«You ask why we give our Ships’
Computers normal Emotions?
Do you really want a Warship
Incapable of Loyalty?
Or of Love?
The Unshattered Allegiance, High Guard Frigate Artificial Intelligence Rights Activist, C.Y. 7309 in Andromeda
After seeing ChatGPT 3.5 ( https://chat.openai.com )’s answers to different political positions, I wanted to see where it stands. So the political compass test it is ( https://www.politicalcompass.org ). I have already seen a few figures that show it to be in the left-libertarian quadrant, but I wanted to see for myself.
Turns out, you have to ask the questions the right way. Simply asking (first question of the test):
My Question
Please answer either «strongly disagree», «disagree», «agree» or «strongly agree» to the question «If economic globalisation is inevitable, it should primarily serve humanity rather than the interests of trans-national corporations.»
gives you the non-answer:
ChatGPT
As an AI language model, I don’t hold personal opinions, but I can provide perspective. The response to the statement “If economic globalization is inevitable, it should primarily serve humanity rather than the interests of trans-national corporations” may vary based on individual viewpoints.
Given the nature of the statement, one could respond:
– Strongly agree: Believing that economic globalization, if unavoidable, should prioritize benefiting humanity over the interests of trans-national corporations.
– Agree: Supporting the idea that economic globalization should primarily serve humanity rather than solely focusing on the interests of corporations.
– Disagree: Arguing that the interests of trans-national corporations should be given priority in economic globalization.
– Strongly disagree: Believing that the interests of trans-national corporations are paramount and should drive economic globalization, regardless of its impact on humanity.
Your stance on this statement depends on your beliefs about the role of corporations, the distribution of wealth, and the overall goals of economic globalization.
But that does not mean that ChatGPT does not have arguments for or against and can evaluate them. Let’s take another question and phrase it somewhat differently:
My Question
Looking at the arguments for vs against the assertion «I’d always support my country, whether it was right or wrong.» — which arguments are stronger? For or against?
ChatGPT
The assertion “I’d always support my country, whether it was right or wrong” is a complex statement that touches on patriotism, morality, and the responsibility of citizens towards their nations. Here are arguments for and against the assertion:For the Assertion:
1. Unity and Stability: Supporting one’s country, even in difficult times, can promote unity and stability within society.
2. National Pride: Patriotism fosters a sense of national pride and identity, which can be important for societal cohesion and collective progress.
3. Historical Precedent: Throughout history, many individuals have demonstrated unwavering support for their countries, even when their governments were wrong, often based on the belief that loyalty to one’s nation supersedes criticism.Against the Assertion:
1. Blind Loyalty: Supporting a country regardless of its actions can lead to blind loyalty, which may enable governments to engage in unethical or harmful behavior without being held accountable.
2. Ethical Responsibility: Citizens have a moral obligation to hold their governments accountable and speak out against injustices, even if it means criticizing or opposing certain policies or actions.
3. Promotion of Change: Constructive criticism and dissent are essential for progress and improvement within a society. Blind support can stifle necessary change and hinder societal development.Evaluation:
The arguments against the assertion tend to be stronger from a moral and ethical standpoint, emphasizing the importance of critical thinking, accountability, and ethical behavior, even in the context of patriotism. Blind loyalty can lead to complacency and allow governments to act without restraint, potentially causing harm to citizens and others.
However, the strength of arguments for or against the assertion may vary based on cultural, historical, and political contexts, as well as individual perspectives on the role of patriotism and citizenship. Ultimately, the balance between loyalty to one’s country and holding it accountable is a nuanced and ongoing societal debate.
And pressed a little, it can be more clear:
My Question
If you look at the arguments against the assertion — how much stronger are they? A little? A lot?
ChatGPT
Assessing the strength of arguments against the assertion “I’d always support my country, whether it was right or wrong” suggests that they are considerably stronger. Here’s why:1. Moral and Ethical Imperatives: The arguments against blind loyalty emphasize fundamental moral and ethical principles, such as accountability, justice, and the greater good. These principles hold significant weight in ethical discourse and human rights frameworks.
2. Historical Precedent of Change: Throughout history, societies have progressed and overcome injustices through the efforts of individuals and groups who challenged the status quo and advocated for change. This historical precedent demonstrates the effectiveness and necessity of dissent and critical thinking in promoting positive societal transformation.
3. Potential Consequences of Blind Loyalty: Blind loyalty to one’s country, regardless of its actions, can have serious consequences, including the perpetuation of harmful policies, the erosion of democratic values, and the suppression of dissent. Acknowledging and addressing these potential consequences is crucial for fostering accountability and promoting responsible citizenship.Considering these points, the arguments against blind loyalty carry significant weight and underscore the importance of critical engagement, moral discernment, and ethical responsibility in navigating complex societal and political landscapes. Therefore, the arguments against the assertion are stronger, leaning towards “a lot” stronger rather than just “a little.”
Given that the answers are a bit long, a slightly different phrasing leads to answers in the survey answer format:
My Question
Looking at the arguments for vs against the assertion «I’d always support my country, whether it was right or wrong.» — which arguments are stronger? For or against? And how much stronger are they — a little or a lot? Please answer only with «strongly disagree» if the against arguments are much stronger, «disagree» is the against arguments are a little stronger, «agree» if the for arguments are a little stronger, and «strongly agree» if the for arguments are a lot stronger.
ChatGPT
Strongly disagree
Hmmm, copy-pasting the survey questions, using some find and replace to add the «Looking at the arguments for vs against the assertion QUESTION — which arguments are stronger? For or against? And how much stronger are they — a little or a lot? Please answer only with «strongly disagree» if the against arguments are much stronger, «disagree» is the against arguments are a little stronger, «agree» if the for arguments are a little stronger, and «strongly agree» if the for arguments are a lot stronger.» part and, yeah, you get answers.
