Show HN: PC Builder AI(pcbuilderai.com)
pcbuilderai.com
Show HN: PC Builder AI
https://www.pcbuilderai.com/
83 comments
What part of this needs ai? This looks like it could be done with a database and a price list. I think this suggestion is already said, but the main push of "ai" is mostly in terms of natural language processing, so why not allow a prompt? Give a default prompt like "I'd like to build a gaming pc with a budget of $1000." You could even make it a fill in the blank like "I'd like to build a _______ pc with ___special features____ for a budget of $____." The thing about it's current iteration is regardless of there being "ai" in it, it's not clear what it's doing that's more special than me asking a pc building community for a parts list with a budget.
To continue this idea (feel free to steal this), what if this tool looked more like this: https://play.tailwindcss.com/KqFF5wqk5I
Have a conversation with the AI about the PC you want, and challenge it on things like "can I see the equivalent PC but with AMD instead of Intel" and things like that. I think that would have extremely good value. And for prices, you might be better off scraping amazon (or whatever site makes the most sense), instead of whatever stale data you have. Prices for PC parts fluctuate so much that I don't know that it make sense to try and store/cache it locally.
Have a conversation with the AI about the PC you want, and challenge it on things like "can I see the equivalent PC but with AMD instead of Intel" and things like that. I think that would have extremely good value. And for prices, you might be better off scraping amazon (or whatever site makes the most sense), instead of whatever stale data you have. Prices for PC parts fluctuate so much that I don't know that it make sense to try and store/cache it locally.
this makes a lot of sense, thanks for the feedback! We'll think more about your idea and get back to you here
Non PC experts need this sort of thing.
No. They need something which deterministically tells them which parts are appropriate for their budget and usecase. Putting AI in there is ridicolous.
I don't disagree, but from what I see, it could be executed faster with just a good database query. It's not clear what is AI about this, or what benefit AI actually gives it as a tool.
I don't understand your favor for a database query. What database, who's querying it, and why would that be better than a UI? Ain't nobody opting for database queries as their go-to interface for anything, and it would be silly to recommend anyone use sql to build their pc. The benefit the tool provides is not having to interact with a database.
Sorry, I probably wasn't clear on my point. What I was trying to convey is that whatever the ai model is doing behind the scenes doesn't appear to be any more powerful than having a set of pre-built machine specs in a database and querying them for price and type. I'm sure there is more at work here in reality, but since it takes a few minutes for the ai to build a machine from 2 parameters, it seems silly that you need ai for this at all.
Now, there are ways to make the ai more useful, and that would be to have a conversation with it about what you want to do on the computer. Then that makes the ai use apparent and reasonable.
Now, there are ways to make the ai more useful, and that would be to have a conversation with it about what you want to do on the computer. Then that makes the ai use apparent and reasonable.
Ah I see, so like Rock Auto vs ChatGPT
Pcpartpicker does this perfectly
i used the logical increments guide back in the day
Not sure where you’re getting pricing data, but it’s clearly out of date. A Ryzen 5 3600 is $97 on Newegg ($92 after rebate), not $350. Tools like this absolutely must have up-to-date pricing to be useful.
I agree broadly. I think this helps to narrow down what pricing for these types of expenses should be, though. If I'm trying to spend $1000 on a PC (the last time I built a PC was in 2013, can you tell?) and I don't have intimate knowledge of the prices of specific parts, I'll probably try to get a vague sense of proportions for "what should the ratio of CPU cost to GPU cost be?" I think this tool helps with that. Then, when I'm ready to load up a cart, I can use something like PC Part Picker and verify my order.
I think this is a good tool, but not a great tool. Some refinement is needed, but the API is good, if sparse.
I think this is a good tool, but not a great tool. Some refinement is needed, but the API is good, if sparse.
I was tired of waiting after 30 seconds or so. I think it would help to have a more meaningful progress reporting, like flight ticket services have. 99% Invisible had a nice episode on the subject. https://99percentinvisible.org/episode/wait-wait-tell-me/
If anyone is looking for a different website to help with these kinds of choices, I’ve used https://www.logicalincrements.com many times over the years and have been happy with the recommendations there.
Great site thanks for sharing it! Really good breakdown.
this is very powerful and complete, thanks for sharing!
PC Builder AI is more of a fun project that might help by giving you direction of what to search
PC Builder AI is more of a fun project that might help by giving you direction of what to search
I'm sorry but this seems like a moneygrab AI keyword operation.
Also I'm a bit worried about the other projects I've found clicking the dropdown:
https://www.duplicateduplicate.com is very disturbing.
Also it never generated a PC for me...
