Do you really need an AI computer?
The true potential of AI PCs lies in their future use, not in replacing our main computers right now
AI, the buzzword of recent times, seems to be creeping into literally everything these days. After powering chatbots, virtual assistants, web tools, and our favourite applications, it has started to spill into absurd roles like burger flippers, fortune tellers, beauty contest judges, and even toothbrushes and vacuum mops.
It was only a matter of time before artificial intelligence made its way into personal computing, and that time is now. But do we really need it?
The failure of devices like the Humane AI Pin or Rabbit r1 and the overkill of Flippy, the burger flipper, have taught us that not everything needs artificial intelligence. Just because AI is selling right now doesn't mean we need to reinvent the wheel with artificial intelligence.
Many features promised on AI PCs, apart from the native AI assistant, will still require an internet connection. Currently, these computers mainly use artificial intelligence to accelerate on-device or cloud-based tasks.
Keep updated, follow The Business Standard's Google news channel
However, AI PCs are here, and now we must wonder: do we really need to upgrade to an AI PC from our traditional computers, or is it just a gimmick likely to die out in a few years? To answer that, we need to understand what AI PCs are, why they were made, how they work, what problems they solve, and whether end users are ready for built-in AI processing in their main devices.
To most, the term AI PC might not seem new, as we have already been using AI in our computers since 2022 with the release of ChatGPT. However, there is a big distinction between using AI-powered tools on a computer and having an AI computer.
Currently, most of the processing that enables AI tools to work happens in the cloud. For example, asking ChatGPT or Google Gemini a question, or having DALL-E or Midjourney create an image from a prompt, does not require any computational power on the user's end. Everything happens on OpenAI, Google, or the respective tool's servers, which are miles away from the user. This creates two problems: latency and vulnerability.
Since the processing happens on a remote server, commands and data take time to be received and processed, resulting in noticeable delays. Additionally, data shared with these tools needs to be sent and accessed remotely, threatening confidentiality.
Both latency and vulnerability can be addressed by doing the computation locally on the user's devices. However, traditional computers are not capable enough to run AI models without affecting performance, battery life, or device longevity. Enter the AI PC.
An AI PC is like any other traditional computer but with one added component: a Neural Processing Unit (NPU) dedicated to processing AI-related programs and tasks. This frees up the CPU and GPU from the additional workload, ensuring a smooth experience. Any computer with an NPU or a dedicated artificial intelligence processor can be classified as an AI PC.
As artificial intelligence advancements become exponential, traditional PCs will soon be unable to keep up with AI capabilities, hence the emergence of PCs with dedicated AI-processing capabilities.
Companies like Intel, AMD, and Microsoft have fully committed to integrating artificial intelligence into their products, and PC brands have adopted this integration into their computers.
Intel, a leading manufacturer of semiconductors and computer circuits, calls this AI integration the biggest advancement in personal computing in two decades. Their new Intel Core Ultra processors include built-in AI acceleration to ensure efficient distribution of work among the CPU, GPU, and NPU.
This significantly improves performance, especially for tasks requiring neural processing. Since it has an AI processing unit built in, the computer running on this chip would not need an internet connection to run large language models (LLMs) or basic AI-enabled functions. Everything happens locally on the PC, making the process faster and more secure.
AMD has joined this race with its AMD Ryzen AI, the world's first dedicated AI engine for an x86 computer. With this AI engine, computers can act as personal virtual assistants, generating ideas and producing visuals from prompts. In addition to substituting tools like ChatGPT and Gemini, this AI engine can enhance video calls, offer threat protection, and self-heal from cyber-attacks.
Microsoft also launched its Copilot+PC series with Surface and Surface Pro laptops to enter the AI PC market. These laptops are powered by Qualcomm Snapdragon X chips, built specifically for AI processing. Major PC brands, including ASUS, Lenovo, HP, Dell, and Acer, have also adopted Copilot+PC and introduced their own AI PC lineups.
These PCs come with many AI-powered features. Similar to Gemini, ChatGPT, Rytr, QuillBot, and other generative AI tools, your computer will be able to write essays, social media captions, or emails from prompt instructions. Visual content generation needs will also be addressed by built-in AI. The native AI will enhance video conferencing by adjusting backgrounds and, mimicking Apple's Centre Stage, always keeping you centred in the frame. Using machine learning, these PCs will scan your device for malicious attacks, prevent them, and recover from their effects.
Along with text-to-image, predictive text fill-out, visual editing, and essentially everything that cloud-based AI tools offer, these computers provide specific niche features that might make life easier.
Recall is one such feature that allows you to view past actions on the device and play them back like a video. The integration also enables real-time live captions in more than 40 languages. Advanced search allows users to search for anything they might have written, clicked, browsed, saved, or generated.
However, many features promised on AI PCs, apart from the native AI assistant, will still require an internet connection. Currently, these computers mainly use artificial intelligence to accelerate on-device or cloud-based tasks.
The features offered by AI PCs are niche and handy but are not revolutionary, and do not prompt an immediate upgrade. The problem is not with feasibility or usability; these PCs offer perks, and people have the know-how to use them. With our experience using prompt-based generative AI tools, the transition from traditional PCs to AI PCs would not be as challenging as it seems. These computers can be used like traditional PCs. If you're not accustomed to prompt-based actions, you can buy a new Microsoft Surface Pro laptop, never use a prompt, and still accomplish everything.
The problem is that almost everything these PCs offer can be achieved with third-party cloud or web-based AI tools we already have. It doesn't make much sense to swap your existing setup with an AI PC for the limited offerings of these seemingly future-proof machines.
The true potential of AI PCs lies in their future use, not in replacing our main computers right now. However, similar to the integration of voice assistants with smartphones, there will come a day when a computer without an NPU or dedicated AI engine will be obsolete, and it certainly feels like we are heading that way.
