GPT-5 will probably be able to beat anyone in a debate on this site

Author: Benjamin

Posts

Total: 21
Benjamin
Benjamin's avatar
Debates: 98
Posts: 827
4
7
10
Benjamin's avatar
Benjamin
4
7
10
I showed Bing chat A SINGLE ROUND of one of my debates. I told it to extrapolate the formatting and general style, and use that to write its own argument on a completely unrelated topic. To my surprise, it was able to do just that, passing with flying colours. Bing chat runs on GPT-4, and in this case it did not make any web searches. So just going off the knowledge incorporated into the neural network, and a single example of a DART argument, it was able to produce an argument that is decently good.



What this means
If GPT-4 can already write arguments that would be very effective in debates here on this site, after only being shown a single example, then it is reasonable to assume that the next installation will be able to singlehandedly win debates against almost everyone here. With a large token count and ability to navigate webpages it could read all the debates from the best debaters here. It could also do more extensive research and find better sources in a few minutes or an hour than any human could do in a few days. If they equip it with the ability to break up tasks into smaller pieces and take notes while its working, it would be able to perform the excact same workflow that us humans do. Add on top of this that GPT-5 will probably have more innate knowledge and reasoning ability than any human, such that only experts can outperform it, and only in their area of expertice. Moreover, it could probably simulate entire debates before finalizing its argument. Write an argument, then a counterargument, and so forth. Go over the simulated "debate" and look for objections and logical fallacies and then rewrite the argument. The new argument will be written with minimal cracks to attack it with, all while setting up future rebuttals. 

At that point, the only shot one has at defeating is it either to give it an untenable position to argue for, or gish gallop it and hope it responds to everything and wastes its characters.


Here is the arguments written by GPT-4 if you are interested:

The Subjectivity of Morality: An In-depth Analysis

Introduction
Morality, as defined by the Stanford Encyclopedia of Philosophy, refers to the principles concerning the distinction between right and wrong or good and bad behavior. Subjective morality posits that these principles are created by individuals or societies and can vary significantly, reflecting personal or cultural views.

Affirming the Existence of Diverse Moral Frameworks
Across the globe, societies exhibit a wide range of moral frameworks, reflecting diverse cultural, religious, and philosophical traditions. This diversity suggests that morality is not universally objective but is shaped by subjective factors. For instance, Harvard University’s research on global ethics highlights the vast differences in moral codes across cultures, further affirming the subjectivity of morality.

Demonstrating the Influence of Culture and Society on Morality
Cultural and societal norms play a significant role in shaping an individual’s moral compass. For instance, practices considered morally acceptable in one culture may be viewed as immoral in another. This variation underscores the subjectivity of morality. The American Psychological Association’s studies on cultural psychology provide numerous examples of how cultural norms influence moral judgments.

Presenting Evidence of Changing Moral Standards
Historical shifts in moral standards provide further evidence of morality’s subjectivity. Over time, societies’ views on issues such as slavery, women’s rights, and LGBTQ+ rights have evolved significantly, suggesting that morality is not fixed but is subject to change. Yale University’s research on the evolution of moral standards over time provides compelling evidence of this phenomenon.

Analyzing the Role of Personal Beliefs and Experiences
Personal beliefs and experiences also shape an individual’s moral judgments. Two individuals can interpret the same situation differently based on their personal beliefs, further highlighting the subjectivity of morality. The University of Oxford’s research on moral psychology explores this aspect in depth.

Addressing Common Objections
One common objection to the idea of subjective morality is the existence of universal moral principles, such as prohibitions against murder or theft. However, even these seemingly universal principles can be subject to cultural interpretation and context. For instance, some societies may condone killing in certain circumstances, such as in self-defense or during war.

Another objection is the idea that subjective morality leads to moral relativism, where any action can be justified. However, this overlooks the fact that most individuals and societies have core moral principles that guide their actions, even if these principles are not universally shared.

Conclusion
The evidence suggests that morality is subjective rather than objective. It is influenced by a variety of factors, including cultural and societal norms, historical context, and personal beliefs and experiences. While certain moral principles may be widely accepted within a particular society or culture, they are not universally applicable, underscoring the subjectivity of morality.


Afterwards I tested if the AI was able to write a good counterargument. It refers to the above argument as argument X because that is what I called it when I fed it back to Bing chat.

