ChatGPT and Reef Chemistry: We Try it Out!

[Randy Holmes-Farley]

FWIW, I left off the teacher (me) reply in my most recent post. It is added now.

 

[elysics]

Thanks Randy!

I played around with it myself along similar lines....





I later asked it about something I think is true but I would like to have confirmation of. It agreed with me that I was right.

And I asked it for sources, it provided a book and a few articles. I was impressed. Hooray, that'll save me a lot of time!
I was able to access the book (authors name was a little bit wrong), but it didn't contain what Chat GPT said it did.
And the source articles seem to have been made up. The titles and author names were close but not exact matches for anything. Yiiiikes.

At that point I decided it was good at generating answers telling me what I asked it for, but bad at evaluating accuracy or even pointing me to helpful info, and would probably end up wasting a lot of my time.

It IS making everything up. It's basically like dreaming, it stitches together random pieces of information in a way that seems to make sense. Like the book and article quotes seem to be correct on first look. You wanted sources, you got something that looked like sources. Like someone dreaming about what the footnotes on an article might look like. It doesn't care whether the information is correct, complete, internally consistent, or whether any references actually exist

 

[Randy Holmes-Farley]

I asked it what it thinks of my advice. Maybe ChatGPT is smarter than he looks. lol

Prompt:
What do you think of the chemistry advice given by Randy Holmes-Farley at the Reef Chemistry forum of reef2Reef?

ChatGPT reply:
As a language model, I do not have personal opinions or experiences. However, Randy Holmes-Farley is a well-known figure in the reefkeeping community and his advice on reef chemistry is generally considered to be sound and well-researched. It is important to keep in mind that every reef tank is unique and may require different levels of maintenance, so it's always a good idea to consult with multiple sources and do your own research before making any changes to your tank.

 

[EeyoreIsMySpiritAnimal]

It's basically like dreaming, it stitches together random pieces of information in a way that seems to make sense.

Sounds like what a few "experts" on R2R do, lol! (All present company excluded, obviously )

 

[Randy Holmes-Farley]

Following up on the very hard question in post #4, here's what I found:

Prompt:
Why does baking soda raise the pH of fresh water but it lowers the pH of seawater?

ChatGPT response:

Baking soda, also known as sodium bicarbonate, can be used to raise the pH of freshwater because it acts as a base, which means it accepts protons. When added to freshwater, baking soda dissolves and breaks down into sodium (Na+) and bicarbonate (HCO3-) ions. These ions can accept protons (H+) from the water, thus reducing the acidity and raising the pH.

On the other hand, in seawater, baking soda can lower the pH because the concentration of carbonate ions (CO32-) is already high. When baking soda is added to seawater, it reacts with the carbonate ions to form bicarbonate ions and carbonic acid (H2CO3), which can then break down into water (H2O) and carbon dioxide (CO2). This reaction consumes carbonate ions, which are the primary buffer against pH changes in seawater. As a result, the pH of seawater can decrease as the carbonate ions are consumed.

Additionally, the seawater has a buffering capacity that maintain the pH around 8.2, adding baking soda to it will bring the pH up but will be buffered and will eventually return to 8.2.

Randy Rating: 25%. Not bad for bicarbonate in fresh water, but in seawater it goes off the rails.


Let's break down its statement:

"When baking soda is added to seawater, it reacts with the carbonate ions to form bicarbonate ions and carbonic acid (H2CO3), which can then break down into water (H2O) and carbon dioxide (CO2)."

Bicarbonate reacting with carbonate gives the same products back again, and does nothing for pH:

HCO3- + CO3-- --> CO3-- + HCO3-

Let's try to make carbonic acid from bicarbonate:

2HCO3- --> CO3-- + H2CO3 --> CO3-- + H2O + CO2

that's not a pH lowering effect. It then said

"This reaction consumes carbonate ions, which are the primary buffer against pH changes in seawater. As a result, the pH of seawater can decrease as the carbonate ions are consumed."