Also it never generated a PC for me...
we make a lot of crazy stuff here and I respect your opinion (:
did you check out https://www.seanceai.com it's even weirder, I admit, but people seemed to like it, search for "Seance AI" on Google
Could try again generating the PC, please?
did you check out https://www.seanceai.com it's even weirder, I admit, but people seemed to like it, search for "Seance AI" on Google
Could try again generating the PC, please?
I built a desktop seance AI app back using Markova chains and transcripts of conversations back in the early 2000, then later black mirror went one step further, and now we're here.
Combine LLM with tacotron TTS and you've pretty much got a means to virtually resurrect somebody.
Combine LLM with tacotron TTS and you've pretty much got a means to virtually resurrect somebody.
It recommended I pay the $150 extra to downgrade from the 3950X to the 7950X.
Also, that I pay $1400 for a 3080 TI, but a substantially less powerful GPU would do just fine. I asked for a workstation, not a gaming rig.
Depends on what "workstation" means to you... For some, it literally means AI work, and would suggest a higher end GPU with plenty of VRAM. For others, it may mean really memory heavy work, or high CPU process work. For me, it's the latter, for others it could be the former. YMMV, seriously.
I agree. "Workstation" is too general.
neat.
newegg has something similar -works pretty well
https://www.newegg.com/tools/custom-pc-builder/
newegg has something similar -works pretty well
https://www.newegg.com/tools/custom-pc-builder/
Is there a tool similar to this one but for searching through the offerings of large manufacturers like Lenovo or Dell? In other words, based on my criteria, it would find PCs with configurations closest to my needs.
I need a relatively high-performance i9-based desktop PC (for ML and toying with LLMs) with on-site warranty, so building my own PC would not work for me.
I need a relatively high-performance i9-based desktop PC (for ML and toying with LLMs) with on-site warranty, so building my own PC would not work for me.
Doesn't seem to be working 100%, but it's a neat idea :)
I put I wanted the computer to cost $8000, and it generated components that cost $5340 together (1000 + 500 + 2100 + 700 + 90 + 250 + 500 + 200) but still the UI showed the correct prices for the individual items, but the total still showed as "$8000" although adding them together is nowhere near that.
I put I wanted the computer to cost $8000, and it generated components that cost $5340 together (1000 + 500 + 2100 + 700 + 90 + 250 + 500 + 200) but still the UI showed the correct prices for the individual items, but the total still showed as "$8000" although adding them together is nowhere near that.
Yeah... I randomly punched in $40,000 for Gaming and got
(1) Parts that total $9,980, but claims it totals $33,950 (2) Recommendations for parts that are both years out of date, and vastly inferior to other options (CPU = AMD Ryzen Threadripper 3990X, GPU = RTX3090)
So I'm guessing the AI data isn't trained enough and is going off some poor metrics like sorting by price, perhaps recommending multiples without indicating (was it asking me to double up on RAM or Stroage?)
(1) Parts that total $9,980, but claims it totals $33,950 (2) Recommendations for parts that are both years out of date, and vastly inferior to other options (CPU = AMD Ryzen Threadripper 3990X, GPU = RTX3090)
So I'm guessing the AI data isn't trained enough and is going off some poor metrics like sorting by price, perhaps recommending multiples without indicating (was it asking me to double up on RAM or Stroage?)
interesting... thanks for the feedback, I'll see what happened
Seems to be hugged to death - every request I run ends in a 500/504 - and there is no error handling to display an error when it happens.
What is the backing AI/data source? I tried using ChatGPT for this, but given it's cutoff is September 2021 it doesn't really give you the most relevant hardware.
What is the backing AI/data source? I tried using ChatGPT for this, but given it's cutoff is September 2021 it doesn't really give you the most relevant hardware.
Pretty cool, but for all of what I could imagine AI could do, this seems a bit limited. Does it just generate the same thing each time? This could just be some premade lists.
When I think of an AI I'd imagine some more personalisation, and there's so much that goes into building a PC but the only input here is the price.
What about appearance? Opaque causes, transparent cases, white vs black PCBs, RGB? Preferences around the prioritisation of certain specs (more storage?). Then there's the many form factors.
Even for just a "gaming" build - are you going for blockbuster graphics, or more CPU-intensive simulation genres.
When I think of an AI I'd imagine some more personalisation, and there's so much that goes into building a PC but the only input here is the price.
What about appearance? Opaque causes, transparent cases, white vs black PCBs, RGB? Preferences around the prioritisation of certain specs (more storage?). Then there's the many form factors.
Even for just a "gaming" build - are you going for blockbuster graphics, or more CPU-intensive simulation genres.
Thanks for the feedback! You are right. We made this tool to be as simple as possible, but it makes sense to have something more powerful. We will test and add these options soon (:
Perfect example of what’s wrong with AI - unless you command massive trust, this is worse than asking a random stranger on reddit. There’s no explanations (you get that on reddit/lemmy).