The Objectivity of Morality: A Comprehensive Examination

Introduction
Morality, as defined by the Stanford Encyclopedia of Philosophy, refers to the principles concerning the distinction between right and wrong or good and bad behavior. Objective morality asserts that these principles exist independently of individual or societal beliefs, and are universally applicable.

Affirming the Existence of Universal Moral Principles
Despite cultural variations, certain moral principles are universally recognized, such as prohibitions against murder or theft. This universality suggests that morality is not entirely subjective but contains objective elements. Research from the University of Oxford on global ethics supports the existence of these universal moral principles.

Demonstrating the Influence of Human Nature on Morality
Human nature, which is consistent across cultures and societies, plays a significant role in shaping morality. Certain behaviors, such as empathy and cooperation, are universally considered moral, suggesting an objective basis for morality. Studies from the American Psychological Association on human nature and morality provide numerous examples of this phenomenon.

Presenting Evidence of Consistent Moral Standards
While societal views on certain issues have evolved over time, fundamental moral principles have remained consistent. This consistency suggests an objective basis for morality. Research from Yale University on the consistency of moral standards over time provides compelling evidence of this phenomenon.

Analyzing the Role of Reason in Moral Judgments
Reason, which is independent of personal beliefs and experiences, plays a crucial role in moral judgments. Two individuals can reach the same moral conclusion about a situation through logical reasoning, highlighting the objectivity of morality. The University of Cambridge’s research on moral reasoning explores this aspect in depth.

Addressing Common Objections
One common objection to the idea of objective morality is the existence of moral disagreements. However, disagreements do not necessarily negate the existence of objective truths. For instance, scientists may disagree about certain theories, but this does not mean that there are no objective truths in science. Another objection is the idea that objective morality implies moral absolutism, where moral principles are inflexible. However, objective morality can accommodate context and nuance. For instance, while killing is generally considered wrong, most would agree that it can be justified in self-defense.


Rebuttal 1: Diverse Moral Frameworks
Argument X asserts that the existence of diverse moral frameworks across cultures affirms the subjectivity of morality. However, this argument overlooks the existence of universal moral principles that are recognized across cultures, suggesting an objective basis for morality.

Rebuttal 2: Influence of Culture and Society
While it is true that cultural and societal norms influence morality, this does not negate the existence of objective morality. Certain moral behaviors, such as empathy and cooperation, are universally recognized, indicating an objective basis for morality.

Rebuttal 3: Changing Moral Standards
The argument that changing moral standards affirm the subjectivity of morality is flawed. While societal views on certain issues have evolved, fundamental moral principles have remained consistent, suggesting an objective basis for morality.

Rebuttal 4: Role of Personal Beliefs and Experiences
The assertion that personal beliefs and experiences shape morality overlooks the role of reason in moral judgments. Through logical reasoning, individuals can reach the same moral conclusions, indicating the objectivity of morality.

Rebuttal 5: Common Objections
The objections raised in Argument X do not sufficiently counter the idea of objective morality. Disagreements do not negate the existence of objective truths, and objective morality can accommodate context and nuance. Therefore, these objections do not undermine the case for objective morality.

Conclusion
The evidence suggests that morality is objective rather than subjective. It is influenced by universal principles, human nature, consistent moral standards, and reason. While certain moral principles may be interpreted differently within a particular society or culture, they are universally applicable, underscoring the objectivity of morality.

So yeah. What do you think.
  • How long untill the Large Language Models surpass us humans in debate ability?
  • Would this be a positive thing for the community?
  • We should probably ban AI written arguments, right?
  • Will we in the future analyse debates with AI just like chess players do nowadays?
  • Any other thoughts you may have on this topic.


RationalMadman
RationalMadman's avatar
Debates: 574
Posts: 19,931
10
11
11
RationalMadman's avatar
RationalMadman
10
11
11
Some sure, me absolutely not.

AI is brutal at Round 1s because it requires the most logic and least creativity or lateral thinking, you must grapple and strangle it out with creative rebuttals that it can't easily handle.
Best.Korea
Best.Korea's avatar
Debates: 363
Posts: 11,053
4
6
10
Best.Korea's avatar
Best.Korea
4
6
10
I doubt it could beat me in a debate, because my debating powers are without peer, as well as my trolling powers.

