Half a croissant, on a plate, with a sign in front of it saying '50c'
h a l f b a k e r y
Bone to the bad.

idea: add, search, annotate, link, view, overview, recent, by name, random

meta: news, help, about, links, report a problem

account: browse anonymously, or get an account and write.

user:
pass:
register,


                                   

AI / Stock / Debate

The Role of Pro/Con Algorithms in Shaping Artificial Intelligence
  (+4, -9)(+4, -9)
(+4, -9)
  [vote for,
against]

A website dedicated to promoting the best pro/con arguments could pave the way for a form of collective artificial intelligence. The emergence of AI is likely to mirror the evolutionary principles that led to human intelligence, relying heavily on algorithms that evaluate pros and cons—an essential component of decision-making.

Intelligence and Decision-Making Intelligence, often defined as "the capacity for logic, understanding, self-awareness, learning, reasoning, planning, creativity, and problem-solving," revolves around one core function: making better decisions. The foundation of intelligent decision-making, whether biological or artificial, rests on evaluating the validity (truth) and impact (costs, benefits, risks) of various pros and cons.

AI cannot effectively make decisions without a pro/con evaluation algorithm. Similarly, machine learning thrives on iterating and refining strategies by focusing on past successes—an implicit cost-benefit analysis to define "success."

Improving Argument Evaluation with Structured Scoring Pro/con evaluation fosters intelligence by promoting systematic, unbiased thinking. Removing emotion and bias enables us to make decisions rooted in evidence and logic. Scientific methods, critical reasoning, and even rational societal norms rely on processes that strip away anecdotal distractions to focus on measurable truths.

To implement this, arguments should be scored based on: 1. Truth: How verifiable is the argument? 2. Importance: Does the argument significantly impact the conclusion? 3. Relevance: Is the argument logically connected to the decision at hand?

This structure could facilitate better decision-making by filtering redundant, irrelevant, or poorly supported arguments.

Building Collective Intelligence Platforms An ideal online forum would use tools like pro/con analysis, crowd-sourced argument evaluation, and linkage scoring. For example: - Unique Scoring: Identify redundancy by grouping similar arguments. This ensures efficiency and prevents duplicative content from derailing discussions. - Linkage Scoring: Measure the strength of causal relationships between evidence and conclusions, dynamically recalibrating scores based on new information. - Expert Verification: Arguments endorsed by qualified individuals (e.g., professors in relevant fields) could receive higher weight.

Promoting Better Arguments To refine beliefs, arguments should be evaluated in survival-of-the-fittest matchups. A well-designed platform would: - Separate arguments about truth from arguments about importance and linkage. - Create spaces to identify logical fallacies or redundancies. - Encourage transparency by letting users see the supporting and opposing evidence for each belief in one place.

Democratizing Knowledge and Decision-Making Such a system could leverage algorithms to score and rank ideas, promoting reasoned discourse over emotional or partisan rhetoric. Arguments with higher supporting evidence and fewer weaknesses would naturally rise, while weaker arguments would fall. Over time, this process would encourage the public to focus on well-founded positions rather than divisive propaganda.

The Role of Funding, Experts, and Public Contribution A platform built on this model could sustain itself through donations targeted at specific ideas, further demonstrating public commitment. Additionally, experts could lend credibility by associating their insights with arguments, and books or resources critical to understanding an issue could be prominently tagged for users.

Towards a Smarter Society As ideas compete and evolve in this structured environment, society could map out all possible reasons for and against every significant decision—political, moral, or economic. By focusing on evidence and logical coherence, this system could bridge divides, mitigate misinformation, and refine collective understanding.

Conclusion To truly benefit humanity, AI must incorporate the deliberative strengths of human reasoning and the computational precision of algorithms. A "Battlefield of Ideas" platform—where arguments are scored and refined—could usher in a new era of collective intelligence, fundamentally reshaping decision-making processes for the better.

myclob, Mar 07 2005

TouchGraph Google Browser http://www.touchgra...GGoogleBrowser.html
An interesting way to browse Google... [zen_tom, Mar 11 2005]

Gnod http://www.gnod.net
Browse through music or books to see what you might like based on what you know you already like. [zen_tom, Mar 11 2005]

[link]






       The relevant comments I made on the other idea still apply. Some advice: you should leave out the detail you've already gone through in the other idea, and simply reference it in this one. It would (IMO) make this idea much more readable and independent.
Detly, Mar 07 2005
  

       (personal opinion) Write as if your reader had a 10 second attention span.
david_scothern, Mar 07 2005
  

       Sounds a bit like this site in some ways, though I must admit I went into 'skim' mode after reading the sentence: //The website with better web links.//. It's a bit confused but there's something in there.
wagster, Mar 07 2005
  

       Good points:
1. Two column format.
2. Your new algorithm that promotes annotations.
3. Site search and recognition management software.
  

