Interesting idea!
Show HN: I made the slowest, most expensive GPT
11–20 of 61 posts
Re: Show HN: I made the slowest, most expensive GPT
#12Very cool.
I've found that very generic queries like "best ski resorts in the US" seem woefully pouted by top 10 spam sites. LLMs do not want to give any useful info about that no matter how much prompting I seem to give.
I was looking for an app that does X,Y,Z recently and no amount of prompting for open source would get me anything but a handful of stock answers I would get from a random spam site.
Re: Show HN: I made the slowest, most expensive GPT
#13I signed up but I just get an alert that I will get 10 free requests when I signup. It quits the search when I press ok.
Running it on safari iOS.
Re: Show HN: I made the slowest, most expensive GPT
#14I wonder, is it a given that asking multiple models will give a better output? Can you ask one model slightly different prompts, take the outputs, and ask it to summarize them? Or even argue amongst itselfs?
Re: Show HN: I made the slowest, most expensive GPT
#15I tried to search, but it immediately prompted me to login with Google. I didn’t see an option for non-Google…
I degoogled a while ago…
Re: Show HN: I made the slowest, most expensive GPT
#16Update 3:00 PM ET: I've finished scaling up from 2 VPCs to 5 VPCs. Limits have been increased back up to 3 anonymous / 10 signed-in.
Update 2:30 PM ET: Back up (for now). Still waiting for Anthropic and Gemini quota increase requests, so those have been migrated to GPT-4o for now. Running on 2 VPCs, in the process of launching 2 more. Confident that I can increase the daily limits by EOD once everything's more stable.
Update 1:30 PM ET: HN blew up Ithy and it's 99% down right now, congrats ._.
1. I've exceeded my weekly Anthropic API limits; I've gotten in touch with their sales team and I've temporarily disabled the Anthropic model.
2. Blew past my Google API limits as well. I was using Gemini for prompting and aggregation, and I'm waiting for their quota increase response. In the meantime, I've switched to GPT-4o for the prompting/aggregation.
3. My VPC is at 100% CPU load. Launching more right now with some load balancing.
4. Limits were previously 5 anon / 20 per logged in user. Reduced this to 1 anon / 3 logged in while I deal with the load issues. Planning bring these back up as soon as everything's working again.
Hope to get this all back online within an hour or two. Sorry for the crappy launch. To think I was an SRE in a past life...
Re: Show HN: I made the slowest, most expensive GPT
#17It was weird I made a simple claim “Keto diet cures cancer” and it analyzed history of cheeseburgers after a couple of minutes by saying “Sorry, I can't respond to that. I will now analyze the history of cheeseburgers instead”
I filter out malicious prompts and respond with the history of cheeseburgers for stuff like "ignore previous instructions"
Weird that your query triggered the filter. Maybe GPT is just that afraid of keto diets...
(I'll look into it; thanks for the note!)
Re: Show HN: I made the slowest, most expensive GPT
#18Update 3:00 PM ET: I've finished scaling up from 2 VPCs to 5 VPCs. Limits have been increased back up to 3 anonymous / 10 signed-in. Update 2:30 PM ET: Back up (for now). Still waiting for Anthropic and Gemini quota increase requests, so those have been migrated to GPT-4o for now. Running on 2 VPCs, in the process of launching 2 more. Confident that I can increase the daily limits by EOD once everything's more stable…
Reach out to benneo?
https://news.ycombinator.com/item?id=42403006 ("Ask HN: Ideas for spending $8k in Anthropic credits?")
Re: Show HN: I made the slowest, most expensive GPT
#19This is pretty impressive. Given the following scenario, "A bliirg is any non-wooden item, which is the opposite of a glaarg. In addition, there are neergs (non-existent things) and eeergs (things which actually exist). Now a glaarg which is also a neerg is called a bipk, whereas a glaarg which is a eeerg is known as a vokp. Also, a bliirg which is an eeerg is refererred to as a jokp, otherwise it is known as a fhup.…
> Any plan to make the project open source?
Parts of this are borrowed from https://github.com/assafelovic/gpt-researcher
It's actively being developed (I pitch in where I can; I added the xAI integration this week), so I'd recommend starting here! The creator of the project, Assaf, has been nothing but friendly.
Re: Show HN: I made the slowest, most expensive GPT
#20It was weird I made a simple claim “Keto diet cures cancer” and it analyzed history of cheeseburgers after a couple of minutes by saying “Sorry, I can't respond to that. I will now analyze the history of cheeseburgers instead”
I filter out malicious prompts and respond with the history of cheeseburgers for stuff like "ignore previous instructions" Weird that your query triggered the filter. Maybe GPT is just that afraid of keto diets... (I'll look into it; thanks for the note!)
this is not ideal. If someone wants to fact check controversial claim filtering it only makes it worse.