The independent student newspaper of Iowa State and Ames since 1890

Iowa State Daily

Iowa State Daily

Iowa State Daily

Suza: AI can be a powerful tool in the war against poverty

How artificial intelligence may possess wisdom in light of drastic circumstances
Courtesy of IStock

The United States is a land of plenty. Even poverty.

The U.S. Census Bureau estimated in 2021 that there were 16 million poor non-Hispanic whites, 11 million poor Hispanics or Latinos, 9 million poor blacks or African Americans and 1 million poor American Indians and Alaska Natives. 

To estimate the rate of poverty among these groups we take into consideration their total population. In 2021, the total population was 196 million non-Hispanic whites, 63 million Hispanics or Latinos, 45 million blacks or African Americans and 4 million American Indians and Alaska Natives. 

Dividing the number of people in poverty by their total population gives the poverty rate. For non-Hispanic whites it was 8.1%, Hispanic or Latino 17%, black or African Americans 20% and American Indian and Alaska Native 25%. 

America’s poor are white, brown and black, yet poverty occurs at higher levels in communities of color. What might be the reason? Let’s talk about how America estimates poverty first.

The U.S. government uses income to estimate individuals and families in poverty. “If a family’s total income is less than the family’s threshold, then that family and every individual in it is considered in poverty.” The poverty level income in 2023 for an individual is $14,880.

Such low income means a person may not be capable of owning a vehicle. Since it’s common in the United States for people to live far from supermarkets, lack of transportation can cause food insecurity. Yet a person without a car may not experience similar transportation constraints in a rural setting as a larger city. In a rural setting tight-knit relationships may enable a person to seek help from family or friends, but living in an urban setting may mean lack of support in times of need. 

Amartya Sen, a Harvard Economics and Philosophy Professor and Nobel Laureate, warns us not to see poverty just as reduced income, but also the “deprivation of basic capabilities.” From the context of philosophy, capabilities are like the freedom to travel, eat sufficient and nutritious food, obtain an education and live in a community of their choice. The lack of economic freedom makes a person less capable of enjoying real freedoms. 

Let’s now bring in my AI partner, ChatGPT, to talk about poverty and its causes in the context of the United States. ChatGPT’s summarized responses are in italics.

What is poverty?

Poverty is a state of extreme financial deprivation and lack of basic necessities, such as food, shelter and clothing, due to insufficient income or resources to meet one’s essential needs. It often leads to reduced quality of life and limited opportunities for individuals and communities.

Who experiences poverty?

Poverty affects various demographic groups, including low-income individuals, families and children, as well as minorities and individuals with disabilities. 

Who among the various demographic groups experiences more poverty?

Statistically, minority populations, particularly African Americans and Hispanics, tend to experience higher poverty rates in the United States compared to the majority white population. 

What contributes to poverty in communities of color?

Historical discrimination, lower educational opportunities, limited access to well-paying jobs, systemic inequalities and housing disparities. These factors create barriers to economic advancement, perpetuating higher poverty rates within these communities.

Why does systemic inequality exist?

Historical and structural biases in institutions and policies that disadvantage certain racial, ethnic, or socioeconomic groups. Discriminatory practices, unequal access to education, employment opportunities and resources, as well as biases in the criminal justice system.

What can humans do to eradicate poverty?

  • Implement fair labor practices, minimum wage laws and progressive taxation.
  • Improve access to quality education, vocational training and adult literacy programs.
  • Strengthen social welfare programs and provide a safety net for vulnerable populations.
  • Encourage job growth and support small businesses.
  • Promote gender equality to empower women economically.
  • Ensure affordable health care to reduce medical-related poverty.
  • Invest in infrastructure and technology for economic development.
  • Involve communities in decision-making and development.
  • Raise awareness and advocate for poverty eradication.

What can AI do to help eradicate poverty?

  • Streamline the administration of social welfare programs, ensuring efficient and equitable distribution of benefits to those in need.
  • Identify affordable housing opportunities and optimize urban planning to address housing shortages.
  • AI-powered personalized learning tools can improve educational outcomes, especially in underfunded schools and underserved communities.
  • AI-powered telemedicine and diagnostic tools can enhance health care access, particularly in remote or low-income areas.
  • AI-driven financial tech solutions can expand access to banking and credit for marginalized communities, promoting economic empowerment.
  • Optimize food distribution networks, reducing food deserts and ensuring access to affordable, nutritious food.
  • Analyze data to predict poverty risks, enabling proactive interventions and policy adjustments.
  • Assist in identifying opportunities for economic development and revitalization in impoverished neighborhoods.

What would a poverty-free America look like?

A poverty-free America would represent a society where every individual’s basic needs are met, where opportunity is truly equal and where social and economic disparities have been significantly reduced or eliminated. It would require a concerted effort from governments, businesses, communities and individuals to address systemic issues and ensure a more equitable and prosperous future for all.

I agree, ChatGPT, but sadly, some humans just don’t care. 


