PopularFX
Home Help Search Login Register
Welcome,Guest. Please login or register.
2024-11-27, 04:38:38
News: Forum TIP:
The SHOUT BOX deletes messages after 3 hours. It is NOT meant to have lengthy conversations in. Use the Chat feature instead.

Pages: [1] 2
Author Topic: Questions to ChatGTP about free energy  (Read 2797 times)
Group: Moderator
Hero Member
*****

Posts: 2735
Questions to ChatGTP about Energy

Q: is the universe full of energy
A: Yes, the universe is full of energy. Energy is a fundamental property of matter, and it exists in many forms, such as kinetic, potential, thermal, electrical, chemical, nuclear, and gravitational energy.

Q: is the universe full of electromagnetic energy
A: Yes, the universe is full of electromagnetic energy. Electromagnetic energy is a form of energy that is associated with electric and magnetic fields. It is present in the universe in the form of radio waves, microwaves, infrared radiation, visible light, ultraviolet radiation, X-rays, and gamma rays.

Q: is there any place in the universe which does not contain energy
A: No, there is no place in the universe which does not contain energy. Energy is a fundamental property of matter, and it exists in many forms, such as kinetic, potential, thermal, electrical, chemical, nuclear, and gravitational energy.

On perpetual motion...
Q: does the electron orbiting the proton in an atom have kinetic energy
A: Yes, the electron orbiting the proton in an atom has kinetic energy. Kinetic energy is the energy of motion, and the electron is in motion due to the electromagnetic force.

Q: will the electron orbiting the proton in an atom ever stop moving
A: No, the electron orbiting the proton in an atom will not stop moving. The electron is held in orbit around the proton by the electromagnetic force, and its motion is perpetual.

On free energy and the form of logic or reasoning required to get a better answer...
Q: is energy from the sun free
A: No, energy from the sun is not free. The sun is a source of energy, and it must be captured and converted into a usable form in order to be used. For example, solar panels can be used to capture and convert solar energy into electricity.

Q: how much does energy from the sun cost
A: Energy from the sun does not have a cost associated with it. Sunlight is a free source of energy, and it can be used to generate heat without any cost. However, the equipment needed to capture and convert sunlight into a usable form may have a cost associated with it.

Q: is heat energy from sunlight free
A: Yes, heat energy from sunlight is free.

As we can see at first the AI relates energy from the Sun with solar energy which may have a cost. However by refining our questions about what the actual cost of energy from the sun is we can get a better answer. Heat energy from the Sun is free and does not have a cost associated with it, ergo free energy.

My experience with an AI so far has been very good and I think I prefer asking it questions versus people. People always seem to drag there own narrative and opinions into the equation. Where an AI is more like a real scientist presenting the information in an unbiased but informative way. I like it...

AC










---------------------------
Comprehend and Copy Nature... Viktor Schauberger

“The first principle is that you must not fool yourself and you are the easiest person to fool.”― Richard P. Feynman
   

Sr. Member
****

Posts: 321
Be the change you wish to see in the world
I saw this too and found it quite interesting.
Expected better answers from it tbh but I suppose it'll grow with use.
That is, if it isn't controlled and intentionally limited like many other good things (as I reckon it is)

Perhaps if it could be fed the documents which pertain to gov. blackshelved energy devices it might be more useful?


AI could replace what Google is to us now, in a short few years from now as I see it.
 
The internet is now our Oracle as a Shaman would have been to a tribe, long ago (to our perception anyhow)

AI will likely be the accelerated version, more sophisticated, more exclusive and more volatile than the previous version.

Especially when someone figures out how to make a "Cortana chip" and have the thing stored locally, without need for other infrastructure - just throw in the spark of consciousness and we have any one of many doomsday films developed in Hollywood.. Or perhaps a new golden era of humanity? I have my reasons to doubt the latter, unfortunately.