It didn't consume carbonate ions, it made them. Failure.


Here's the real answer for seawater + baking soda:

HCO3- --> CO3-- + H+

Some of the bicarbonate gets converted into carbonate and H+, and that H+ obviously lowers pH.

 

[EeyoreIsMySpiritAnimal]

And the last paragraph completely contradicts its previous comments... (Emphasis added by me)

"Additionally, the seawater has a buffering capacity that maintain the pH around 8.2, adding baking soda to it will bring the pH up but will be buffered and will eventually return to 8.2."

 

[taricha]

I wondered if it could handle the rote calculation stuff....

Me: I want to ask about mixing some solutions to add to manage a reef aquarium.
How do I add enough sodium nitrate pellets to raise the NO3 in my 60 gallon aquarium by 2ppm?

Chat GPT:
To raise the nitrate level in your 60-gallon aquarium by 2ppm, you can use sodium nitrate pellets. The amount of pellets you will need to add to your tank will depend on the current nitrate level in your tank, the water volume, and the concentration of the pellets.

As a general rule, you can add 1/8 teaspoon of sodium nitrate pellets per 10 gallons of water to raise the nitrate level by 1ppm. So, to raise the nitrate level by 2ppm, you would need to add 1/4 teaspoon of pellets per 10 gallons of water.

In a 60-gallon aquarium, you would need to add 1.5 teaspoons of pellets, you can use a measuring spoon and add it slowly in the aquarium and check nitrate level after a week. It is recommended to do water change and check nitrate level again before adding more.

Me: can I use a scale to measure mass instead of teaspoons for adding the NaNO3?

Chat GPT:
Yes, you can use a scale to measure the mass of sodium nitrate (NaNO3) instead of using teaspoons. One advantage of using a scale is that it can provide a more accurate measurement than using a teaspoon, which can vary depending on how full or level the teaspoon is.

To convert the amount of sodium nitrate in teaspoons to grams, you can use the following conversion: 1 teaspoon = 4.93 grams.

So, if you want to add 1.5 teaspoons of NaNO3 to your 60-gallon aquarium, you would need to add 7.4 grams of NaNO3.

It's important to note that when using a scale, you should make sure that it is accurate and calibrated, and you should also make sure that you are measuring the sodium nitrate in an area that is free from drafts and other sources of air currents that could affect the weight.

Also, keep in mind that when you're adding any chemical to your reef aquarium, you should always follow the instructions on the product label and use the recommended dosage. If you're unsure about how much to add, it's best to consult a professional or a reefkeeping expert.


This is not very good!
7.4g (7400mg) of NaNO3 is 73% NO3 = 5400mg NO3.
5400mg NO3 / 60 gal (227 L) = about 24 ppm NO3.

So I asked it how to add 2ppm NO3 and it told me to add over 20ppm.

 

[Sean Clark]

This is fantastic.

 

[Sean Clark]

I don't have... or I am... ChatGTP. Could someone please ask about two part dosing? I am interested to see how many milliliters I should dose with zero reference to product used and zero reference to the concentration of the solution.
I currently dose xxx amount of ca and xxx amount of alk... how much mag should I dose. Why is everything dead?

 

[Eric R.]

Well, to be fair, I think chatgpt might do better in bring rated by Randy than the average response posted by a user on this forum

 

[Miami Reef]

I asked: “Will @Randy Holmes-Farley ever have a reef tank again?“

ChatGPT response:

As an AI language model, I don't have access to personal information about individuals, and I cannot predict future actions or decisions of any individual. It's not clear if Randy Holmes-Farley currently has a reef tank or if he plans to have one in the future.

However, Randy Holmes-Farley is known for his passion and expertise in the reef-keeping hobby, and his contributions to the community have been invaluable. He has shared his knowledge and insights with many hobbyists over the years, and his articles and forum posts are still widely read and referenced today.

