Brother
Professional
- Messages
- 2,590
- Reaction score
- 539
- Points
- 113
Is the age of machine intelligence coming?
At the recent Intel Innovation conference, Intel CEO Pat Gelsinger announced the imminent release of "AI PCs" -personal computers optimized to accelerate a wide range of artificial intelligence tasks without an Internet connection.
Intel is not alone in its aspirations. Major IT companies, such as Apple and Qualcomm, are actively developing hardware and software solutions for local execution of neural networks on user devices instead of remote servers in the cloud. The goal is a personalized AI that is so naturally integrated into our lives that we will no longer notice "artificiality" in it.
However, there were obstacles in the way of realizing this dream. Today, the most powerful AI models require expensive specialized hardware and a high-speed Internet connection. Even when accessing remote computing power, the response of systems such as ChatGPT may slow down due to server congestion.
Pallavi Mahajan, vice president of Intel, emphasized that today about 50% of AI tasks are performed on users ' devices, mainly thanks to natural language processing and computer vision. However, the most powerful AI models currently depend on data centers, which limits their speed and availability.
Oliver Lemon, professor of computer science at the University of Edinburgh, spoke about his experience working with large language models, noting that they are not fast enough for real-world tasks. Instead, his team used the smaller Vicuna-13B model from Meta.
Companies are now actively developing tools for local AI execution, such as Intel's OpenVINO and Apple's Core ML. This not only speeds up the AI, but also improves data privacy, since all processing takes place on the user's device.
An example of this approach is the Rewind app for Mac and PC, which allows you to perform complex AI tasks, including searching for deleted emails and files, locally on the user's device.
New Qualcomm Snapdragon chips will enable powerful AI models like Meta's Llama 2 to be used directly on smartphones without the need for an internet connection.
These innovations not only improve productivity and privacy, but also open up new opportunities for smart assistants and home devices, enabling intelligent capabilities even with limited resources.
It seems that the industry is moving towards a new paradigm - when powerful AI "on the edge of the network" will become a familiar part of our devices.
At the recent Intel Innovation conference, Intel CEO Pat Gelsinger announced the imminent release of "AI PCs" -personal computers optimized to accelerate a wide range of artificial intelligence tasks without an Internet connection.
Intel is not alone in its aspirations. Major IT companies, such as Apple and Qualcomm, are actively developing hardware and software solutions for local execution of neural networks on user devices instead of remote servers in the cloud. The goal is a personalized AI that is so naturally integrated into our lives that we will no longer notice "artificiality" in it.
However, there were obstacles in the way of realizing this dream. Today, the most powerful AI models require expensive specialized hardware and a high-speed Internet connection. Even when accessing remote computing power, the response of systems such as ChatGPT may slow down due to server congestion.
Pallavi Mahajan, vice president of Intel, emphasized that today about 50% of AI tasks are performed on users ' devices, mainly thanks to natural language processing and computer vision. However, the most powerful AI models currently depend on data centers, which limits their speed and availability.
Oliver Lemon, professor of computer science at the University of Edinburgh, spoke about his experience working with large language models, noting that they are not fast enough for real-world tasks. Instead, his team used the smaller Vicuna-13B model from Meta.
Companies are now actively developing tools for local AI execution, such as Intel's OpenVINO and Apple's Core ML. This not only speeds up the AI, but also improves data privacy, since all processing takes place on the user's device.
An example of this approach is the Rewind app for Mac and PC, which allows you to perform complex AI tasks, including searching for deleted emails and files, locally on the user's device.
New Qualcomm Snapdragon chips will enable powerful AI models like Meta's Llama 2 to be used directly on smartphones without the need for an internet connection.
These innovations not only improve productivity and privacy, but also open up new opportunities for smart assistants and home devices, enabling intelligent capabilities even with limited resources.
It seems that the industry is moving towards a new paradigm - when powerful AI "on the edge of the network" will become a familiar part of our devices.
