How would ChatGPT fix the housing crisis?

Commenting on the housing crisis, the bot goes on: “The crisis has been exacerbated by the COVID-19 pandemic, which has led to job losses and economic uncertainty, making it harder for people to afford housing. The situation is particularly dire in London and other major cities, where housing is in high demand and prices are particularly high.

“The crisis is not only affecting low-income households, but also the middle-class families and young people. This is making it harder for them to access decent and stable housing.”

This might feel out of date considering the pandemic has arguably contributed to the opposite problem of labour shortages. It seems a reasonable summary, if you accept that longstanding structural issues which have not been addressed can reasonably be described as a ‘crisis’.

ChatGPT does at least understand that the problem persists across incomes.

“ChatGPT even goes where politicians fear to tread on tax reform”

But what would the robot do about this?. I asked the AI: “How can we fix the housing crisis in the UK?” and received this answer: “There are several ways that the UK government and private sector could work together to address the housing crisis, including:

1. Building more affordable homes – this could involve increasing funding for social housing and incentivising private developers to build more affordable homes.
2. Rent control – the government could introduce measures to control rents and make them more affordable for renters.
3. Tax reform – the government could reform taxes to make it more expensive for investors to buy properties and leave them empty, and more attractive for owner-occupiers.
4. Planning reform – the government could make changes to the planning system to make it easier for developers to build new homes.
5. Increase in wages – increase in wages would help people afford housing.
6. Increase in supply – increase in the number of houses available to buy or rent would help to lower prices.”

ChatGPT goes on: “Ultimately, addressing the housing crisis in the UK will require a multi-faceted approach that involves the government and private sector. It will also require co-operation between different levels of government, such as local and national government, to ensure that policies are implemented effectively.”

It might be a simplistic one, but it does seem a reasonable summary of the familiar fixes. Also, ChatGPT even goes where politicians fear to tread on tax reform.

The AI software seems to comprehend that we need many different solutions, but there is no real sense that it understands the problem. For example, increasing wages might make homes more ‘affordable’, but not for very long without greater supply and possibly not at all if accompanied by rising interest rates. 

I wanted to test the AI even further by asking why I can’t afford a home. Here’s the response: “There are several reasons why you may not be able to afford a home, including:

1. High house prices – if house prices in your area are high, it may be difficult to save enough money for a down payment and to make mortgage payments.
2. Low income – if you have a low income, it may be difficult to afford a home, even if prices are relatively low.
3. Lack of affordable housing – if there is a shortage of affordable housing in your area, it may be difficult to find a home that you can afford.
4. High debt – if you have high levels of debt, it may be difficult to qualify for a mortgage.
5. Lack of savings – if you have not been able to save enough money for a down payment, it may be difficult to buy a home.
6. Limited access to credit – if you have a low credit score or lack of credit history, it may be difficult to get a mortgage.
7. High unemployment or low wages – if you are unemployed or have a low paying job, it may be difficult to afford a home.
8. Lack of affordable rentals – if you are renting, and the rental market is expensive, it may be difficult to save enough money to buy a home.”

ChatGPT provided further context by adding: “It’s important to note that the factors which contribute to affordability of housing vary depending on different regions and demographic groups. It’s important to consider the specific reasons that make it difficult for you to afford a home.

“Additionally, there are various government programs that can help you to afford a home, such as Help to Buy and shared ownership, which you can look into.”

“The chatbot may be able to synthesise lots of answers from those that are already out there, but expecting anything more original might be asking too much”

The wording of your command is important here, as the AI assumes that this is all about personal finance. If Help to Buy and shared ownership are the answers, then maybe there is a problem with the question? But it’s a question I chose deliberately as an echo of the 2018 book by Josh Ryan-Collins, Why Can’t You Afford a Home?, which ChatGPT should add to its reading list.

There is a sense that the housing question might just be more complex than the bot makes out – and that the solutions might not be so simple. The politics of achieving those solutions is also completely missing, which comes back to my point about ‘crisis’.

Many people, and a majority of those most likely to vote, are profiting off high house prices and politicians are terrified of betraying them: see the government’s surrender to its backbenchers on planning reform for the latest example of this.

Perhaps that’s not ChatGPT’s fault, though. The chatbot may be able to synthesise lots of answers from those that are already out there, but expecting anything more original might be asking too much.

However, it does seem that AI can come up with a reasonable, if limited, summary of solutions familiar from countless reports on housing. Is it just ChatGPT that’s missing something then – or is it us, too?

Jules Birch, columnist, Inside Housing