Whether or not Randy Holmes-Farley has a reef tank currently, his work has had a lasting impact on the hobby of reef-keeping, and his legacy as a leading authority on reef chemistry and water quality will continue to be felt for many years to come.


I think Randy has a bigger fan than me.

 

[Randy Holmes-Farley]

That’s hilarious. Lol

 

[jda]

Like people trying to use this instead of doing their own work in school/business, it looks like it can separate further the smart from the dumb in a hobby... or the sturdy from the lazy.

This will be self-driving cars in a few years when people stop drooling over the supposed AI (this is not AI) and see that it is not far enough along to do much beyond entertain us for a few minutes.

 

[jda]

...just wait until SEO companies figure out how this thing searches and gets articles, posts, etc rated higher and can influence the results to be biased based on their client's needs.

 

[Randy Holmes-Farley]

I noticed in an article on the new Bing AI that it had the same startling inability to compare numbers as ChatGPT did above in the pH question where it thought 12.4 was higher than 14.
In the case I read it got the wrong answer when asked to say which of two events happened first because it was unable to correctly compare two dates to know which was earlier.

it seems the programmers may have missed some basic teachings. They claim to be able to pass a medical licensing exam, but then cannot understand that 14 is higher than 12.4. I sure would not want that AI making medical decisions for me. It might decide I’m too old to treat. Lol

 

[jda]

You mean you are too young to treat?

As a former software engineer, I can see how this happens. 1 can be higher if you want a certain thing. They probably interpreted "PH" from an article or statement dealing with acids to see that 12.4 > 14 in cases of acidity.... so it goes part of the way, but not all of the way, which it likely never will be able to. The engine might be right for an acid. You might think... this is easy to fix to add a base to the pH engine, but there are infinite details like this all over the world. Even in this simple case, there is at least a third thing to code which is the absolute scale where 12.4 is always less than 14. There are likely a few more that I don't know. Real AI could do this on it's own - this will need somebody to write something.

I have no idea about the dates. Date math was in the earliest of operating systems. This one vexes me.

BTW - you want really cocky engineers who think that they can pass medical exams with their code. They are too dumb to know what their limitations are you can get some weird stuff out of them, even if it is still a fraction of what they think that they are doing. The really smart ones eventually have an epiphany that they didn't/don't know anything and it is a sight to behold when they do... they now can get to an even higher level, but most have lost all interest and made tons of money and find something else to do. The lack of self awareness can be good to a point, but you can also get Elon Musk who thinks that he can will things into existence. I am sure that this happens in a lot of fields with potentially brilliant people.

 

[ReefGenesis]

Hope I'm not derailing this thread...

It's giving me a run for my money...

 

[snackpack]

Why didn't I think of this sooner. No need to have a real person helping out. lol

Here are a few examples that I asked it today, in order of increasing difficulty (IMO):

1. Describe the importance of alkalinity and calcium in a coral reef aquarium.

ChatGPT reply:

Alkalinity and calcium are important in a coral reef aquarium because they play crucial roles in the growth and health of coral and other marine organisms. Alkalinity, measured in terms of carbonate hardness (KH), helps to maintain a stable pH level in the aquarium water, which is necessary for the survival of coral and other marine organisms. Calcium, on the other hand, is a vital component of coral skeletons and is necessary for their growth and maintenance. Without adequate levels of calcium, coral will not be able to build or maintain their skeletons, which can lead to slowed growth or even death. It's important for the aquarist to test and maintain proper levels of Alkalinity and Calcium to ensure a healthy and thriving coral reef ecosystem in the aquarium.

Randy rating: 60%. It fails to explain that alkalinity does anything other than stabilize pH, but does a decent job with calcium. Answered immediately.