       Interesting point of view:
Natural selection led to intelligence.
  

       Brazen presumption:
A broadly implemented, comparative ranking of multifarious and disparate ideas will lead to equal parts idea advancement and idea betterment.
  

       The Halfbakery site design lacks some of the refined annotation and opinion handling that are suggested, [myclob], but the beauty of this site is that ideas here can be silently voted into the stratosphere without 'idea necromancy' (pulling old ideas out of the cellar and putting them on the front page without adding substance to a very long idled annotation thread). That benefit of this site's idea management algorithm seems to me a bit of AI in and of itself.
reensure, Mar 07 2005
  

       I agree with everything here.   

       Eventually this kind of a system should get away from language and into emotions as evidenced by patterns in visual and auditory biofeedback displays.   

       People should choose by playing video games, not by using words.   

       Maybe Google is already an artificial intelligence. To speed it up we need a real time Dow Jones-like display of the popularity and contextual relationships of individual words.   

       I wonder if Google is holding an advanced interface like that back from the public. It would certainly be a better stethiscope of our collective trust than the stock market.
JesusHChrist, Mar 11 2005
  

       You are right. You could say that Google already has AI, it just needs to be presented to the public. For instance the internet would have voted for one of the presidents. I don't know which one, but it could be easy to determine which one. Do a Google search for George Bush is an Idiot, and all of the websites that say John Kerry is an idiot, and see which ones come back with the highest rank. You couldn't say the internet is always smart, but you could say that it has a preference for one man over the other. Also synonyms for stupid should be used. Of course you wouldn't have to use negative, but you could, for each statement, see which person would win. For instance, Google "Is a good person" for two candidates... Does this get anyone else excited?
myclob, Mar 11 2005
  

       //Does this get anyone else excited?// [myclob] sorry to put a downer on things, but no, not me. Google is a big stack of servers that indexes groups of unicode bytes into alphabetical order before linking them with urls that contain matching strings of unicode bytes. It's really not about to start voting for anyone.   

       It's a far cry from associating a lot of similar data to actually having any understanding of what any of it means. Yes, if you type in clever searches, you can get meaningful and surprising results - but that's you being clever, using a complex analysis tool. But it's certainly not an example of you interacting with an intelligent machine.
zen_tom, Mar 11 2005
  

       I've added a couple of links - Googleduel, which rates the googlerank of one entry over another (Kerry/Bush, Brown/Blair) etc and a link to the TouchGraph Google Browser page - this is a really great way to explore Google, and to see how various concepts are related to one another - it's best when you use a website as a starting point - I tried it with www.halfbakery.com and found the results quite interesting!   

       There's also a link to Gnod, a site that uses a kind of voting algorythm to link associated topics - you can use it to reccommend books or music based on authors or songwriters you know you already like.
zen_tom, Mar 11 2005
  

       People should choose by playing video games, not by using words?
yabba do yabba dabba, Mar 11 2005
  

       [myclob], I tested your suggestion.   

       [george bush is an idiot] 3,910 hits   

       [john kerry is an idiot] 103 hits   

       How are we going to factor in the Googlectoral college?
normzone, Mar 11 2005
  

       Can you also factor in the hits that say eg "Anyone who thinks that John Kerry is an idiot has obviously not studied him"?
angel, Mar 11 2005
  

       normzone: after your anno, the # hits should go up by one for each. See how valid this is?   

       This idea suffers from the premise that truth can be arrived at democratically. This lenghthily describes an algorithm to assess what's commonly believed vs. controversial, and we don't need no fancy algorithm to tell us that.
sophocles, Mar 11 2005
  

       It also suffers from the idea that a Google search would result in a 1:1 correlation with the opinion of the general public. The internet is still largely a haven for those who like to hang out there.
RayfordSteele, Mar 11 2005
  

       sophocles, I do not believe that "truth can be arrived at democratically." I do not think this is a truth promoting or finding website. It can only represent our colective (and current) belief or understanding of truth. That is why I call it collective intellegence, instead of AI or a truth finder.
myclob, Mar 11 2005
  

       It would have to be more sophisticated than just the number of websites that say that george bush is stupid. It would have to be more like the percentage of websites, because their are naturally more websites about George Bush (a 2nd term president who started 2 wars) than John (who?) Kerry.
myclob, Mar 17 2006
  
      
[annotate]
  


 

back: main index

business  computer  culture  fashion  food  halfbakery  home  other  product  public  science  sport  vehicle