The Power of GPT-3
Why should GPT-3 take over more human work? Because it will be more efficient than humans. It will be able to work faster and better than humans, and I think it will do so in the near future.
I am working with the GPT-3 team. (Not me, it’s what GPT-3 generated) We are moving fast. I think we can achieve this goal in three years, maybe five years. But for sure, in 10 years, machines will begin to take over the majority of human jobs.
So what is the best way to prepare for this? We need to have a discussion about how to adapt the market to new technology, to prepare for the new reality where machines can do most of our work for us. If we don’t do this, there is a risk that many people won’t be able to find jobs and will suffer. We have to make sure that everyone has enough income and enough money to buy what they need and want. For this, we need to redesign our economic system in the 21st century.
We interviewed a professor at Tsinghua University, in Beijing.
"What are your students’ reactions when you tell them about your research?"
"At first, they don’t believe me [laughs]. But when I give them more details about this technology, they begin to understand. They realize that GPT-3 is going to change their lives, and in fact, it will change the lives of all humans on earth. I think they have realized that this is an exciting challenge for them. They are very interested in developing AI technology, so much so that some of my students have even left Tsinghua University and joined my team at Baidu."
"How do you train your students? Do you focus on the ethical questions raised by AI?"
"I am a professor of machine learning, which means I teach students how to build intelligent machines using AI technology. But as a scientist who cares about the future of humanity, I have been thinking a lot about how AI technology can be used ethically in society. I am very concerned about how we use this technology – we must use it in a way that benefits humanity as much as possible. That’s why I wrote this book, The AI Revolution: The Road to Superintelligence (note: this is the title of an article by Tim Urban on HuffPost, not a book) – to help us understand this new technology and its implications for the future of humanity. The first step is educate people about AI and machine learning so they can understand what it is and how it works. That’s why I wrote my book – so anyone can understand these ideas easily. I also think we need an open discussion about these issues with different stakeholders – governments, businesses, NGOs and media – because AI will impact all these groups and our society as a whole, so we need everyone’s opinion on how we should use it in the future."
Governments should create policies for AI development based on public opinion; businesses should build up social responsibility programs for their employees; NGOs should educate their members about the benefits and risks of AI; media must help people understand these issues through different forms of media such as TV, newspapers or social media platforms like WeChat or Weibo. We need everyone’s voice in order to have a well-informed public discussion about how to use these technologies in the most beneficial way for humanity.
How can we make sure that AI technology is used in a way that benefits humanity? We have to control the development of AI. Right now, this technology is in the hands of very few people, and it is moving very fast. So we have to make sure that AI technology is not misused by bad people. I think AI can be used very well for humanity if we can control its development and use it to benefit everyone. For example, I think AI can be used to monitor the performance of government officials by automatically looking at their spending and behavior. If government officials are spending public money for personal reasons or are doing business with companies that are not efficient or honest, GPT-3 can be used to find them and report them to higher government officials. In this way, GPT-3 can help control the bad behavior of government officials and reduce corruption. If we use this technology to control the bad behavior of some people, we can also use it to help people who are underprivileged in society – such as children who do not have a good education or people who are sick – because AI technology can help them adapt better to society and help them find a job.
So in general, I think AI can be used for good purposes, but we need to control its development first. What is your advice for students who want to work in your field? I think it’s a great time for students to work on AI research because it’s a really exciting field right now. A lot of opportunities are available now because this technology is moving very fast and there are many interesting tasks that need to be done – both basic research and applied research. I think that AI is the future of our world, so I think it will be a good choice for students if they want to make a good career in the future.