OpenAI's large-scale UBI experiment results are out, will universal basic income make people "lie flat" directly?

share
OpenAI

I have been closely following the CEO of OpenAI, Sam Altman, on the topic of Universal Basic Income (UBI). The large-scale UBI experiment he leads has been gradually revealing its results. Will people who receive basic income every month choose to just "lie flat"?

OpenAI Conducts Large-Scale UBI Experiment

Sam Altman initiated his own UBI experiment through OpenResearch in 2020, paying over 3,000 participants between $50 and $1,000 monthly.

  • Approximately one-third of the participants received less than $1,000 per month, referred to as recipients
  • Two-thirds were in the control group, receiving less than $50 per month, referred to as control participants

At the start of the program, participants had an average annual household income of $29,990, aged between 21 and 40, from rural, suburban, and urban areas in Illinois and Texas.

Can Receiving $1,000 Monthly Lead to "Lying Flat"?

Due to the high unemployment rates during the COVID-19 pandemic when the study began, only 58% of recipients and 59% of control participants were employed. By the end of the program, the employment rate increased to 72% for recipients and 74% for control participants.

The trend lines show quarterly average employment rates for recipients and control participants over time. The colored bands around the lines represent 90% confidence intervals for the estimates: larger shaded bands indicate more variability, while smaller shaded bands indicate higher precision.

The red shaded band around recipients in the graph below is significantly larger than the purple shaded band around control participants, indicating greater variation in employment rates among recipients. However, recipients also did not seem to give up on job opportunities.

Cash Provides More Choices

The research found that UBI can empower individuals to make employment decisions that align with their personal circumstances, goals, and values. Recipients were more likely to be actively job-seeking but with more selectivity.

The study results showed that recipients experienced an average of about 1 month longer in consecutive unemployment compared to the control group but had fewer suitable job options. This suggests that recipients were able to be more selective in their job search. One recipient, Jessie, stated:

I didn't have to take a crappy job just because I needed income right now. I had a chance to stick it out and try to find something more suitable.

Some recipients used the cash to make up for lost income due to resigning or reducing work hours, choosing to become full-time caregivers or increase time spent with family. Bethany quit her job to homeschool her three children, two of whom were recently diagnosed with autism and ADHD, struggling in traditional education.

For individuals facing mental and physical health challenges, the extra cash support may enable them to seek care or prioritize their well-being. Some recipients were able to use the time off work to improve their mental and physical health and eventually return to the workforce during career transitions.

In the Era of AI, Universal Basic Income (UBI) Evolves into Universal Basic Compute

As artificial intelligence (AI) continues to reshape the labor market, Sam Altman, CEO of OpenAI, proposed a new concept related to Universal Basic Income (UBI) known as "universal basic compute."

Sam Altman's concept of "universal basic compute" involves allowing individuals to access a portion of computing power from advanced AI systems like GPT-7. "Everyone can get a small piece of the computing power of GPT-7," explained Sam Altman. He envisions people being able to directly use this resource, sell it, or donate it for purposes such as cancer research, potentially creating a more valuable digital asset than traditional currency. "You own a piece of the productivity," he noted, suggesting that this model has the potential to empower individuals economically.

OpenAI's Sam Altman proposes "universal basic compute," transforming UBI with AI.