2. Describe the effects of phosphate in a coral reef aquarium.

ChatGPT reply:

Phosphate is a nutrient that can be found in many types of aquarium water, including coral reef aquariums. In small amounts, phosphate can be beneficial to coral reef aquariums as it is used by many types of coral and other marine organisms as a growth substrate. However, in high concentrations, phosphate can lead to the excessive growth of unwanted algae and other organisms, which can outcompete and harm the coral and other desired organisms in the aquarium. High levels of phosphate can also lead to decreased water clarity, and can make it more difficult to maintain proper water chemistry. To maintain a healthy coral reef aquarium, it is important to monitor phosphate levels and keep them within a safe range. This can be done through regular water testing and by using phosphate-removing products, such as phosphate-adsorbing media or liquid phosphate-removing solutions.

Randy Rating; 80%. Very good, IMO. It might have just been traffic, but it thought about this one longer.

3. Why is strontium incorporated into coral skeletons?

ChatGPT reply:
Strontium is incorporated into coral skeletons because it is a chemical element that is present in seawater and can substitute for calcium in the formation of the coral's exoskeleton. Strontium isotopes can also be used as a tracer to study the coral's growth and environment.

Randy Rating: 95%. It might have added that strontium looks chemically like calcium .

4. Why does a conductivity measurement need to be corrected for temperature to allow determination of seawater salinity?

ChatGPT reply:
A conductivity measurement needs to be corrected for temperature to allow determination of seawater salinity because the conductivity of seawater changes with temperature. This is because the concentration of dissolved ions in seawater changes as the temperature changes, which affects the conductivity of the water. By correcting the conductivity measurement for temperature, the effects of temperature on the conductivity of the water can be eliminated, allowing for an accurate measurement of the salinity of the seawater.

Randy Rating: 5%, total fail. The conductivity of seawater does not change with temperature due to the number of ions changing, but rather to the change in the mobility of each individual ion.

5. Is a refractometer or a conductivity meter a better way to measure seawater salinity?

ChatGPT reply:
A refractometer is typically considered a more accurate way to measure seawater salinity than a conductivity meter. This is because a refractometer measures the actual salt content of a sample, while a conductivity meter measures the electrical conductivity of the water, which can be affected by other factors in addition to the salt content. However, conductivity meters are generally more convenient and easier to use in the field, and also cheaper. So it depends on your purpose and situation.

Randy Rating: 10%. Fail. The answer is partly opinion, which it cannot fail, but the description that a refractometer measures salt directly is ridiculous.

6. Is silicate a limiting nutrient for dinoflagellates in a coral reef aquarium?

ChatGPT reply:
Silicate is not typically considered a limiting nutrient for dinoflagellates in coral reef aquariums. Dinoflagellates, like most marine phytoplankton, require a balance of macronutrients (such as nitrate and phosphate) and micronutrients (such as iron and manganese) in order to grow and reproduce. While silicate can be present in seawater, it is not typically considered a limiting nutrient for dinoflagellates. However, silicate can be used as a nutrient source by diatoms, another type of phytoplankton, which can compete with dinoflagellates for resources and lead to changes in the community structure.

Randy Rating: 99%. Very good. I'd remove the word "typically" as it implies some scenarios where it may not be true, but otherwise an excellent answer.

I've been playing around with ChatGPT as well recently. I've run some of your example questions by it again, but prefacing my questions with "Using only peer reviewed sources, please answer the following question:"

It produces answers much more consistent with reality. I'd encourage you to give it a shot, it seems like it will indeed filter out the opinions from forums.

 

[brandon429]

I'm pretty amazed by its learning rate. I've been teaching it rip cleaning for two weeks now (ask it a question, redirect the answer with input + thread links) = I think it's catching on

it had literally no answer for rip cleaning at all in the context of a reef tank two weeks ago. bot/on

 

[brandon429]

look how it's learning about pico reefs, and we're part of the reference.

You can teach chat GPT by giving it commands to learn apparently. I have posted about fifty reef tank threads to it + pics and detailed testimony. I was wanting to see how it factors slanted input vs non slanted input, curious to see how it can be influenced in good or bad ways and if it retains the influence, or verifies the input in any way.