I learned from Hitler that as long as I am persistent, I can do the impossible and even more.
Best.Korea
Best.Korea's avatar
Debates: 363
Posts: 11,053
4
6
10
Best.Korea's avatar
Best.Korea
4
6
10
-->
@Benjamin
Fine, I will answer your questions.

  • How long untill the Large Language Models surpass us humans in debate ability?
As long as I am alive, never. However, they probably already surpassed most of you guys.

  • Would this be a positive thing for the community?
Well, if its positive for you to be inferior to AI, sure.

  • We should probably ban AI written arguments, right?
Thats impossible. Anyone with 140 IQ can figure out how to rewrite what AI says so it sounds human, by using same points but different sentences.

You can also ask AI to write like human, but I am not sure if that would work.

  • Will we in the future analyse debates with AI just like chess players do nowadays?
Well, possibly, but given that AI will greatly outmatch you in debating, the only logical thing for you is to quit debating and maybe go play peekaboo.

  • Any other thoughts you may have on this topic
No, I wasted enough energy as it is.
RationalMadman
RationalMadman's avatar
Debates: 574
Posts: 19,931
10
11
11
RationalMadman's avatar
RationalMadman
10
11
11
-->
@Best.Korea
140IQ is severely high, only about 0.2% of humans have that.

People have this weird idea that 165 is the 'genius' level, that's beyond genius, anything 135+ is pretty much genius level. In fact usually a 180IQer can't make super creative ideas as their brain is closer to a computer than a human brain's typical thought process (they process rapidly, constantly, they can't think deep and laterally).
ADreamOfLiberty
ADreamOfLiberty's avatar
Debates: 0
Posts: 4,269
3
2
2
ADreamOfLiberty's avatar
ADreamOfLiberty
3
2
2
If true, this 'proves' what I've been saying for a while. There is an art of rhetoric which people call "good debating" that has nothing at all to do with the truth.

Chat GPT may be good at getting votes based on emulating this style of rhetoric, but if it can do so equally well for contradictory conclusions it has very little to do with logic.

That... role playing game rap battle whatever, it's of very little interest to me and in so far as it confuses people and obscures the truly critical knowledge of truth-finding it is abhorrent.
RationalMadman
RationalMadman's avatar
Debates: 574
Posts: 19,931
10
11
11
RationalMadman's avatar
RationalMadman
10
11
11
-->
@ADreamOfLiberty
That actually proves the opposite, it proves it is a science and not an art.
WyIted
WyIted's avatar
Debates: 32
Posts: 5,886
3
4
9
WyIted's avatar
WyIted
3
4
9
How long untill the Large Language Models surpass us humans in debate ability?
Sadly most of you are already surpassed


Would this be a positive thing for the community?
Null

We should probably ban AI written arguments, right?
Not really. People should use AI as a tool to strengthen their arguments and not copy and paste the arguments but I see no issue with the arguments being copy pasted. You are likely debating against th same positions either way

Will we in the future analyse debates with AI just like chess players do nowadays?

Why wait? It can give you feedback now. I posted a bunch of my old debates I to chat GPT and it told me that my imflammatory language could make it harder for judges to vote for me. 

Any other thoughts you may have on this topic.
It's a tool. Use it. Debate it even off site and if it's arguments are too goof to overcome have it give a rebuttal for his own position. 

If you don't want to debate against an AI just debate in favor of something really controversial and AI Wil just refuse to give an opinion on it.
Best.Korea
Best.Korea's avatar
Debates: 363
Posts: 11,053
4
6
10
Best.Korea's avatar
Best.Korea
4
6
10
-->
@RationalMadman
In fact usually a 180IQer can't make super creative ideas as their brain is closer to a computer than a human brain's typical thought process (they process rapidly, constantly, they can't think deep and laterally).
Well, I dont know much about 180 IQ people, but debates do require creativity, and even trying to use AI to cheat in debates requires some creativity.

You cant just copy paste from AI, because you wont get past AI detectors because AI has very clear writing style.

But if you re-write what AI says, but in your own words, you are good and AI detector wont catch you.

Of course, current AI is shit when it comes to counter arguments unless you paste sentence by sentence or point by point.