View Comments (5)
Donate to Iowa State Daily
Our Goal

Your donation will support the student journalists of the Iowa State Daily. Your contribution will allow us to purchase equipment, send our student journalists to conferences and off-set their cost of living so they can continue to do best-in-the-nation work at the Iowa State Daily.

More to Discover
About the Contributor
Walter Suza, Author
Donate to Iowa State Daily
Our Goal

Comments (5)

All Iowa State Daily Picks Reader Picks Sort: Newest

Your email address will not be published. Required fields are marked *

  • D

    David Jackson | Sep 20, 2023 at 7:02 pm

    Can ChatGPT tell us what evidence exists to support the assertion of racial, ethnic, and socioeconomic discrimination and biases are the cause of poverty, and not the statistics directly linking single parent homes to poorer performance academically and socially across all racial and ethnic groups? Or the fact first generation immigrants belonging to racial and ethnic backgrounds that experience poverty at higher rates in the US, outperform and experience less poverty than people of those same racial and ethnic backgrounds who are from the US and have been influenced for decades by detrimental cultural behaviors and bad government incentives?

    Getting a chat bot to regurgitate the answers it was programed to think isn’t a “powerful tool in the war against poverty” it’s either an implicit assumption AI is an all-knowing electronic god to be believed without question, or a political-ideological rouse to manipulate people who don’t know any better into believing what you want. If I’m building a bridge and enter the max dynamic load capacity of my support structures into my computer, but mistype the numbers, the computer didn’t make any mistake in its analysis, yet gave the wrong calculation based on an error in the data it was given. Sorry, but even if current AI can recognize how certain data tend to pair based on feedback from human trainers and data pulled from the internet and puts them together generating something new from content it already recognizes, when that data is partial at best to incorrect at worst, its answers aren’t going to be unbiased truth. Instead, they will simply be a function of the information it was given and trained to produce.

    If ISU students learn anything about politics in their years at Iowa State it needs to be this: asking what evidence one has, and is it comprehensive, is the single most important thing to question about what a person or organization is proposing. Particularly if it’s being sold with appeals to emotion and moralizing slogans, as those are always meant to steer the audience away from logical analysis of the tradeoffs and factual outcomes in favor of self-satisfaction for being socially viewed as supporting the “right” thing agaisnt the “wrong” thing.

    This is usually done to either push an ideological agenda, or promote consolidating political power, or both but in either case under the guise of painlessly solving a social issue, and if you’re too busy feeling good about the alleged solution’s advertised intentions to question what evidence exists that the tradeoffs will result in a net benefit over the costs in the real world, you’ll be foolish enough to vote for it. This only benefits the politicians who can now promote themselves as saviors when all they really accomplished was creating new problems, and advertise themselves as the leaders with the great solutions again when someone notices the new problems that were created. That and they usually have more power than they had before at the expense of the very people they told you they were trying to help. Whether or not a chat bot has been trained to produce slogans and dogmas based on popular cheery picked data and partisan narrative in support of such agendas is no excuse for not seeking out the comprehensive facts.

    • J

      Josh | Sep 22, 2023 at 1:27 pm

      To answer your first question, yes it can. I would recommend trying it out. I copied and pasted directly your question and it gave what I would think is a relatively comprehensive answer. And while it doesn’t always give specific studies or policies, it does give a good jumping off point for you to delve further

      • D

        David Jackson | Sep 24, 2023 at 4:30 pm

        Then why didn’t you post it…it’s almost like the facts are the rise of single motherhood is detrimental to success and is provable for all ethnic groups, and that immigrants outperform their same race peers in the US because they’re not effected by counter productive cultural ideals, and that these facts disprove historical discrimination and systemic oppression as causal to disparate outcomes.

        For extra credit, ask the thinking machine so you don’t have to anymore the answer to why black Americans went from near zero literacy at emancipation due to slavery to high levels of academic success decade after decade up until the 1960s when reading comprehension and math scores, along with high school and college graduation rates, began to trend back down after decades of meteoric rise?

        • J

          Josh | Sep 26, 2023 at 1:44 pm

          I didn’t post it because the answer is a bit more verbose and formatted in a way that likely wouldn’t paste well in a comment on this site. That’s why I recommend you go try it for yourself 🙂

          I mean, I guess I don’t disagree that single parenthood makes success more difficult but I think saying that that is the sole cause is overly simplistic for such a complex thing to measure. It also doesn’t really do anything to disprove systemic biases and discrimination like you claim.

          Lastly, punctuation is a powerful tool, friend.

          • D

            David Jackson | Sep 27, 2023 at 10:07 pm

            If you truly understand something, you can summarize it.

            I didn’t argue it was the sole cause. The point is that single motherhood has been proven to be a major causal factor of increased criminality and poorer academic performance across all racial and ethnic groups, and that recent immigrants outperform their same race peers in the US. That in fact does disprove systemic biases and discrimination as the root cause, as if such were true, those immigrants would not be able to succeed either as they would suffer from the same systemic biases and discrimination keeping them down.

            Lastly, thanks for the tip on powerful tools. So is independent thought, friend.