assisted epistemology
2009-Aug-15, Saturday 07:37 pm![[personal profile]](https://www.dreamwidth.org/img/silk/identity/user.png)
I coined a phrase back on the afternoon of 2006 October 04 when I tried to imagine a computer technology that would assist humans in collaboration on large, complex projects. Suppose a democracy allowed its citizenry to craft legislation directly. That's the kind of massive scale that I had in mind, the involvement of millions of people to produce a single document that met preset standards of quality. We would need a tool that allowed huge numbers of people with varying skill levels and specialties to work together while identifying contributors either as reputable or disreputable based on the validity of their concrete arguments and sources.
On that date, I searched the quoted phrase "assisted epistemology", and I came up with a completely empty result list from Google. How often can you do that using only a 2-word search phrase? *proud grin* I emailed ANI-L (an autism listserv) that day to tell them about my idea. Even today, google finds only 3 hits on the phrase, and none of them are appropriate to the concept that I intend.
http://www.google.com/#hl=en&q="assisted+epistemology"
Me, I'm still just "getting by" in life, as it were, so I haven't even begun to build such tools. I'm glad to learn today that other people are putting serious effort into building them. Some people are studying the logic of structured debate. Others are creating tools to assist with confrontational computing, like Intel's browser plugin tool called "Dispute Finder" for annotating and searching text with disputed claims. The most recent entry seems to be an essay on "disputation arenas" written by my favorite sci-fi author of all time, David Brin.
Just imagine the power of a tool that can help people to contribute on the difficult construction of an important document. It could be a document of legal significance, or religious, or medical. Contributions would be rated according to the logic of their argument and also according to the truth of their argument. Logic should be rather uniform (I think?), but truth would depend upon the knowledgebase of the subject matter. People would gain reputations (tracked by the software, of course) for their consistency in making logical contributions... or for frequently adding to the knowledgebase (like Wikipedia perhaps but with mandatory source article verification or at least plausible verifiability by experiment).
Confrontational computing. Disputation arenas. Assisted epistemology.
It'll happen someday. We'll have tools that help us, even with our slow chemical-sidetracked minds, to find the truth of our statements. I'm very hopeful. :)
On that date, I searched the quoted phrase "assisted epistemology", and I came up with a completely empty result list from Google. How often can you do that using only a 2-word search phrase? *proud grin* I emailed ANI-L (an autism listserv) that day to tell them about my idea. Even today, google finds only 3 hits on the phrase, and none of them are appropriate to the concept that I intend.
http://www.google.com/#hl=en&q="assisted+epistemology"
Me, I'm still just "getting by" in life, as it were, so I haven't even begun to build such tools. I'm glad to learn today that other people are putting serious effort into building them. Some people are studying the logic of structured debate. Others are creating tools to assist with confrontational computing, like Intel's browser plugin tool called "Dispute Finder" for annotating and searching text with disputed claims. The most recent entry seems to be an essay on "disputation arenas" written by my favorite sci-fi author of all time, David Brin.
Just imagine the power of a tool that can help people to contribute on the difficult construction of an important document. It could be a document of legal significance, or religious, or medical. Contributions would be rated according to the logic of their argument and also according to the truth of their argument. Logic should be rather uniform (I think?), but truth would depend upon the knowledgebase of the subject matter. People would gain reputations (tracked by the software, of course) for their consistency in making logical contributions... or for frequently adding to the knowledgebase (like Wikipedia perhaps but with mandatory source article verification or at least plausible verifiability by experiment).
Confrontational computing. Disputation arenas. Assisted epistemology.
It'll happen someday. We'll have tools that help us, even with our slow chemical-sidetracked minds, to find the truth of our statements. I'm very hopeful. :)
:o)
Date: 2009-Aug-16, Sunday 01:10 am (UTC)Re: :o)
Date: 2009-Aug-16, Sunday 01:18 am (UTC)"The one method by which human beings can reliably be made aware of their delusions is through interaction with others -- (a crucial point that we need to make clear to burgeoning artificial intelligences! )
YOU are capable of noticing the delusions that I am too in-love-with to spot or correct. In pointing them out, you do me the service of reciprocal accountability (RA) -- or criticism -- a great boon, allowing self-improvement, and a boon which I'll be only too happy to serve back to you, in plenty. As a favor, of course."
So maybe it can be an AI that helps us out, as long as we remind it that it needs to stay within an argumentative community (in the philosophical sense) to ensure that it retains its grip on reality. (I point now at modern American neoconservatives as a perfect example of what happens when you first lose the ability to debate and then lose the ability to judge reality itself.)
Re: :o)
Date: 2009-Aug-16, Sunday 02:15 am (UTC)no subject
Date: 2009-Aug-16, Sunday 02:14 am (UTC)I would postulate that this exactly is what a true democracy is. Although I would be wary of the scope of power the citizenry should have. Some things should not be allowed to be voted on, like basic human rights, such issues should be reach through logic.
Of course logic is only as good as the underlying value hierarchy assumed.
So I would postulate that value hierarchy needs to be the first issue to be resolved. Using logic of course. What values are useful? :o)