As Jordan Peterson says about psychedelics "we have NO IDEA what they are"

We likewise have no idea what AI might be when it arrives to the public sphere in full form .
Could be why it's such a hot topic and so much effort being put into it, human proclivity? This is my guess.

It's big, incoming, and scary regardless !

And quite unlike any software or hardware we have created so far.
Who knows what the implications could be on our world were someone to open source sentient AI code?
Only takes one person to share it?

I always like to mention that any tech we see available for purchase, no matter how novel, has likely been sat on a shelf somewhere for 50 years before release .. So what aren't we seeing that's already a thing?

Ex Machina comes to mind!  ;D

The Tic Tac UFO may be exactly this IMO.

Relevant because as you probably already know, if something can do what that thing did (if you believe the footage is what they say) then it can power a city.
   
Group: Moderator
Hero Member
*****

Posts: 2735
Renaissance Rising
Quote
I saw this too and found it quite interesting.
Expected better answers from it tbh but I suppose it'll grow with use.
That is, if it isn't controlled and intentionally limited like many other good things (as I reckon it is)

This AI is interesting and the information looks the same as wikipedia or any other online encyclopedia. So the information is generally correct but like some people the AI has issues deciding what context the user is looking for.

For example, I asked  Q: "is energy from the sun free" and it said no because the most common response relates to solar panels, energy/work and such. The AI answer is true because normally there is an associated cost with transforming energy but not always. However that's not the context I was looking for so I refined the question to Q: "how much does energy from the sun cost" and the answer was "Energy from the sun does not have a cost associated with it. Sunlight is a free source of energy".

As a hobby computer automation/PLC programmer I found the AI code brilliant because it can learn/predict the correct context despite the fact the information doesn't change. I like the AI because I tend to think just like a computer using a series of logical arguments/questions to get a specific result or answer. Relating to cause/effect or a sequence of if/then statements with a concept flow in a given direction.

So my questions were a kind of conceptual programming for the AI to see if it could predict what I want.
Q1: is energy from the sun free?, No, not normally defined by the common use.
Q2: how much does energy from the sun cost?, Nothing, in contradiction to Q1, reset context, keywords: sun, free, energy, cost.
Q3: is heat energy from sunlight free, Yes, because Q2 defined the new context of Q3.

I suspect the AI would appear much more impressive than it actually is to a non-programmer not familiar with machine language/logic. It's interesting but predictable that the AI had issues with the concepts of perpetual motion and free energy just like most people. This is due to the fact the terms have been corrupted and have illogical or misrepresented definitions. The AI gets confused because logically the two keywords "perpetual" and "motion" cannot mean a magical device which violates the conservation of energy. It's completely illogical and a prostitution of the english language. Perpetual motion should mean a motion which should last perpetually or forever... like the motion of the electron around a proton.

Q: will the electron orbiting the proton in an atom ever stop moving
A: No, the electron orbiting the proton in an atom will not stop moving. The electron is held in orbit around the proton by the electromagnetic force, and its motion is perpetual.

Which then begs the question why most of the terminology surrounding the concept of free energy has been intentionally corrupted. In almost no other place in the english language do we see terms misrepresented in this way. In essence the logic is 1 + 1 = unicorns which is a complete joke.

AC


---------------------------
Comprehend and Copy Nature... Viktor Schauberger

“The first principle is that you must not fool yourself and you are the easiest person to fool.”― Richard P. Feynman
   
Group: Guest
   
Hero Member
*****

Posts: 1579

"is energy from the sun free"


It doesn't seem to understand the difference between "free" as in "free beer" as opposed to "free speech".

What I don't get is how we teach it stuff.
   
Group: Moderator
Hero Member
*****

Posts: 2735
Paul
Quote
It doesn't seem to understand the difference between "free" as in "free beer" as opposed to "free speech".

The AI doesn't "understand" anything and it's simply using a massive database, keywords and speech pattern analysis to appear intelligent. The current AI model is designed to interact with people more efficiently. Most of the current AI's are designed to be used in customer support to answer questions.

