DingTalk app enables enterprise customers to use Tongyi Qianwen, the ChatGPT-like service launched in April by Alibaba Cloud. DingTalk’s president said all products within the app will be ‘re-made with AI’ in the future, and that AI will underpin research and development. — SCMP
Alibaba Group Holding has been utilising self-developed large language models (LLM) to help Chinese enterprises improve efficiency via its workplace collaboration app DingTalk, as the ecommerce giant pushes ahead with its investment into this hot artificial intelligence (AI) space.
The DingTalk app enables enterprise customers to use Tongyi Qianwen, the ChatGPT-like service launched in April by Alibaba Cloud, said Ye Jun, president of DingTalk, at the annual digital economy conference hosted on Tuesday by Luohan Academy, the open-research platform initiated by Alibaba in 2018.
In a keynote speech given by Ye in Alibaba’s home city of Hangzhou, Ye presented a user case of how DingTalk has used generative AI tools to help Belle International, a women’s shoe retailer in China, to increase work efficiency in many areas. He said the AI had improved information acquisition, content creation, as well as customer services.
In a live demonstration the generative AI tool was asked how to promote a new product, and it was instructed to give various responses including mind maps, proposal outlines, as well as images. The app was also able to field customer service questions, using the company’s internal database.
Founded in Hangzhou in 2018, Luohan Academy is an open research institute initiated by Alibaba. It has gathered social scientists, including seven Nobel Prize laureates in economics, to research critical issues where digital technology impacts society.
“Five years ago, people’s impression of AI was that it could be used for facial recognition, but now with the emergence of ChatGPT and other intelligent capabilities, I believe there is a new beginning,” said Ye.
He added that all products within DingTalk will be “remade with AI” in the future, and that AI will underpin research and development.
LLMs drive AI-powered chatbots like ChatGPT from Microsoft-backed OpenAI. LLMs allow the chatbots to answer sophisticated queries with humanlike responses, and generate writing, code, and other content.
Ye’s statement came a few weeks after Alibaba said it had integrated its LLMs into Tingwu, its voice recording and transcribing tool that was launched in October 2021, with wider roll-out plans including DingTalk integration.
Alibaba’s cloud unit in April unveiled its own alternative to ChatGPT – Tongyi Qianwen – which is based on DAMO Academy’s LLMs, marking an early foray into the ChatGPT space, along with search engine giant Baidu, which launched Ernie Bot in March.
The service has received more than 200,000 beta testing applications from corporate clients, Alibaba chairman and CEO Daniel Zhang Yong said in a conference call with analysts last month.
Alibaba has started to work with partners to develop industry-specific AI models, said Zhang, who on Tuesday announced he would relinquish his roles at the parent holding company in September to focus on Alibaba Cloud.
In a dialogue with three Nobel Prize laureates of economics, Zhang talked on Tuesday in Hangzhou about the opportunities and challenges of AI. He said computing power, data and industrial applications are three layers for the development of generative AI, although IP protection is an area of concern.
“We’re in a fast-changing world where the foundation model (has been) proven,” said Zhang at the event. “(It) will generate new industries or new services which we’ve never seen (before), and I think this will make the whole society move forward.” – South China Morning Post