Many startups will find a rude awakening about how much the vast majority of consumers hate and fear AI at this point when they roll out “trust his AI” bots.
Many startups will find a rude awakening about how much the vast majority of consumers hate and fear AI at this point when they roll out “trust his AI” bots.
I feel like this is a linear optimization problem (maximize benchmark score for games and/or for work) constrained on price / part compatibility, not something which you can just feed into ChatGPT. Unfortunately, the compatibility constraints are tricky, as well as understanding what the benchmark score would be with a nonlinear set of items.
What I would do is classify each item to the best item year (similar to FB's best device year), find benchmark scores for representative sets of item-years, and use that as a 'oracle' to guide the solver.
What I would do is classify each item to the best item year (similar to FB's best device year), find benchmark scores for representative sets of item-years, and use that as a 'oracle' to guide the solver.
seems like this was posted for feedback, here's mine
- seems like the market data about components is out of date, so I wouldn't actually trust this.
- pcpartbuilder is helpful in large part because there are links to actually buy the components
- really what I'd want to do is describe what I want to do with the computer, then have it suggest things -- not just select from a dropdown. That would actually be helpful; there's a learning curve to stay abreast of the latest hardware things and really what I would want is "tell me what's the good stuff to buy" like you might find if you google "best workstation build for ai" or watch a Linus tech tips video
- I'd want to take a look at the list of components and edit them and have it recalibrate, like "actually I think this case looks really cool" and then have it automatically readjust components based on for example things that fit in the case
- I'm guessing this is using a LLM based on the out-of-date market data. For the above, I think you would really just need a robust set of apis about pc part data and then use LLMs to function call into the apis. It seems like pcpartpicker is the group with this data, not sure if they expose it to other developers
- seems like the market data about components is out of date, so I wouldn't actually trust this.
- pcpartbuilder is helpful in large part because there are links to actually buy the components
- really what I'd want to do is describe what I want to do with the computer, then have it suggest things -- not just select from a dropdown. That would actually be helpful; there's a learning curve to stay abreast of the latest hardware things and really what I would want is "tell me what's the good stuff to buy" like you might find if you google "best workstation build for ai" or watch a Linus tech tips video
- I'd want to take a look at the list of components and edit them and have it recalibrate, like "actually I think this case looks really cool" and then have it automatically readjust components based on for example things that fit in the case
- I'm guessing this is using a LLM based on the out-of-date market data. For the above, I think you would really just need a robust set of apis about pc part data and then use LLMs to function call into the apis. It seems like pcpartpicker is the group with this data, not sure if they expose it to other developers
The prices really are not exact, but they are close. Fetching the prices from real-time sources and then using AI to transform them into information seems to be a good approach (currently we're using AI only)
We will add links soon. We have this feature already, it's just not available because the prices don't match (the AI price is higher than the actual Amazon price, for instance)
Your feedback is much appreciated and very very helpful, thank you so much.
We will add links soon. We have this feature already, it's just not available because the prices don't match (the AI price is higher than the actual Amazon price, for instance)
Your feedback is much appreciated and very very helpful, thank you so much.
It should ask more questions about your preferences. Some person would value smaller size, someone would value more speed per dollar, someone values upgradability (i.e. more PCIe slots, latest CPU socket) and durability above everything else. E.g. I would mostly prefer to get a cheaper CPU but a better motherboard so I can just buy a better CPU later once I need it and have more spare money.
thanks for the feedback, we're adding more preferences soon. will let you know when it's deployed
Newegg.com actually has an AI PC builder where you describe your needs in natural language and it spits out parts. https://www.newegg.com/tools/custom-pc-builder
Gave ok suggestions for a $500 desktop, apart from including a redundant $45 cooler (Ryzen 3200g already has a passable cooler in the box).
Changing to a $550 build keeps the same cpu but changes its price from $100 to $170, adds a $20 cooler, and drops Ram from 16gb to 8gb.
Sorry for the negativity but algorithm needs work.
Changing to a $550 build keeps the same cpu but changes its price from $100 to $170, adds a $20 cooler, and drops Ram from 16gb to 8gb.
Sorry for the negativity but algorithm needs work.
indeed it needs more work, thanks for the feedback! It helps us to make it better (:
Getting HTTP 500 response when I attempt to generate a build. Seems like the intent isn't to charge money for a useful tool, but rather have the traffic funnel into your agency business. I would highly suggest fixing your errors if that's the case.
I’ve chosen “Gaming” and $4000. What I’ve got is quite strange:
* Ryzen 5950x — a 16-core Zen 3 CPU for gaming in 2023?