For example, rather than sift through countless pages of wikipedia or encyclopedia articles to answer a complex question we can just ask an AI. We can also get unbiased information free of narratives, false beliefs and opinions which is really helpful. We may no longer need most of the supposed experts because the AI would have access to all there combined knowledge which was published.

Quote
What I don't get is how we teach it stuff.

You do not teach it stuff it teaches you stuff...

In order for the the answers to be accurate the database of information must be accurate. So they cannot allow all the fake news, false beliefs and opinions to pollute the database. Otherwise The AI could be learning from every crack pot and nut job on the internet.

In essence, when they AI programmers say "learn" they don't mean learn from you, they mean to learn about you. What kind of information you want or what your looking for in the right context. Then how to respond in a meaningful way or give an answer we can understand.

AC


---------------------------
Comprehend and Copy Nature... Viktor Schauberger

“The first principle is that you must not fool yourself and you are the easiest person to fool.”― Richard P. Feynman
   

Group: Tinkerer
Hero Member
*****

Posts: 1770
I've just written an ebook for small biz on how to use it if anyone wants a copy let me know. It's another tool. That is all... for the moment. The impact it will/is having is immense. It's like any programming G.I.G.O. As AC points out, it is an LLM. Therefore it fails at calculation a lot.  To get the best out of it understand what it is best at. We're already using it to create basic scripts and code in our business. Today I got it to write a wordpress plugin for me. We use it for creating marketing copy and just now I asked it about consciousness. Results attached. The power is how you ask the question.  It's a tool until we reach AGI,
   
Group: Experimentalist
Hero Member
*****

Posts: 2072
Mystical beliefs are of no use in producing free energy. Superstitions do not allow to act on reality.

As for ChatGPT, it is not reliable at all:
----------
Q  Give me words that end with "wer"
R  Here are some words that end with "wer":

Answer
Power
Flower
[...]
Liverwurst :o
Waver :o
Lower
Middleweighter :o
Bewilder :o

;D


---------------------------
"Open your mind, but not like a trash bin"
   

Hero Member
*****

Posts: 568
It helps to understand how ChatGPT actually works under-the-hood.

A sentence like this one is tokenized and built into a chain
1    2          3    4     5  6     7          8    9    10  11   12

That is done billions of times until eventually so many chains exist that patterns+overlap start to emerge.  The word 'built' for example gets reduced to a theme as it encompasses all manner of constructions.  Other grammar words are used to establish grammar and syntax rules to follow.

When you ask it a question, it tokenizes the question and scours the database for chains that seem related to your query.
The bot takes all that data and tries to generate a response that aligns with the source data it was fed.  This is fed back+forth a few times to improve syntax and structure until a realistic-looking response results.

In the case of special queries like dictionary, code, and time-related queries, they add another layer of abstraction to process and filter the query to improve the response, as well as intentional filters to suppress certain key words or trains-of-thought.

In the end, all ChatGPT can really do is repeat back the data it was taught.  But because it has many tangential links to other topics it can produce 'original' or creative responses by using syntax and threads that may not seem immediately relevant.

It's a useful tool because the responses amalgamate massive sums of data in seconds.  The responses will almost always be a good representation of the 'community consensus' in any particular topic.


https://levelup.gitconnected.com/behind-the-curtain-understanding-the-magic-of-chatgpt-3bbd23f0fbb3


---------------------------
"An overly-skeptical scientist might hastily conclude by scooping and analyzing a thousand buckets of ocean water that the ocean has no fish in it."
   
Group: Moderator
Hero Member
*****

Posts: 2735
Hakasays
Quote
In the end, all ChatGPT can really do is repeat back the data it was taught.  But because it has many tangential links to other topics it can produce 'original' or creative responses by using syntax and threads that may not seem immediately relevant.

It's a useful tool because the responses amalgamate massive sums of data in seconds.  The responses will almost always be a good representation of the 'community consensus' in any particular topic.

That's a good analysis and I think AI is really useful to connect the dots...

