[DigitalToday reporter Yoonseo Lee] What results come out if you ask ChatGPT and Google Gemini to draw up custom PC build quotes? On April 20, IT outlet TechRadar reviewed the beginner quotes suggested by the two AI chatbots, alongside recommendations from in-house computing expert Matt Hanson.

The comparison focused on who could put together a more convincing PC for the intended use. Building a PC has a high barrier to entry for beginners because it requires checking part compatibility, cable management, complex product names and detailed specifications. Recent supply issues for memory and CPUs have also made choosing parts more difficult.

Each AI asked questions in a slightly different way. ChatGPT asked relatively detailed questions about intended use and preferences, and included technical terms that may be unfamiliar to beginners. Gemini asked fewer questions but used simpler wording. ChatGPT showed a relative strength in reflecting a beginner's requirements in detail.

The key was the actual recommendations. The comparison confirmed that AI can help with PC building, but cannot do everything. Limits stood out in particular when it came to real-time purchase links and price estimates. ChatGPT produced relatively better results than Gemini, but neither model could accurately find the best real-time prices on the internet. Gemini also provided inaccurate links in some cases.

Part compatibility itself was relatively stable. Cross-checking the recommendations from the two AIs with Hanson produced a positive assessment on whether the selected parts were compatible with each other. Even so, market price fluctuations were not sufficiently reflected, and when the models failed to fit the suggested budget, they repeatedly revised their answers without showing the level of clarity consumers might expect.

In the end, the purchase stage required checking model names again and searching for individual parts separately. Both services showed limits in carrying out the task consistently through to the end, and they also struggled to explain the gap between the budget and actual market prices smoothly.

The case shows how far generative AI can be effective. AI can be useful in reducing the burden of complex research and information searches in technology and science. It also showed that human judgment carries greater value in areas such as PC building, where hands-on hobbies and product choice matter.

Ultimately, generative AI proved meaningful as a supporting tool for helping beginners draft an initial quote, but the comparison confirmed that human review is essential in the final steps, such as checking purchase prices and links and adjusting the budget. It also highlighted that the faster product choice and market price changes interact, as in PC building, the more cross-checking is needed rather than following AI answers as-is.