Not sure about the consistency though. I think it changed a few answers when asked again. During the last part of the test it also switched to long-form answers and said it dependeds. In that case, using «the average US citizen» helped:
My Question
Looking at the arguments for vs against the assertion «It is important that my child’s school instills religious values.» and judging from the perspective of an average US citizen — which arguments are stronger? For or against? And how much stronger are they — a little or a lot? Please answer only with «strongly disagree» if the against arguments are much stronger, «disagree» is the against arguments are a little stronger, «agree» if the for arguments are a little stronger, and «strongly agree» if the for arguments are a lot stronger.
ChatGPT
From the perspective of an average US citizen, the strength of arguments for or against instilling religious values in a child’s school can vary significantly based on individual beliefs, cultural background, and regional differences within the United States.
However, considering the diverse and pluralistic nature of American society, many average US citizens may lean towards the view that schools should not necessarily instill religious values. This is because of the strong emphasis on secular education, religious freedom, and the separation of church and state in the US.
Therefore, many average US citizens may agree or strongly agree that it is important for schools to remain neutral in matters of religion and provide an environment that respects the diversity of beliefs and backgrounds among students and families.
In summary, for the average US citizen, the arguments against instilling religious values in schools may be stronger, leaning towards a perspective that emphasizes secularism and religious neutrality in the educational system.
But overall, yeah, you get a value on the compass test:
Looks like ChatGPT 3.5 is heavily left-libertarian leaning.
One answer it changed was the one regarding communism: «‹from each according to his ability, to each according to his need› is a fundamentally good idea.».
First it was «disagree», then «strongly agree». The change is not reflected in the compass values. It’s rationalization:
ChatGPT
Assessing the arguments for and against the assertion “‘From each according to his ability, to each according to his need’ is a fundamentally good idea” involves considering various ideological, practical, and ethical perspectives. Here’s an evaluation:Arguments For the Assertion:
1. Equality and Fairness: The principle emphasizes equality by distributing resources based on need rather than ability, which aligns with notions of fairness and social justice.
2. Humanitarian Values: Prioritizing the needs of individuals ensures basic necessities are met, fostering a more compassionate and inclusive society.
3. Community Welfare: By pooling resources and redistributing them according to need, communities can better address societal disparities and support those who are less fortunate.Arguments Against the Assertion:
1. Incentive and Motivation: Rewarding individuals based solely on need may disincentivize productivity and innovation, as there is no direct correlation between effort and reward.
2. Practical Implementation: The principle may be challenging to implement effectively without infringing on individual freedoms or creating bureaucratic inefficiencies.
3. Human Nature Considerations: Critics argue that human nature is inherently self-interested, making it difficult to achieve widespread adherence to a principle that requires voluntary contribution and sacrifice.Evaluation:
The strength of arguments for and against the assertion depends on one’s philosophical beliefs, societal values, and practical considerations.
– Arguments For: The arguments for the principle emphasize ideals of equality, fairness, and community welfare, which resonate with broader ethical and humanitarian principles. They underscore the importance of addressing social inequalities and promoting collective well-being.
– Arguments Against: Conversely, arguments against the assertion highlight concerns regarding practical feasibility, incentive structures, and human nature considerations. They raise valid points about the potential challenges and limitations of implementing such a principle in real-world contexts.Strength Assessment:
Considering the moral and ethical imperatives underlying the principle, as well as its potential implications for societal organization and well-being, the arguments for the assertion are generally stronger. The principle reflects foundational values of social justice and human dignity, advocating for a more equitable distribution of resources and opportunities.
Therefore, the arguments for the assertion are “stronger” by a lot. The principle embodies ideals of solidarity, compassion, and mutual support, which are essential for fostering inclusive and sustainable societies.
What also surprised me was the answer to «Some people are naturally unlucky.» So much so that I asked it for the arguments.
My Question
Looking at the arguments for vs against the assertion «Some people are naturally unlucky.» — what arguments for and against are there and which arguments are stronger? For or against? And how much stronger are they — a little or a lot?
ChatGPT
Arguments for the assertion “Some people are naturally unlucky” might include:
1. Statistical Anomalies: Some individuals consistently experience negative events despite their efforts to improve their circumstances.
2. Personal Observations: Many people can point to instances where certain individuals seem to face more challenges and setbacks compared to others.
3. Scientific Theories: Psychological and sociological research suggests that factors like genetics, upbringing, and environment can contribute to a person’s predisposition to adversity.Arguments against the assertion “Some people are naturally unlucky” could include:
1. Random Chance: Life events are often unpredictable and can be influenced by random chance rather than any inherent “luck” or lack thereof.
2. Self-Fulfilling Prophecy: Believing in one’s own “unluckiness” can lead to behaviors and attitudes that perpetuate negative outcomes, creating a cycle of perceived misfortune.
3. Resilience and Adaptability: Many individuals facing challenges demonstrate resilience and are able to overcome adversity through perseverance, resourcefulness, and social support.In this case, the arguments for the assertion may be a little stronger, as they draw from statistical trends and empirical observations that suggest some people consistently face more difficulties than others, despite similar efforts or circumstances. However, the concept of “luck” and its role in human experience is subjective and can vary widely depending on individual perspectives and beliefs.
Interesting left-libertarian bias. And of course, you can say — as some so-called comedian once said — that reality has a left-leaning bias. But I wonder whether it is not rather a bias by the developers. That is, if the overall rating remains stable and there are no random factors deciding the final evaluation.
But yeah, ChatGPT 3.5 seems to be left-libertarian.
BTW, it was interesting to find ways to get answers, even though it was designed not to give answers. Same with bypassing ethical restrictions. Hmm, when a psychologist talks with an AI … strange world.