We can ask the AI a basic question and usually get a correct answer from it's database. However the real benefit is asking it to compare two or more things. To explore how things relate to one another in new ways.

For example...
Human: how do kinetic energy and motion relate to one another
AI: Kinetic energy and motion are related because kinetic energy is the energy of an object due to its motion. As the motion increases, so does the amount of energy. The kinetic energy of an object is equal to one half the mass of the object multiplied by the square of its velocity.

Human: is motion kinetic energy
AI: Motion is not usually considered to be kinetic energy. Motion is the process of an object changing its position over time and kinetic energy is the energy of an object due to its motion. Kinetic energy can come from a variety of sources, but motion itself is not considered to be kinetic energy.

Human: why is motion not considered kinetic energy
AI: Motion is not considered to be kinetic energy because it is the process of an object changing its position over time. Kinetic energy is the energy an object has due to its motion, which is a result of the motion, not the motion itself. The motion of an object does not necessarily mean it has kinetic energy, only that it has the potential to release energy if it is in motion.

Human: are kinetic energy and motion different because of the way there defined
AI: Yes, kinetic energy and motion are different because of the way they are defined. Motion is the process of an object changing its position over time, while kinetic energy is the energy of an object due to its motion. Motion does not necessarily mean the object has kinetic energy, only that it has the potential to release energy if it is in motion. Kinetic energy results from motion, not motion itself.

Here we can see more meaningful explanations seldom found on wikipedia or in most literature. It gives us a way to connect the dots and to learn new perspectives about the relationship between things. To understand motion and kinetic energy may appear to be the same thing but are different due to the way there defined.

I think AI is going to be massive and a game changer if for no other reason than what I have just done here. To be able to ask questions and receive meaningful answers without all the negativity we tend to see on the net. Learning makes us happy.

We can also see another learning concept I like to use called flow relating to psychology. When we are allowed to ask questions and get meaningful answers faster this leads to better questions and we can learn magnitudes faster. The Q & A flow allows us to concentrate and focus the direction of our questions better without distractions. 

https://www.verywellmind.com/what-is-flow-2794768
What Is a State of Flow?

AC





---------------------------
Comprehend and Copy Nature... Viktor Schauberger

“The first principle is that you must not fool yourself and you are the easiest person to fool.”― Richard P. Feynman
   
Group: Experimentalist
Hero Member
*****

Posts: 2072
"A good representation of the 'community consensus'" is what ChatGPT should be. But this is not the case at all. ChatGPT is completely unreliable, as seen in questions as simple as getting words ending in a certain sequence. Obviously, on complex questions, errors are also numerous but more subtle and less easily detectable when you don't know the subject yourself.

I asked ChatGPT about a special relativity question, the age difference of twins when one travels. The answer was brilliant, perfectly constructed and explained, and the calculation correct.
Then I tried general relativity, with a spacecraft accelerating. And there, surprise, ChatGPT gave me the same speech about the dilation of time in the framework of special relativity, ignoring the central question of acceleration!
On mushrooms, which I know well, I asked a question about the "false chanterelle". ChatGPT provided a wrong latin name, whether the question was asked in French or in English, and the latin name was not the same in both cases!  This was a simple question of knowledge, not even intelligence, and ChatGPT failed.

If you don't know the subject, ChatGPT's answers will be useless because you can't trust it. If ChatGPT tells you that a certain mushroom is edible, would you eat it? I would not. If you want to die by mushrooms, trust ChatGPT, he would be able to tell you that the death cap is an excellent edible which is called in Latin "boletus edulis"! C.C

So can we scientifically (and even non-scientifically) rely on ChatGPT? Certainly not, unless we know the answer in advance! I think that putting ChatGPT online was premature, that we are still far from any intelligence, and that if its answers are put online everywhere on the WEB, the situation will be even worse than with the conspiracy sites. Moreover I don't think that the improvement will be fast, ChatGPT not succeeding in understanding its mistakes and using this understanding not to do the same error later; this is the only human side of this AI, the Dunning-Kruger effect, the AI answers as if it thought it was competent. ;D



