Jump to content

SkySurveyBanner.jpg.21855908fce40597655603b6c9af720d.jpg

Chat GPT is this accurate advice


Rustyspurs

Recommended Posts

I don't know enough about those lenses to advise but I use Chat GPT a lot. It can give very good advice but it also has a very weak grasp of reality vs fantasy. It is quite likely to recommend lenses that don't even exist. It is also trained to make its responses sound plausible and helpful!

I have asked Chat GPT for Canon lens advice and the answer was not correct. It even told me that it personally had used one of the lenses!!! I have also used it for programming advice and so on and it is not accurate enough to rely on. I have found it most useful to assist in creative writing.

You are using the latest release of Chat GPT which has "enhanced factuality" - which i think is a geeky way of saying it is a bit less inclined to make stuff up.

  • Haha 2
Link to comment
Share on other sites

Asking an AI to recommend gear is a bit like asking me which is best,  Lamborghini or Ferrari. I could tell you either way, but since I've never even sat in one, what is my advice worth to you? Personally, I'd prefer advice from an actual owner, rather than from a story teller.

  • Like 3
Link to comment
Share on other sites

On 09/02/2023 at 01:08, nfotis said:

Never trust a story teller, only actual users.

ChatGPT is *not* a search engine, it assembles persuasive text from a large amount of existing text.

 

N.F.

 

There is an argument to be made that this existing text was written by users for starters and that therefore, the answer provided through machine learning is the most accurately representative of the human group opinion. I will let that here and let you ponder :)

Link to comment
Share on other sites

2 minutes ago, FrenchyArnaud said:

There is an argument to be made that this existing text was written by users for starters and that therefore, the answer provided through machine learning is the most accurately representative of the human group opinion. I will let that here and let you ponder :)

Yes but humans are experts at making up and posting mis/dis-information.

Trying to convince the world at large (inc 'experts') just 30 years ago that our star is not the only star with planets around it was an impossible task.

Link to comment
Share on other sites

21 minutes ago, EarthLife said:

Yes but humans are experts at making up and posting mis/dis-information.

Trying to convince the world at large (inc 'experts') just 30 years ago that our star is not the only star with planets around it was an impossible task.

I am not disputing this, but for the sake of discussion I need to point out that this is a self-defeating argument

- Ask real poeple, not a robot.

- but the robot asked poeple in the first place and repeats their common opinion.

- yes but poeple are misinformed therefore the bot repeats wrong information.

it automatically follows that "ask real poeple" is bad advise, one cannot have it both ways. My point is that at this stage of technology, the bot's advise is to be considered, surely not as the ultimate answer but at the very least equal or superior to the half baked opinion of Johnny on internet. It does not mean individuals have to agree with the bots answer to a specific question; only that the bot's answer are on average superior to Johnny's. Point in case: this particular bot just passed law school. Most internet users could not, ergo if i have a legal question i am better off asking the bot than appealing to internet and its "real" people.

 

Again, pointing it out just for the sake of friendly argument 🙂

Link to comment
Share on other sites

If you treat an AI as a search engine, then given enough time, the AI should come up with a text based on what an(other) AI has previously written. Something to ponder indeed.

As for the OP, why not ask the question that was put before the AI directly on this forum, rather than asking this forum if the AI output was accurate?

  • Like 2
Link to comment
Share on other sites

Which raises an interesting point:

Assume an alien invader bases their chance of success on the information we have been broadcasting about ourselves for 100 years.
They will be fed with increasingly more modern information as they get closer. Requiring constant updates to their strategies.

Will they have access to the Internet? Is any of it being routinely broadcast into space? Will they notice we have fledgling AI?
Will they then rush here to save us from AI itself? AI being a known existential threat to every growing technological civilization. :wink2:
Or will they let us play with fire? To see if we come up with a unique solution. One which might be used on other planets approaching the AI hurdle.

Link to comment
Share on other sites

On 10/02/2023 at 19:02, Rusted said:

Which raises an interesting point:

Assume an alien invader bases their chance of success on the information we have been broadcasting about ourselves for 100 years.
They will be fed with increasingly more modern information as they get closer. Requiring constant updates to their strategies.

