Live data from Hacker News

Show HN: Summarizing product reviews into simple bullet-point lists with GPT-3

buyforlife.com

11–20 of 43 posts

Re: Show HN: Summarizing product reviews into simple bullet-point lists with GPT-3

#14
One of the things that makes these GPT3 startups hard, and in my opinion most likely non viable, is that GPT3 is not ready for production.

It can produce text that at a glance looks like it could have been written by a human, but a human that is not very competent at the task it set out to do and makes basic mistakes. For example:

https://www.buyforlife.com/products/efa8b0c1-d7da-4485-b7f0-...

- Blade will rust easily unless oiled.

- The locking mechanism is not as secure as it could be, and can be difficult to disengage when the blade is being used for heavy cutting tasks.

- The blade is softer than most stainless steel blades, so the knife will dull quicker than others.

There's also no way for the startup to fix these issues, unless they have the resources to actually improve GPT3 itself.

Re: Show HN: Summarizing product reviews into simple bullet-point lists with GPT-3

#15
Not really related to the GPT-3 thing, which seems cool, but a point of confusion and / or criticism:

> if warranty = lifetime, then price/expected lifetime

The #4 item showing for me is a tape measure, which costs $20 and has an estimated monthly cost of $.28. That means you are estimating this tape measure with a lifetime warranty will only last 6 years.

The #3 item showing for me is an iron skillet, which costs $180 and has an estimated monthly cost of $.25. That means you are estimating this skillet with a lifetime warranty will last 60 years.

Later on down the list you have another cast iron skillet, which cost $20 and has an estimated monthly cost of $.29. This means you are estimating this skillet will last less than 6 years. I happen to own this skillet. It's made of solid iron. I promise you no one who is taking care of this thing at all is going to see less than 10 years of use out of it. I plan to hang on to mine for decades.

What's the basis for this estimation? Pretty much every one I've seen seems completely random and mostly unjustified. E.g. if my tape measure with a lifetime warranty breaks after 5 years I'm definitely taking it in for a free replacement, so what's the deal here? (My high quality tape measures have never broken that quickly anyway.)

All that said I think the implementation of "badges" was really neat and what I can see of the GPT reviews so far look pretty good (although I'm a bit worried that scraping certain review sites may lead to a garbage in, garbage out problem). I'll be checking out your site again in the future.

Re: Show HN: Summarizing product reviews into simple bullet-point lists with GPT-3

#16
post #14

One of the things that makes these GPT3 startups hard, and in my opinion most likely non viable, is that GPT3 is not ready for production. It can produce text that at a glance looks like it could have been written by a human, but a human that is not very competent at the task it set out to do and makes basic mistakes. For example: https://www.buyforlife.com/products/efa8b0c1-d7da-4485-b7f0-... - Blade will rust easil…

What am I missing? What is wrong with those? I read the complete review and honestly, if I didn't know better, would have thought that a human wrote it.

Re: Show HN: Summarizing product reviews into simple bullet-point lists with GPT-3

#17

It would be good to have links to the original reviews which were used for bullet points.

The reviews are actually listed/linked at the bottom of each product page.

Possibly, the parent commenter is looking for per-bullet links? (I'd want that as well.)
Post reply on HN