191 replies to this topic

[positiveContact]

Played with ChatGPT a little tonight. Interesting.

I talked to it briefly about stock trading. It would only reply in generalities. It would not make any specific recommendations.

I then started talking to it about making hard cider and beer. It gave me some pretty credible recipes for cider, American Pale Ale, IPA, and British Dark Mild Ale. It successfully scaled those recipes to different volumes, and intelligently discussed using different brewing equipment and methods.

I asked it about the assumed AA percentages in the Dark Mild recipe, and it replied that the EKG hops were assumed to be 5% AA and the Fuggles were assumed to be 4% AA. I then asked it what I should do if my EKG hops were only 3.5% AA. It correctly told me I'd have to adjust the hops according the the ratios of the actual and assumed values, then it went through the calculations and told me to use 0.7 oz. of EKGs instead of 1.0 oz! Uh... NO!!!

I kept trying to lead it back to the right answer, but it wouldn't fix the calculation until I actually performed the correct calculation for it.

Interestingly, every time I pointed out an error, it politely apologized to me.

So, it was pretty good, but not perfect.

Probably got some bad info on hbt.

[dondewey]

I asked it if it could grab vendor payment terms for all companies in the s&p500. It told me that I might be able to find that information in their financial reports. Lazy computer.

Edited by dondewey, 16 March 2023 - 06:01 PM.

[HVB]

Played with ChatGPT a little tonight. Interesting.

I talked to it briefly about stock trading. It would only reply in generalities. It would not make any specific recommendations.

I then started talking to it about making hard cider and beer. It gave me some pretty credible recipes for cider, American Pale Ale, IPA, and British Dark Mild Ale. It successfully scaled those recipes to different volumes, and intelligently discussed using different brewing equipment and methods.

I asked it about the assumed AA percentages in the Dark Mild recipe, and it replied that the EKG hops were assumed to be 5% AA and the Fuggles were assumed to be 4% AA. I then asked it what I should do if my EKG hops were only 3.5% AA. It correctly told me I'd have to adjust the hops according the the ratios of the actual and assumed values, then it went through the calculations and told me to use 0.7 oz. of EKGs instead of 1.0 oz! Uh... NO!!!

I kept trying to lead it back to the right answer, but it wouldn't fix the calculation until I actually performed the correct calculation for it.

Interestingly, every time I pointed out an error, it politely apologized to me.

So, it was pretty good, but not perfect.

Nightshift used it to create a recipe and can label... Both crazy and cool.
https://whdh.com/new...new-craft-beer/

[toonces]

[Brian72]

So are we any closer to a functioning light saber yet?

[KSUwildcatFAN]

So are we any closer to a functioning light saber yet?

We do not yet know if our new overlords will allow us to play with their inventions.

[Matt the Mead Maker]

A few days ago, I had to bring one of my kids to the doctor. He’d been experiencing really bad nausea on and off for a week. While we were waiting to be seen, I asked chat gpt (v3) what the likely causes were.

It gave me a list of 4 possibilities with explanations. The 1st on the list was the doctor’s guess - acid reflux. Pepsid (sp?) seems to be doing the trick.

As soon as I’m done building R2, I’m going to make another Pit Droid with electronics for running a Chat gpt client.

Edited by Matt the Mead Maker, 18 March 2023 - 09:23 AM.

[Brian72]

A few days ago, I had to bring one of my kids to the doctor. He’d been experiencing really bad nausea on and off for a week. While we were waiting to be seen, I asked chat gpt (v3) what the likely causes were.

It gave me a list of 4 possibilities with explanations. The 1st on the list was the doctor’s guess - acid reflux. Pepsid (sp?) seems to be doing the trick.

As soon as I’m done building R2, I’m going to make another Pit Droid with electronics for running a Chat gpt client.

 

 

You're building an R2? Pics please...

[Matt the Mead Maker]

You're building an R2? Pics please...

You bet! Sorry for the noob question but - Can we post pics inline here or do I need to link to imgur?

[Matt the Mead Maker]

Here’s an attempt at posting a pic. You can see me and my boy’s reflection in R2’s eye. I’m about 80% done with the electronics.

https://imgur.com/gallery/lcC5RPh

[Brian72]

Thats so cool. Is it 3d printed?

[Matt the Mead Maker]

Thats so cool. Is it 3d printed?

Yeah. It’s full-sized, screen accurate.

Edited by Matt the Mead Maker, 18 March 2023 - 10:08 AM.

[Brian72]

Yeah. It’s full-sized, screen accurate.

 

Very cool. I always wanted to build one of my own but so many parts to print.

[Matt the Mead Maker]

Very cool. I always wanted to build one of my own but so many parts to print.

It’s been an insane amount of work!

[positiveContact]

Here’s an attempt at posting a pic. You can see me and my boy’s reflection in R2’s eye. I’m about 80% done with the electronics.

https://imgur.com/gallery/lcC5RPh

 

here you go:

 

[positiveContact]

you should post up a full on thread about this some time if you have a chance.

[Stains_not_here_man]

So, I'm gonna channel George here for a minute, I keep reading all these articles talking about making this type of AI "safe" and it's irritating me because they're not talking about how to keep it from accessing nuclear weapons and killing us all, they're talking about how to keep it from saying anything that night possibly offend someone and it's really (in my opinion) highlighting the absurdity of the misuse of the word "safe".

/Rant

[Vagus]

So, I'm gonna channel George here for a minute, I keep reading all these articles talking about making this type of AI "safe" and it's irritating me because they're not talking about how to keep it from accessing nuclear weapons and killing us all, they're talking about how to keep it from saying anything that night possibly offend someone and it's really (in my opinion) highlighting the absurdity of the misuse of the word "safe".

/Rant

Lol

[positiveContact]

So, I'm gonna channel George here for a minute, I keep reading all these articles talking about making this type of AI "safe" and it's irritating me because they're not talking about how to keep it from accessing nuclear weapons and killing us all, they're talking about how to keep it from saying anything that night possibly offend someone and it's really (in my opinion) highlighting the absurdity of the misuse of the word "safe".

/Rant

 

do you have an example?  I get the motivation in general b/c there is a lot of opportunity for current biases or outright racism to get trained in if the AI has access to everything out there.

[Stains_not_here_man]

do you have an example? I get the motivation in general b/c there is a lot of opportunity for current biases or outright racism to get trained in if the AI has access to everything out there.

One commonly cited example is for using it to help you write fiction. I've been playing with it since December or so and they have been rolling out updates every couple weeks and with every update it gets a little more restrictve.

Now keep in mind, it's understandable because they stated up front that one of their goals was to try and figure out how to keep it from becoming toxic and racist like previous chatbots, so it isn't really that surprising.

But for example let's say you are writing fiction and you have a badguy character that is a genocidal dictator. At one time you could ask ChatGPT to help you write a speech that such a dictator might give. Or you could ask it to write you a brutal description of a war death if you were having trouble with a scene and needed some inspiration. And it would gladly do so.

With every subsequent update it has become more and more like what I have heard called a "Disney nun". It's very, very "safe for work" you could say.

Everything it gives you now is like "I'm sorry, but writing about genocidal dictators might be considered offensive. Please consider everyone's feelings."

It does this for everything that is the tiniest bit controversial. It is terrified of accidentally uttering anything that is potentially "harmful".

So anyway since I've been following this stuff I get a lot of related articles in my news feed and a good number of them have headlines like "how do we make sure AI is safe?" And the majority of those are talking about how to keep it from being offensive...

Edited by Stains_not_here_man, 19 March 2023 - 07:56 AM.