Bitten by the AI bug

A platform powered by artificial intelligence and launched by a 13-year-old has gone viral among middle-school students since its debut in Shenzhen, Guangdong province, in September 2023.

AI-Button, the platform founded by middle-school student Qiu Yumo, who entered the eighth grade this year, enables students to customise learning guidance by simply uploading material.

With ChatGPT, a chatbot developed by United States company OpenAI, taking the world of technology by storm, Qiu used the large language model – a key ChatGPT technology – for the first time to develop her learning platform.

The AI revolution, which many technology industry experts believe is underpinned by ChatGPT, is driving a new form of entrepreneurship, described as “smaller, faster, cheaper and weirder” by Nathan Baschez, co-founder of US daily newsletter startup Every Inc.

Along with Qiu, young people from China, especially those born in the 1990s and 2000s, are increasingly plunging into AI entrepreneurship.

Qiu, who was born in 2010, started programming when she was 6. She wrote a computer game one year later, which triggered her interest in developing complicated software.

With the idea of making a small learning tool for herself, she used the large language model and other emerging AI technologies to create AI-Button earlier this year.

“It was only after ChatGPT began to take off that I sensed the opportunities brought by the new AI generation, and I realised that such technologies can change the world,” Qiu said.

ChatGPT is driving the AI revolution, which a Goldman Sachs research report said has the potential to bring sweeping changes to the global economy, and could produce a rise of 7 per cent, or almost $7 trillion (£5.7 trillion), in global GDP, over the next decade.

Baschez, from Every Inc, said in a note that AI startups in this new age begin small. “They use open source and cloud computing to get off the ground quickly and perform recurring tasks. Thanks to AI, they will stay smaller for much longer, and the most successful will achieve staggering results with only a handful of employees,” he said.

His comments are supported by the example of Midjourney, a system that uses AI to generate pictures from text descriptions. The Midjouney team at his beginning comprised only 11 full-time employees, with four of them students who had yet to graduate and were not studying at leading universities.

Baschez said, “More important, this new type of organisation can try out fresh ideas. The entry barrier is already lower than ever, but now that AI is good enough to design interfaces, write code, and execute marketing campaigns, we’re going to see a whole new set of experiments come online.”

In China, Miaoya Camera, an AI-powered digital product that generates self-portraits online, launched as a mini-program on WeChat in July, just five months after the launch team of six Chinese software professionals, who were born in the late 1980s and early 1990s, conceptualised and developed the product.

To obtain an AI portrait by using Miaoya Camera, users pay 9.90 yuan (£1.11) and upload at least 20 recent portraits to get a digital clone. They then turn the clone into different types of portraits, such as ID photos, business photos, or those in an ancient Chinese style.

Within a month of its launch, Miaoya Camera was viewed by market observers as a potential disruptor of the AIGC, or artificial intelligence-generated content, niche market arising from the arrival of ChatGPT.

Zhang Yueguang, leader of a group of students who graduated from Tsinghua University in 2012, said the portrait generator’s market performance is way beyond his expectations. He attributes part of this success to the new wave of AIGC, with a large number of consumers looking forward to applications that use ChatGPT-related technologies.

“I think the business logic of AIGC and the internet is fundamentally different. Business on the internet is essentially about information flow, channels and platforms. But the essence of the AI era can be compared with building a factory. Technologies that point to market demand are more easily accepted by customers,” he said.