'Just use the manual': I followed ChatGPT's PC-building instructions, and all I got was a huge headache and a schooling on prompting

Two tech journalists looking confused at a motherboard
(Image credit: Future)

I am a PC builder. There, I said it. I don't feel like one yet — but the facts are undeniable. I built a PC.

Okay, so I had some help along the way, but now that I've crafted my own gaming rig for the latest season of TechRadar's AI Planned It video series, I can hold my head high among the experienced builders on the team.

Article continues below

The more risky (and potentially costly) stage of the experiment, of course, was actually putting the thing together — and that process taught me a lot about both PCs and optimizing LLM usage.

So, here's how I did it, including the specific instructions ChatGPT gave me for my PC build, plus my thoughts on what worked and what did. Hopefully they'll be helpful to others — and please do share your thoughts in the comments below.

I-I-I hallucinate when you call my name

I've already shared the prompt I used to train up Gemini and ChatGPT for this project, so be sure to give that a read for some extra context.

As far as the building stage went, my principal concern with using ChatGPT was that it might hallucinate different parts or outdated configurations. My experiences in the first stage of the experiment, in which it shared outdated information on a couple of occasions, did nothing to reassure me.

So, at the very start, I doubled down and reminded it of my configuration via the below prompt:

The first signs of trouble

Initially, ChatGPT took the word "detailed" and really rolled with it, presenting me with reams of instructions. In doing so, it made it pretty hard to follow as a first-timer, especially due to its excessive use of (often unnecessary) jargon.

While I had the benefit of building the PC alongside TechRadar's resident computing expert, Matt Hanson, I wanted as much as possible to emulate what someone trying this at home might experience. So, I reminded ChatGPT I was a beginner and it came back with the below instructions:

IT'S ALIIIIIIIIIVE

For the full, unfiltered trauma I experienced when following these instructions, you'll have to watch the full episode on TechRadar's YouTube channel. The long and short of it, though, is that barring one or two missteps where ChatGPT encouraged us to do a step too early or wasn't clear enough on its instructions, the steps were broadly correct.

That being said, it hit me a few times with a generic "just use the manual" tip. In an ideal world, it would search for those instructions and be a true assistive companion, but ultimately, for a fair chunk of this process, I ended up looking at the respective components' printed/online guidance.

I don't mind that too much, and should have probably guessed that'd be the case — particularly given that I'd asked ChatGPT to return such a long-form result. Setting aside the lofty ease-of-use promises made by AI evangelists in the years since LLM technology began its rapid ascent to mass adoption, ChatGPT itself advised that I should instead "message and I’ll guide you one stage at a time." Granted, that tip was at the end of its building instructions and I was by then elbow-deep in cables, but I do respect that it made the suggestion.

In the end, I did actually take it up on this offer, though not for the main building process; that would have been a nightmare while also trying to film, present and take notes on the day, which was hard enough by itself.

Instead, when I later had major issues troubleshooting a problematic bootable Windows drive, ChatGPT was there to help in real-time, both by message and through photos of the interface to help me navigate jargon-y menus.

Emphatic message sent to ChatGPT after seeing some success in PC building

Typically, I'm a huge hater when it comes to messaging AI like its your pal. The joys of PC building, however, broke my brain. (Image credit: Future / OpenAI)

You'll have to watch the full AI Planned It series to know exactly what went right and wrong in our experiment, but to cut to the chase, yes, the PC works and no fires were started in the process.

That said, I did have the aforementioned computing expert Matt Hanson on hand to fact-check and sanity-test ChatGPT's building instructions, which proved invaluable to stop us from making dumb rookie mistakes that the AI was ignorant of.

My big takeaway? If you try to place all the onus on AI, you're more likely to make a mistake than you are if you instead watch a few videos, read the documentation and learn for yourself.

However, it can be a fantastic companion for troubleshooting, researching and finding information written by experts; just moderate your usage if you care about its environmental impact, how its prosumer rollout leads to job losses, and perhaps most crucially, its impact on critical thinking.

Watch the full episode of TechRadar's AI Planned It below!

I let ChatGPT instruct my first ever PC build - will it run? - YouTube I let ChatGPT instruct my first ever PC build - will it run? - YouTube
Watch On

You can check out the top recommended builds for both ChatGPT and Gemini, and if you subscribe to our YouTube channel, you'll get so see what we let AI plan next.


Follow TechRadar on Google News and add us as a preferred source to get our expert news, reviews, and opinion in your feeds. Make sure to click the Follow button!

And of course you can also subscribe to our YouTube channel and follow TechRadar on TikTok for news, reviews, unboxings in video form, and get regular updates from us on WhatsApp too.

TOPICS
Josephine Watson
Managing Editor, Social & Engagement

Josephine Watson is TechRadar's Managing Editor - Social & Engagement. Josephine is an award-winning (PPA 30 under 30 2024), NCTJ-trained journalist. Having previously written on a variety of topics, from pop culture to gaming and even the energy industry, she joined TechRadar to support general site management. She is a smart home nerd, champion of TechRadar's sustainability efforts as well and an advocate for internet safety and education. She has used her position to fight for progressive approaches towards diversity and inclusion, mental health, and neurodiversity in corporate settings. Generally, you'll find her fiddling with her smart home setup, watching Disney movies, playing on her Switch, or rewatching the extended edition of Lord of the Rings... again.

You must confirm your public display name before commenting

Please logout and then login again, you will then be prompted to enter your display name.