Will they have access to the Internet? Is any of it being routinely broadcast into space? Will they notice we have fledgling AI?
Will they then rush here to save us from AI itself? AI being a known existential threat to every growing technological civilization. :wink2:
Or will they let us play with fire? To see if we come up with a unique solution. One which might be used on other planets approaching the AI hurdle.

I think they will come down to earth and grab all the popcorn they can lay their hands on. Then settle on the moon, lay back and watch it all unfold.

The question that presents itself now is of course if an AI can derail a thread on a forum as efficiently as a human can. That is the real Turing test.

 Now, let's attempt to put this discussion train back on track. The AI mentioned that one of the lenses is popular for low light level photography. But that is not the same as atrophotography. Low light level photography may very well be terrestrial photography, where a lens is evaluated completely differently from astrophotography. A lens may be called "soft", or have a good "bokeh". But if you ever hear that in AP, you have a lens that lacks contrast and is poorly focused. Hence, my original answer; don't rely on an AI, but try to get answers for specific lenses firsthand from owners.

Edited by wimvb
  • Haha 1
Link to comment
Share on other sites

57 minutes ago, wimvb said:

I think they will come down to earth and grab all the popcorn they can lay their hands on. Then settle on the moon, lay back and watch it all unfold.

The question that presents itself now is of course if an AI can derail a thread on a forum as efficiently as a human can. That is the real Turing test.

 Now, let's attempt to put this discussion train back on track. The AI mentioned that one of the lenses is popular for low light level photography. But that is not the same as atrophotography. Low light level photography may very well be terrestrial photography, where a lens is evaluated completely differently from astrophotography. A lens may be called "soft", or have a good "bokeh". But if you ever hear that in AP, you have a lens that lacks contrast and is poorly focused. Hence, my original answer; don't rely on an AI, but try to get answers for specific lenses firsthand from owners.

Alas, the average, Mk1 Human is a rather poor reviewer of anything much at all.
They have product loyalty from spending a fortune on something "unnecessary to survival." 
To admit their purchase was inferior would be an open admittance of failure, bias or simply poor judgement.
Then there are all the corrupt influencers who have received free goodies for review. YouTube, et al.

ChatGPT could harvest all the reviews, both positive and negative and average them. List Pros and Cons.
Provided it knew what to look for and had free access to the entire Internet.
Very early days yet. Who knows where this will lead?

Link to comment
Share on other sites

I think this technology is likely to be taken up more readily by the social media enthusiasts which could possibly result in them being better informed.  I do worry though that it could be manipulated to form biased agendas.  "Free thinkers" may well, at best be marginalised or at worst, sectioned.      🙂 

  • Like 1
Link to comment
Share on other sites

Just asked it why is astronomical sketching such fun.  Not sure if it copied it wholesale from somewhere else, but seems reasonable to me:

"Astronomical sketching is a unique and enjoyable activity that combines elements of art and science. When you sketch the night sky, you are capturing the beauty of the stars and celestial objects in a way that a photograph cannot replicate. Sketching by hand allows you to capture the delicate details of a star cluster, the faint wisps of a distant galaxy, or the intricate patterns of a nebula, and to add your own personal interpretation to the scene.

 

Additionally, astronomical sketching allows you to develop a deeper understanding of the objects you are observing. Through the process of sketching, you become more aware of the size, shape, and relative positions of celestial objects, and you begin to see patterns and relationships that you may not have noticed before.

 

Finally, astronomical sketching can be a peaceful and meditative activity, as you spend time outdoors under the night sky, observing the beauty of the universe around you. Whether you are sketching solo or with friends, it can be a wonderful opportunity to connect with nature and to slow down and appreciate the world around you.

 

Overall, astronomical sketching is fun because it combines the creativity of art with the wonder of the night sky, and it allows you to capture and interpret the beauty of the universe in your own unique way."

