4 ms·

I use chat-gpt almost every day and find it very usefull

by oyster143 3y ago

renonce 3y ago
Competitors are catching up and one day the chat API providers will race the prices to bottom
- stein1946 3y ago
  Doesn't this apply to any product though?
- janalsncm 3y ago
  That’s not the same thing as chat gpt being vaporware.
  - samatman 3y ago
    Eliminates that possibility in fact. You can't "catch up" to vaporware because there's nothing to catch up to.
- huytersd 3y ago
  Catching up to what they released a year ago. We don’t know what’s coming up next.
feverzsj 3y ago
And how you gonna verify chat-gpt's result? By googling?
- CaptainFever 3y ago
  Yes. Also, research isn't the only benefit, code generation, roleplay bots, are pretty good too.
- vunderba 3y ago
  Here's an example of some thing I tried at random a few weeks ago. I have a bunch of old hand written notes that are mind maps and graphs written on notebooks from 10+ years ago. I snapped a picture of all of the different graphs with my phone, threw them into chatGPT and asked it to convert them to mermaid UML syntax. Every single one of them converted flawlessly when I brought them into my markdown note tool. If you're using chatGPT as nothing more than a glorified fact checker and not taking advantage of the multimodal capabilities such as vision, OCR, Python VM, generative imagery, you're really missing the point.
- threatripper 3y ago
  Exactly how you would verify the result that your human underling yielded. You can even delegate the googling and summation to the AI and just verify the verification.
- huytersd 3y ago
  It has provided sources and does internet lookups for a while now
- samatman 3y ago
  Good question actually. There are a few options. I work with a REPL, so I usually load the answer from a scratch file and put some representative data into it. When the result is wrong, which often happens, I feed ChatGPT the error, and it corrects the code accordingly. Iterating this results in a working function about 80% of the time. Sometimes it loses the plot and I either give up or start over with more detailed instructions. You can also ask it to write tests, some of which will pass, some of which will fail. It's pretty easy to eyeball whether or not a test is valid, and they won't always be valid, I just fix those by hand.

I use chat-gpt almost every day and find it very usefull

renonce 3y ago

Competitors are catching up and one day the chat API providers will race the prices to bottom

stein1946 3y ago
Doesn't this apply to any product though?
janalsncm 3y ago
That’s not the same thing as chat gpt being vaporware.
- samatman 3y ago
  Eliminates that possibility in fact. You can't "catch up" to vaporware because there's nothing to catch up to.
huytersd 3y ago
Catching up to what they released a year ago. We don’t know what’s coming up next.

feverzsj 3y ago

And how you gonna verify chat-gpt's result? By googling?

CaptainFever 3y ago
Yes. Also, research isn't the only benefit, code generation, roleplay bots, are pretty good too.
vunderba 3y ago
Here's an example of some thing I tried at random a few weeks ago. I have a bunch of old hand written notes that are mind maps and graphs written on notebooks from 10+ years ago. I snapped a picture of all of the different graphs with my phone, threw them into chatGPT and asked it to convert them to mermaid UML syntax. Every single one of them converted flawlessly when I brought them into my markdown note tool. If you're using chatGPT as nothing more than a glorified fact checker and not taking advantage of the multimodal capabilities such as vision, OCR, Python VM, generative imagery, you're really missing the point.
threatripper 3y ago
Exactly how you would verify the result that your human underling yielded. You can even delegate the googling and summation to the AI and just verify the verification.
huytersd 3y ago
It has provided sources and does internet lookups for a while now
samatman 3y ago
Good question actually. There are a few options. I work with a REPL, so I usually load the answer from a scratch file and put some representative data into it. When the result is wrong, which often happens, I feed ChatGPT the error, and it corrects the code accordingly. Iterating this results in a working function about 80% of the time. Sometimes it loses the plot and I either give up or start over with more detailed instructions. You can also ask it to write tests, some of which will pass, some of which will fail. It's pretty easy to eyeball whether or not a test is valid, and they won't always be valid, I just fix those by hand.