Tencent doubles down on AI, to launch model later this year

Photo credit: Tada Images / Shutterstock

AI is Tencent‘s next frontier. In its Q2 2023 earnings call, the Chinese tech giant’s co-founder and CEO Pony Ma said that it is testing a “proprietary foundation model in different use cases.”

Tencent president Martin Lau further elaborated that the company sees an opportunity in generative AI beyond building intelligent chatbots like OpenAI’s ChatGPT.

“That is a product form factor which can be added to our large daily active user products like Weixin and QQ over time,” said Lau. “The more we look at it, the more excited we are for that asset growth multiplier across our many businesses.”

China’s AI market is slated to reach US$26 billion in value by 2026 – more than double from 2022 – which explains why Tencent and its fellow industry giants are investing in the space.

Tencent Meeting, its cloud-based video conferencing solution, has been using a model that can provide a summary of meeting notes. Similarly, Alibaba has integrated AI capabilities into DingTalk, its workplace messaging app, to summarize meetings and help draft emails, among others.

But that’s just the beginning. Earlier this month, Alibaba opened up access to its large language models to the public. These include Qwen-7B-Chat, which is specialized for conversational AI.

Meanwhile, Baidu is close to launching its Ernie 3.5 generative AI model, which is set to power a ChatGPT-like offering dubbed Ernie Bot. The model will also boost the capabilities of Baidu’s search engine.

A likely differentiator for Tencent is its specialized AI services for businesses, which it integrates into Tencent Cloud and calls a “model-as-a-service” offering. The company has launched a marketplace-like platform where its enterprise customers can pick from different types of AI models that best suit their needs and data sets.

According to the firm, some of the industries its model-as-a-service offering caters to include tourism and public services.

Based on their testing, Lau believes that Tencent’s AI model is among the “top leading foundation models” from its home country. Its advertising business has already benefited from investments into machine learning, with the company now seeing more accurate ad targeting and higher returns on investment for its clients.

Lau also said that the firm is testing its AI tech internally in areas like gaming and fintech.

“So, I think we are actually embracing generative AI on multiple fronts, and they are all making very good progress,” he added. “And over the mid-to-long run, we believe this is actually a very positive driver for our business.”

Of course, progress will also depend on factors larger than these companies. With the US limiting overseas access to chips made by Nvidia, a shortage of components critical to AI is becoming a hurdle for Chinese firms.