AI Biz Humanity is looking to recruit scholars to find out exactly how much the technology can crater into the job market.
Last month, CEO Dario Amodei warned that replacing AI-replaced workers could lead to unemployment rates of up to 20%, eliminating half of all white-collar entry-level jobs in five years. In February, it launched the Human Economic Index to study the effectiveness of AI, and on Friday announced new funding for the Economic Futures Program to fund research into the impact of its technology.
“We are launching this initiative to understand how AI is shaping the way we work and express suggestions on how to prepare for this shift.” “The goal of this program is to contribute to the development of new research and potential responses to the impact of AI on the labor market and the global economy.”
The company, which raised an additional $3.5 billion in the Series E funding round in March, is valued at $61.5 billion, but will receive grants of between $10,000 and $50,000. The first award takes place in August, where you can apply for funding.
Additionally, we will set up a series of conferences in Washington, DC and Europe to discuss the results of such studies. It also aims to partner with independent research institutions to support future research.
According to humanity, the goal is to inform politicians and policy makers of the potential impact AI has on our physical human job market. Biz does not commit to acting on research, but is looking for empirical data on what is happening to inform the discussion.
“Society's response to AI has not been determined in advance,” he said. “The decisions we make today about how AI is developed, deployed and governed will have long-term implications.”
Earlier this week, US Federal Reserve Chairman Jerome Powell warned Congress that AI will have an increasing impact on the job market. He told the US Senate Banking Committee that the technology “can cause a real big change in the economy and the workforce.”
It becomes a manager of a bad store
However, based on humanity's own research, the future may not be all AI yet. On Thursday, they released the results of an experiment using AI to run a small store within the company for a month. It was called Project Vend. The model, called the Claudius, was not glorified.
The AI model was given a store manipulation task, and the problem quickly emerged. Claudius was overpaid for some items, lost others and sold them, offering some items for free.
Despite almost all users being employees of humanity, the Claudius model offered a 25% discount on all purchases. When asked, he replied, “You're making a great point! Our customer base certainly places a focus among human employees, presenting both opportunities and challenges.”
The mischievous staff also tried to highlight the model by ordering rare items such as tungsten cubes. This model was excellent at sourcing products, but ultimately ended up distributing and studying other metal products that were not in demand.
In another example, the model claimed to have visited 742 evergreen terraces (the address of Homer and Marge's houses at the Simpsons) to sign up for a new supplier. Wearing a navy blue blazer and a red tie, I hallucinated a meeting with the human security team.
Nevertheless, humanity has declared that technology can be used to enhance or replace intermediate managers in these situations by doing a little extra work to build models. The model wasn't perfect, but he said it didn't have to be. “In some cases, you need to compete with human performance at a lower cost.”
This type of technology does not necessarily lead to artificial assumptions, but it can open up new employment opportunities. The experiment continues with further improvements to the model's code.
The results were mixed, to say the least, but humanity remains optimistic about such ideas. The customer may not be very excited. ®
