We can raise billions for AI but we can’t solve child poverty

Photo by Piron Guillaume on Unsplash

If you could give away 10% of your VC fundraising round to solve child poverty, would you do it?

Today I opened up one of the best AI newsletters I follow, called Ben’s Bites.

Two of the headlines scream $60M and $80B in fundraising for AI projects. It popped a thought into my mind, as we traverse this time of recession where we are all struggling.

This just feels like too much money concentrated on Technology. I felt poor, stupid, and even wronged.

If we take into account that one of the criteria for diagnosing Psychopathy is to identify that the individual can look at pain and hurt, show no emotion/feel nothing, and even feel relief or happiness, then we are potentially looking at a psychopath.

By looking, reading, and not questioning what we are seeing, we are probably approaching at least a sociopathic category of some sort.

Here are the fundraising candidates:

Reddit has an upcoming AI IPO so it has raised $60M, to let an AI company use their content for a training model. I agree with the writer, the number feels small, for the level content that Reddit has stored, throughout the years

OpenAI, the biggest headliner of recent times, confirms that after the latest round of fundraising, it is now worth $80B. This latest deal which brings the company to triple it’s value, to power the development of a text-to-video tool, according to the New York Times.

As a Technologist, I should feel excited. Yet, my brain is not in peace with this type of news.

We have incredible problems to solve. From war, to poverty, hunger, cold houses (a good number of us cannot afford heating) and yet, I am reading the excitable news that these companies are drowning in money.

What is it we cannot see?

Photo by Hermes Rivera on Unsplash

What if 10% of this money could solve child poverty in the US (where the funding has taken place)?

Vox.com has a brilliant article from September 2023 around this, stating ‘We cut child poverty to historic lows, then let it rebound faster than ever before’.

In 2021, we had reduced the poverty levels to 3.8 million American children, considered to be living below the federal poverty line.

As of 2022, the Census Bureau reported an increase of 139%, brining the number of struggling children to 5 million, with 9 million parents struggling to make ends meet.

Nations must be judged by the way they treat their children

What is the true cost of poverty in the U.S.? According to the Brookings Institution, which has tried to put together 20 packages for government approval, with the intent to contradict and reduce child poverty.

2 Packages submitted to the government stood out with the following:

‘At $90 billion to $110 billion a year, the costs of these two packages would be substantial, although much less than the aggregate costs that child poverty imposes on the nation, which have been estimated to amount to $800 billion to $1.1 trillion each year.’

The previous reduction of poverty had been achieved with the simple package called the Child Tax Credit, existing in other countries (like the U.K.).

The most affected children are of African and Hispanic ethnicity.

Could a Universal Credit system reduce child poverty? Absolutely.

Even if AI is automating work and progressing well, we are not seeing such credit anywhere, what we see instead are back raising all sorts of taxes that curiously impact the poorest households already battered and bruised by economic strains.

Some say the economy is booming, but that has to be questioned as only a handful of companies are truly making astronomic profits and likely the ones funding AI developments.

I have no doubt AI will bring big things and solve other problems but, not at the expense of letting ourselves look at a high figure of poverty for our populations and say nothing.

About the author

Carla is a mother, writer, HR technology expert with a crazy ambition to make humans more human, in a world of rising sociopathy

--

--