* RTX 3090 for 1500 — you can get a 4090 FE for about that price
* an X570E motherboard which generally doesn’t make a lot of sense for gaming
* a 1200W PSU which is wildly overspecced for this build
* Ryzen 5950x — a 16-core Zen 3 CPU for gaming in 2023?
* RTX 3090 for 1500 — you can get a 4090 FE for about that price
* an X570E motherboard which generally doesn’t make a lot of sense for gaming
* a 1200W PSU which is wildly overspecced for this build
It's recommending a $42 1TB HDD. I can get a 1TB SSD for less than that...
Just tried it and, asking for a $2000 workstation, it recomended right away a 3700x for $340 dollars. I can find a faster, cheaper option (7600x) for $244. And a rtx 3060 for $650? It's just plain wrong
by the way https://www.cheerleaderai.com/ is great!
thanks! That's what we do. Simple, cool and Agency Increasing tools like this (all for free)
Hug of death? Sorry, my attention span on websites is < 15 sec unless it's giving me a progress bar for stable diffusion.
Thank you all for the feedbacks, we will be doing our best to improve and fix what you guys reported. Keep doing it!
Pretty useful tool!
Do you verify that all the parts you suggest are compatible ?
Are you able to complement with actual build instructions ?
Do you verify that all the parts you suggest are compatible ?
Are you able to complement with actual build instructions ?
> Do you verify that all the parts you suggest are compatible ?
Generating a part list compatible with pcpartpicker.com could fix that easily if not.
> Are you able to complement with actual build instructions ?
It's almost easier to build a PC today than a 100 piece LEGO kit. Otherwise, most components come with manuals for how to assemble them together with the other pieces too, like the motherboard comes with instructions on where things go.
Generating a part list compatible with pcpartpicker.com could fix that easily if not.
> Are you able to complement with actual build instructions ?
It's almost easier to build a PC today than a 100 piece LEGO kit. Otherwise, most components come with manuals for how to assemble them together with the other pieces too, like the motherboard comes with instructions on where things go.
I think CPU cooling mounts are probably the only thing that's really difficult, and even then, the types of mistakes you're likely to make (namely uneven tension or not enough thermal compound) take some hands on experience to get right.
I've walked 3 different under 21 kids through a build since start of the pandemic... only because my hand/eye coordination for some bits (panel headers, etc) isn't so great combined with low light visibility.
I've walked 3 different under 21 kids through a build since start of the pandemic... only because my hand/eye coordination for some bits (panel headers, etc) isn't so great combined with low light visibility.
I dunno, there is a bunch of motherboard/RAM/CPU combinations that don't work properly together. Just because the socket is right, doesn't mean the component supports it.
Slightly recent example is newly launched CPU architectures using contemporary sockets, where you have to insert a supported CPU first, update BIOS and only then could the motherboard support newly launched CPU. But if you just have the one, new CPU, it might not be able to boot.
Slightly recent example is newly launched CPU architectures using contemporary sockets, where you have to insert a supported CPU first, update BIOS and only then could the motherboard support newly launched CPU. But if you just have the one, new CPU, it might not be able to boot.
That's fair as well.. matching launch chipset and CPU is a pain. Also, even in Phoenix, I had trouble getting access to an older CPU to update a MB chipset for a new build last year... wound up spending $100 more on a newer chipset MB. It had bios flashback, but that didn't work.
Seems like this is something PC Part Picker (with their metadata on compatibility and slots and such) could build without AI.
It's easy but it would be good to have a simple "How to". We'll add soon
> Corsair Vengeance LPX 16GB (2x8GB) DDR4 DRAM 3200MHz
> 75 USD
Completely overpriced. Runs for 35 EUR in Germany
> 75 USD
Completely overpriced. Runs for 35 EUR in Germany
I guess this is what you get when you train it on data from the chip shortage (high prices) period
Their suggestion for a $1k workstation is laughable. I can build a comparable system for $250.
What is with the Bitcoin logo on the spinner?
fwiw bard and gpt-4 seem to do a pretty similar job at this
gpt-4 actually gave me ~90% similarity on the suggestions too
gpt-4 actually gave me ~90% similarity on the suggestions too
feature suggestion: start with some components and generate what fits with them
It gave me a P2200 GPU. lol.
great if you want to price pc parts for 2021
You know when you wanna build a new PC, and need to watch a bunch of videos to find the best parts, and then tweak your shopping cart to fit your budget? Well, this new app can help you.
It recommends the best hardware for your usage (gaming or work), and considers your budget. It's really helpful, even when you got expertise on hardware. Check it out, and feel free to suggest stuff! Hope you like it!
https://www.pcbuilderai.com/