry Into Inquiry • On Initiative 3

Inquiry Into Inquiry
Show threadJon AwbreyMay 20, 2023

Inquiry Into Inquiry • On Initiative 4
• https://inquiryintoinquiry.com/2023/05/15/inquiry-into-inquiry-on-initiative-4/

Re: Terry Tao • PCAST Working Group on Generative AI Invites Public Input
• https://terrytao.wordpress.com/2023/05/13/pcast-working-group-on-generative-ai-invites-public-input/

My Comment —
• https://terrytao.wordpress.com/2023/05/13/pcast-working-group-on-generative-ai-invites-public-input/#comment-678057

I think a lot of people who've been working all along on AI, intelligent systems, and computational extensions of human capacities in general are a little distressed to see the field cornered and re‑branded in the short‑sighted, market‑driven way we currently see.

The more fundamental problem I see here is the failure to grasp the nature of the task at hand, and this I attribute not to a program but to its developers.

Journalism, Research, and Scholarship are not matters of generating probable responses to prompts or other stimuli. What matters is producing evidentiary and logical supports for statements. That is the task requirement the developers of recent LLM‑Bots are failing to grasp.

There is nothing new about that failure. There is a long history of attempts to account for intelligence and indeed the workings of scientific inquiry based on the principles of associationism, behaviorism, connectionism, and theories of that order. But the relationship of empirical evidence, logical inference, and scientific information is more complex and intricate than is dreamt of in those reductive philosophies.

#Peirce #Logic #Inquiry #InquiryDrivenSystems #InquiryIntoInquiry
#TerryTao #PCAST #AI #IntelligentSystems #LLM #LargeLanguageModels

Inquiry Into Inquiry • On Initiative 4

Re: Terry Tao • PCAST Working Group on Generative AI Invites Public Input My Comment — I think a lot of people who’ve been working all along on AI, intelligent systems, and computa…

Inquiry Into Inquiry
Show threadJon AwbreyMay 22, 2023

Inquiry Into Inquiry • On Initiative 5
• https://inquiryintoinquiry.com/2023/05/22/inquiry-into-inquiry-on-initiative-5/

Re: Inquiry Into Inquiry • On Initiative 2
• https://inquiryintoinquiry.com/2022/07/20/inquiry-into-inquiry-on-initiative-2/

Re: Mathstodon • Joeri Sebrechts
• https://mathstodon.xyz/@joeri_s@mstdn.social/110401673746671834

JS:
❝That's not how it works. The model lacks agency. It is a machine whose gears are cranked by the user's prompt. It can ask questions, but only when prompted to. It is not doing anything at all when it isn't being prompted.❞

Sure, I understand that. The hedge “as it were” is used advisedly for the sake of the argument. (I wrote my own language learner back in the 80s.)

Speaking less metaphorically, the program and its database are always in their respective states and the program has the capacity to act on the database even when not engaged with external prompts.

Is there any reason why the program's “housekeeping” functions should not include one to measure its current state of “uncertainty” (entropy of a distribution) with regard to potential questions — or any reason why it should “hurt to ask”?

As it were …

#Peirce #Inquiry #InquiryDrivenSystems #InquiryIntoInquiry
#ScottAaronson #AI #ChatGPT #GPT #LLM #LargeLanguageModels

Inquiry Into Inquiry • On Initiative 5

Inquiry Into Inquiry
Show threadJoeri SebrechtsMay 20, 2023
@Inquiry that’s not how it works. The model lacks agency. It is a machine whose gears are cranked by the user’s prompt. It can ask questions, but only when prompted to. It is not doing anything at all when it isn’t being prompted.
Show threadJon AwbreyMay 20, 2023

@joeri_s

Sure, I understand that. The hedge “as it were” is used advisedly for the sake of the argument. (I wrote my own language learner back in the 80s.)

Speaking less metaphorically, the program and its database are always in their respective states and the program has the capacity to act on the database even when not engaged with external prompts.

Is there any reason why the program's “housekeeping” functions should not include one to measure its current state of “uncertainty” (entropy of a distribution) with regard to potential questions — or any reason why it should “hurt to ask”?

as it were …

Show threadJon AwbreyMay 22, 2023

@joeri_s

Hi Joeri,

Would it be okay with you if I quoted and linked to your comment on my blog?

Show threadJoeri SebrechtsMay 22, 2023
@Inquiry feel free, everything I write on mastodon is in the open