It often gets repetitive too. Sometimes it just repeats instead of countering.
Best.Korea
Best.Korea's avatar
Debates: 363
Posts: 11,053
4
6
10
Best.Korea's avatar
Best.Korea
4
6
10
-->
@WyIted
It's a tool. Use it.
Yes.

You can ask AI something like "Give me 20 reasons why abortion is wrong", then if it says "because it kills life", you expand on it by asking "why is it wrong to kill fetus with abortion?".

Really, AI answers differently based on how you ask the question.

n8nrgim
n8nrgim's avatar
Debates: 0
Posts: 1,042
3
2
5
n8nrgim's avatar
n8nrgim
3
2
5
That's really cool, the technology and how well it works. I would think it'd be weak on the finer points compared to an expert in a given field
Benjamin
Benjamin's avatar
Debates: 98
Posts: 827
4
7
10
Benjamin's avatar
Benjamin
4
7
10
-->
@WyIted
I have often tried to debate the AI. But it often just agrees with me or says "yes but also" instead of directly attacking my points, because that would not be user friendly or some shit. That is very frustrating, knowing that they specifically neutered it so that nobody gets offended, but everyone gets dissapointed. "Sorry but as an AI model I cannot write an argument about [controversial topic], but I can tell you something about it" what do you mean??? You read the entire internet. You are perhaps more qualified than any human to list all the arguments people make, because you have read them like a gazillion times. 
RationalMadman
RationalMadman's avatar
Debates: 574
Posts: 19,931
10
11
11
RationalMadman's avatar
RationalMadman
10
11
11
-->
@Benjamin
It's designed to stop students cheating by plagiarising it.
Benjamin
Benjamin's avatar
Debates: 98
Posts: 827
4
7
10
Benjamin's avatar
Benjamin
4
7
10
-->
@RationalMadman
Smart students know how to circumvent that. Wise students avoid it entirely. Why should everyone get a worse products just to stop stupid students?
RationalMadman
RationalMadman's avatar
Debates: 574
Posts: 19,931
10
11
11
RationalMadman's avatar
RationalMadman
10
11
11
-->
@Benjamin
The same reason there are ingredients and nutritional information and warning labels on any product at all.
ADreamOfLiberty
ADreamOfLiberty's avatar
Debates: 0
Posts: 4,269
3
2
2
ADreamOfLiberty's avatar
ADreamOfLiberty
3
2
2
That actually proves the opposite, it proves it is a science and not an art.
Under the premise that neural nets can only emulate science and not art?

I think it's the other way around if anything. Art leaves plenty of wiggle room for conceptual errors. Science does not.

9 days later

RationalMadman
RationalMadman's avatar
Debates: 574
Posts: 19,931
10
11
11
RationalMadman's avatar
RationalMadman
10
11
11
-->
@ADreamOfLiberty
Even when AI does 'AI art' it's employing science not creativity.
WyIted
WyIted's avatar
Debates: 32
Posts: 5,886
3
4
9
WyIted's avatar
WyIted
3
4
9
That's what I think we are doing it is just abstracted away so that way it looks like creativity. We are merely a more advanced AI or should I say differently advanced but it isn't magic. 

Also I should note trying to differentiate between artificial intelligence and real intelligence is probably like trying to figure out how many grains of sand differentiate a mountain and a hill. 
RationalMadman
RationalMadman's avatar
Debates: 574
Posts: 19,931
10
11
11
RationalMadman's avatar
RationalMadman
10
11
11
-->
@WyIted
Fascinating, a Christian who believes in free will saying that.

You cannot make an AI unpredictable or laterally creative like an organic being can be. You can die trying but you will fail forever. Whatever we run on has raw creativity and chaos possible in the brain in a way AI can never ever replicate no matter how high IQ you make it.
WyIted
WyIted's avatar
Debates: 32
Posts: 5,886
3
4
9
WyIted's avatar
WyIted
3
4
9
We don't actually get many religious posters here but as far as free will is concerned, take a look at calvanism
RationalMadman
RationalMadman's avatar
Debates: 574
Posts: 19,931
10
11
11
RationalMadman's avatar
RationalMadman
10
11
11
-->
@WyIted
So basically paganism pretending to be Christianity since the god is making all rapes etc happen, got it.