The company’s Meta Connect 2024 developer conference took place on Wednesday.The social media giant unveiled several new artificial intelligence (AI) and wearable features during the event. In addition, Meta has also reportedly announced a partnership with tech giant Arm to build custom small language models (SLMs). These AI models are said to be used to power smartphones and other devices and introduce new ways to use these devices. The idea is to provide parameters on the device and edge computing to enable fast AI inference.

According to a report from CNET, Meta and Arm plan to create AI models that can perform more complex tasks on devices. For example, the AI ​​can act as a virtual assistant for the device and can make calls or click images. This is not too far-fetched, as today AI tools can already perform many tasks such as editing images and writing emails.

However, the main difference is that users must interact with the interface or type certain commands to get the AI ​​to perform these tasks. At the Meta event, the duo reportedly emphasized that they want to do away with that and make AI models more intuitive and responsive.

One way to do this would be to move the AI ​​models to the device or keep the servers very close to the devices. The latter is also known as peripheral computing and is used by research institutions and large enterprises. Raghavan Srinivasan, vice president of generative AI product management at Meta, told the publication that developing new AI models is a good way to take advantage of this opportunity.

For this, AI models must be smaller in size. While Meta has developed large language models (LLMs) of up to 90 billion parameters, they are not suitable for smaller devices or faster processing. The Llama 3.2 1B and 3B models are said to be ideal for this.

Another challenge, however, is that AI models must also be equipped with new capabilities beyond simple text generation and computer vision. This is where Arm comes to the rescue. According to the report, Meta is working closely with the tech giant to develop CPU-optimized AI models that can adapt to the workflows of devices such as smartphones, tablets and even laptops. No other details about SLM are available at this time.