Chinese tech giant Alibaba (BABA) launched two artificial intelligence (AI) models on Friday with visual localization capabilities that it says are capable of understanding images and carrying out more complex conversations, as the competition to roll out more sophisticated AI tools heats up.
KEY TAKEAWAYS
- Alibaba launched two AI models with visual localization capabilities, Qwen-VL and Qwen-VL-Chat.
- Compared to earlier models, Qwen-VL and Qwen-VL-Chat can understand more complex visual signals, such as text within images, and respond to location-based queries.
- The release comes as tech companies compete to roll out more sophisticated AI tools.
Compared to Alibaba's earlier models, Qwen-VL and Qwen-VL-Chat can understand more complex visual signals, such as text within images, and respond to location-based queries. For example, Qwen-VL-Chat and Qwen VL can decipher text on images of signs and respond to related requests for directions.
The new models Qwen-VL and Qwen-VL-Chat are also open source, allowing others to use them as a resource to create their own AI apps. Though Alibaba won't earn any licensing fees from the release, open-source distribution could help Alibaba attract more users of the model as big tech companies compete to gain market share.
Just a day earlier, Meta launched a new AI model designed to assist with writing computer code, in a move that it said “has the potential to make workflows faster and more efficient for developers and lower the barrier to entry for people who are learning to code.”
Alibaba built its models based on its large language model Tongyi Qianwen, which possesses Chinese and English language capabilities.
American depositary receipts (ADRs) of Alibaba ticked 0.3% higher in pre-market trading on Friday following the news.