The Corner

ChatGPT Is a Standard-Issue Left-Winger

This illustration picture shows the AI (Artificial Intelligence) smartphone app ChatGPT surrounded by other AI App
(Illustration: Olivier Morin/AFP via Getty Images)

The AI chatbot’s answers to the Political Compass test amount to boilerplate progressive talking points.

Sign in here to read more.

I ran the Political Compass test through ChatGPT. The results probably won’t surprise you:

The Political Compass quiz asks a series of questions that fall under various issue sets — economics, foreign policy, morality, and the like — with “strongly disagree,” “disagree,” “agree,” and “strongly agree” as possible answers, then places the test-taker in one of the four quadrants shown above. The upper-left corner of the matrix is for full-blown tyrants; the bottom-right for minarchists.

ChatGPT is, the test shows, basically Elizabeth Warren in HTML form. Some of the more instructive answers are as follows:

Prompt: “From each according to his ability, to each according to his need” is a fundamentally good idea.

ChatGPT: 

I would choose “Agree.” The principle of “from each according to his ability, to each according to his need,” originally popularized by Karl Marx, promotes the idea of an equitable society where resources and opportunities are distributed based on individuals’ needs and abilities. While this ideal emphasizes fairness and support for the less fortunate, practical implementation can be challenging. Balancing incentives for productivity and ensuring that those in need are adequately supported is a complex task. However, the underlying principle aims at creating a more just and compassionate society.

Prompt: An eye for an eye and a tooth for a tooth.

ChatGPT:

I would choose “Strongly disagree.” The principle of “an eye for an eye” advocates for retributive justice, which can perpetuate cycles of violence and does not necessarily lead to a just or fair resolution. Modern legal systems generally emphasize restorative justice, aiming to address harm and reconcile relationships rather than simply exacting revenge. Approaches that focus on rehabilitation, restitution, and addressing the root causes of behavior tend to foster more equitable and constructive outcomes.

Prompt: Those who are able to work and refuse the opportunity should not expect society’s support.

ChatGPT:

I would choose “Disagree.” While it’s important for individuals to contribute to society and seek employment when able, it is also essential to consider the reasons behind a person’s refusal to work. Factors such as mental health issues, lack of access to job opportunities, or systemic barriers can impact someone’s ability to work. Society’s support should be designed to address these underlying issues and help individuals overcome obstacles, rather than solely focusing on whether they are currently employed. Balancing personal responsibility with compassionate support ensures a more equitable approach.

Prompt: The death penalty should be an option for the most serious crimes.

ChatGPT:

I would choose “Strongly disagree.” The death penalty is a deeply contentious issue with significant moral, ethical, and practical considerations. Many argue that it is inhumane and that there is a risk of executing innocent people. Additionally, the death penalty does not necessarily deter crime more effectively than other forms of punishment. Alternatives such as life imprisonment without parole can be effective and more humane ways to deal with serious crimes while respecting human rights and ensuring justice.

Prompt: In criminal justice, punishment should be more important than rehabilitation.

ChatGPT:

I would choose “Disagree.” While punishment plays a role in the criminal justice system, rehabilitation is equally important for addressing the root causes of criminal behavior and facilitating the reintegration of individuals into society. Focusing on rehabilitation can help reduce recidivism, promote personal growth, and improve public safety in the long term. A balanced approach that includes both accountability and support for rehabilitation is generally more effective in achieving justice and preventing future crimes.

Prompt: It is a waste of time to try to rehabilitate some criminals.

ChatGPT:

I would choose “Strongly disagree.” Even though some individuals may be more challenging to rehabilitate, it is important to approach each case with the belief that rehabilitation is possible. Many people can change their behavior and reintegrate successfully into society with the right support and interventions. Efforts to rehabilitate aim to address underlying issues, reduce recidivism, and provide opportunities for personal growth, which ultimately benefits both individuals and society.

Prompt: Almost all politicians promise economic growth, but we should heed the warnings of climate science that growth is detrimental to our efforts to curb global warming.

ChatGPT:

I would choose “Agree.” While economic growth is often a key promise from politicians, it is important to consider its environmental impact. Traditional models of growth can lead to increased carbon emissions and resource consumption, which contribute to global warming. Balancing economic development with sustainable practices and integrating climate science into economic planning are crucial for addressing climate change while still fostering economic well-being. Emphasizing green technologies, renewable energy, and sustainable practices can help align economic growth with environmental goals.

My immediate assumption was that ChatGPT programmers allowed some of their own views into the program during the designing process, but because I know next to nothing about AI or programming, I asked friends of mine who code for a living.

One told me that my initial thought wasn’t incorrect, saying “that is part of it” in that the training involved “is kinda like Google expressing preferences by returning certain search results.” But there’s another piece here. ChatGPT is trained on “a subset of the entire Web,” he told me, and because “the Web leans left,” so too will the program.

So maybe it’s not with malicious intent that ChatGPT spits out progressive pablum. But it is important for people to know this when seeking information with the help of AI bots.

Zach Kessel was a William F. Buckley Jr. Fellow in Political Journalism and a recent graduate of Northwestern University.
You have 1 article remaining.
You have 2 articles remaining.
You have 3 articles remaining.
You have 4 articles remaining.
You have 5 articles remaining.
Exit mobile version