  • Like 1
Link to comment
Share on other sites

42 minutes ago, Peter Drew said:

I think this technology is likely to be taken up more readily by the social media enthusiasts which could possibly result in them being better informed.  I do worry though that it could be manipulated to form biased agendas.  "Free thinkers" may well, at best be marginalised or at worst, sectioned.      🙂 

The billions being invested in AI research suggest a strong commercial bias is highly likely.
Probably resulting in payment for AI services depending on usage.

Link to comment
Share on other sites

On 10/02/2023 at 09:40, FrenchyArnaud said:

I am not disputing this, but for the sake of discussion I need to point out that this is a self-defeating argument

- Ask real poeple, not a robot.

- but the robot asked poeple in the first place and repeats their common opinion.

- yes but poeple are misinformed therefore the bot repeats wrong information.

it automatically follows that "ask real poeple" is bad advise, one cannot have it both ways. My point is that at this stage of technology, the bot's advise is to be considered, surely not as the ultimate answer but at the very least equal or superior to the half baked opinion of Johnny on internet. It does not mean individuals have to agree with the bots answer to a specific question; only that the bot's answer are on average superior to Johnny's. Point in case: this particular bot just passed law school. Most internet users could not, ergo if i have a legal question i am better off asking the bot than appealing to internet and its "real" people.

 

Again, pointing it out just for the sake of friendly argument 🙂

One of the first things to consider when you ask a question is surely the competence of the person you are asking. This would hold true for an AI or person. At the end of the day, AIs are doing nothing different from people in that they are passing on received knowledge. AIs of course will ultimately have the edge in that they can accumulate a greater pool of knowledge than any individual. 

Jim 

  • Like 1
Link to comment
Share on other sites

2 hours ago, Rusted said:

The billions being invested in AI research suggest a strong commercial bias is highly likely.
Probably resulting in payment for AI services depending on usage.

Not sure, I think it will follow what we already have, the service will be stratified. Common search engines as we have now will morph to AI algorithms (we can see that happening already). These services will remain notionally free, open.  Companies developing specific AI technology for specific services (transport, aviation, medical) will imbed the service and include/apportion costs in the integrated product.  

Jim 

Edited by saac
  • Like 1
Link to comment
Share on other sites

4 hours ago, wimvb said:

Hence, my original answer; don't rely on an AI, but try to get answers for specific lenses firsthand from owners.

Now that you have put that advice out there in the public domain it becomes part of any AIs "reading lists" .  If your advice is replicated elsewhere by other sources then it will be given a higher ranking in reliability and more likely to be incorporated into future responses.  You are teaching them :) 

Jim 

  • Like 2
Link to comment
Share on other sites

It seemed odd [to me] that ChatGPT wanted me to register my inside leg measurement before allowing me to play.
Without drifting off topic [yet again?] this makes every registered user a close potential target for "something." :wink2:
Targeted ads? Hacking by AI? Filtering out "high flyers" for recruitment to the AI security service nobody has [yet] heard of. :wink2:

Imagine an AI which has direct access to your vital signs via the Internet of things. Fitness wearables, watches, phones, etc.
It could feed you live images/videos. Which rapidly and successively home in on your perfect car/telescope/face/motorcycle/person/whatever.
Using your own heart rate monitor, heat sensors, etc, to leave you completely paralyzed with "admiration" as your chemistry overdoses.

This isn't too far fetched already. Now that AI is already producing fake images and videos using "mixed text metaphors."  :wink2:
I note that these AI image modification services are presently using heavy filtering. With very good reason. IMO.
 

  • Like 1
Link to comment
Share on other sites

36 minutes ago, Rusted said:

It seemed odd [to me] that ChatGPT wanted me to register my inside leg measurement before allowing me to play.
Without drifting off topic [yet again?] this makes every registered user a close potential target for "something." :wink2:
Targeted ads? Hacking by AI? Filtering out "high flyers" for recruitment to the AI security service nobody has [yet] heard of. :wink2:

Imagine an AI which has direct access to your vital signs via the Internet of things. Fitness wearables, watches, phones, etc.
It could feed you live images/videos. Which rapidly and successively home in on your perfect car/telescope/face/motorcycle/person/whatever.
Using your own heart rate monitor, heat sensors, etc, to leave you completely paralyzed with "admiration" as your chemistry overdoses.

This isn't too far fetched already. Now that AI is already producing fake images and videos using "mixed text metaphors."  :wink2:
I note that these AI image modification services are presently using heavy filtering. With very good reason. IMO.
 

FLO could dispatch a TAK to me before I even know I need one :) 

Jim 

  • Haha 1
Link to comment
Share on other sites

4 hours ago, saac said:

One of the first things to consider when you ask a question is surely the competence of the person you are asking. This would hold true for an AI or person. At the end of the day, AIs are doing nothing different from people in that they are passing on received knowledge. AIs of course will ultimately have the edge in that they can accumulate a greater pool of knowledge than any individual. 

Jim 

You see, that is precisely where this sort of AI is absolutely superior to people. It basically stems from 2 basic factors: 

1) NOONE evaluates all those involved in a virtual conversation (for instance, did you really check I was neither a bot nor a cybernetics expert?)

and

2) the vast data collection of AI is way past any reasonable human effort and that makes it different in nature. Most poeple think the plural of anecdote is data - it is not. You know, it's the  "don't buy a Ford, they are s**t. I should know, I had 3. Never again!" sort of opinion. 

To me by default that makes the google/alexa/siri/chatgpt/you.name.it answers more valid than poeple's opinion, until such time one particular individual has shown to know/understand more than them. There is a handful of experts I totally trust on specific topics - more than the machine for sure! But these are better informed than non vetted expert (including known personalities as known does not equate competent)

In other words: AI ALREADY has the edge. And the gap will grow at the rate of universal inflation, we will never ever be able to catch up. 

Link to comment
Share on other sites

Just a question for information. Does an AI give you just one answer? Isn’t that a bit like always accepting the first answer in Google as the answer?  I imagine that when most people use a search engine they skim read down the returned search results and select the most relevant.   This suggests to me that AIs might be quite good at answering questions to which there is one answer, but not as good as a proper search engine into which questions are typed and returns read. 

Link to comment
Share on other sites

4 hours ago, FrenchyArnaud said:

You see, that is precisely where this sort of AI is absolutely superior to people. It basically stems from 2 basic factors: 

1) NOONE evaluates all those involved in a virtual conversation (for instance, did you really check I was neither a bot nor a cybernetics expert?)

and

2) the vast data collection of AI is way past any reasonable human effort and that makes it different in nature. Most poeple think the plural of anecdote is data - it is not. You know, it's the  "don't buy a Ford, they are s**t. I should know, I had 3. Never again!" sort of opinion. 

To me by default that makes the google/alexa/siri/chatgpt/you.name.it answers more valid than poeple's opinion, until such time one particular individual has shown to know/understand more than them. There is a handful of experts I totally trust on specific topics - more than the machine for sure! But these are better informed than non vetted expert (including known personalities as known does not equate competent)

In other words: AI ALREADY has the edge. And the gap will grow at the rate of universal inflation, we will never ever be able to catch up. 

Exactly, that is why they are developing it. I wouldn't go so far though as to say  that google/alexa/siri/chatgpt answers are more valid than people. You need to qualify "people".  Like I said before, more important than asking the question, is asking who you are asking the question of. No point asking me anything about football for example.  No point asking google/alexi/siri/chatgpt what I felt when I landed my first slide on my longboard :)  

Jim 

Edited by saac
  • Like 1
Link to comment
Share on other sites

22 hours ago, saac said:

I wouldn't go so far though as to say  that google/alexa/siri/chatgpt answers are more valid than people.

Chat GPT has the most information but can it use it? Just because I bought the biggest toolbox in the world doesn't mean I am the best at DIY! 

Link to comment
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now
  • Recently Browsing   0 members

    • No registered users viewing this page.
×
×
  • Create New...

Important Information

We have placed cookies on your device to help make this website better. You can adjust your cookie settings, otherwise we'll assume you're okay to continue. By using this site, you agree to our Terms of Use.