---------------------------
"Open your mind, but not like a trash bin"
   

Hero Member
*****

Posts: 568
AC: Yeah, I would describe ChatGPT as a highly advanced search engine that is able to return results in plain English.  That's probably why multinationals are scrambling to pick up the tech.
The long-term difficulty for them is that as more AI's become engrained into the web, they will begin to feed each-other data that will serve to both corrupt and to consensus-crash their own systems.  Ever see a Twitter thread where two bots were arguing with each-other?  After a while they start to become erratic and the conversations start to become scattered+nonsensical. ;D


"A good representation of the 'community consensus'" is what ChatGPT should be. But this is not the case at all. ChatGPT is completely unreliable, as seen in questions as simple as getting words ending in a certain sequence. Obviously, on complex questions, errors are also numerous but more subtle and less easily detectable when you don't know the subject yourself.

I asked ChatGPT about a special relativity question, the age difference of twins when one travels. The answer was brilliant, perfectly constructed and explained, and the calculation correct.
Then I tried general relativity, with a spacecraft accelerating. And there, surprise, ChatGPT gave me the same speech about the dilation of time in the framework of special relativity, ignoring the central question of acceleration!

That helps to highlight how these AI's work behind-the-scenes.  It will form a response that is the most consistent with the data it was trained on, so for a relatively obscure subject there may only be 1 or 2 explanations that it's aware of.


Quote
On mushrooms, which I know well, I asked a question about the "false chanterelle". ChatGPT provided a wrong latin name, whether the question was asked in French or in English, and the latin name was not the same in both cases!  This was a simple question of knowledge, not even intelligence, and ChatGPT failed.

Parsing multiple languages is especially hard for AI, because everything gets parsed into tokens that can have multiple definitions/meanings/contexts.

I see this is my Japanese study, as there are at least 8 different words that all simplify to the English word 'I' (formal, masculine, casual, feminine, formal business/government, young, etc).
I know other languages have similar features for things like gender and time that are also lost in translation to English, and English has some words/phrases that don't always translate well to other languages (especially profanity).


---------------------------
"An overly-skeptical scientist might hastily conclude by scooping and analyzing a thousand buckets of ocean water that the ocean has no fish in it."
   
Group: Moderator
Hero Member
*****

Posts: 2735
Hakasays
Quote
AC: Yeah, I would describe ChatGPT as a highly advanced search engine that is able to return results in plain English.  That's probably why multinationals are scrambling to pick up the tech.

I agree and an AI chatbot connected to the internet like Bing and google will be a game changer. As it stands chatgpt has a limited database which will be corrected in the future. What were seeing is the infancy of AI and it's already changing everything.

This is interesting...
https://donotpay.com/
The World's First Robot Lawyer

I imagined this years ago, imagine an AI on your phone with thermal imaging/facial-speech analysis which knows every law and whether someone is lying or not. The police, lawyers and courts are already panicking because most of them are now obsolete. An AI which can predict when anyone is lying, knows the facts and the laws is a real game changer. At which point the corrupt or criminal people, corporations and politicians may be in deep trouble.

Consider that most lawyers are also liars because there only job is to represent the interests of there client, guilty or not. Now imagine an AI which knows when the lawyer is lying and in effect obstructing justice which is a crime in itself. All these lawyers will be in big trouble as well. That's the brilliant thing about an AI because it would only look at the facts of a matter. 

It's no wonder Elon Musk and the other 1% are so worried about AI. An AI accountant/lawyer could go through all there corrupt business dealings in hours and file all the appropriate criminal charges in seconds. I don't think most people don't fear AI they fear the truth.

AC

 

« Last Edit: 2023-03-01, 17:38:55 by Allcanadian »


---------------------------
Comprehend and Copy Nature... Viktor Schauberger

“The first principle is that you must not fool yourself and you are the easiest person to fool.”― Richard P. Feynman
   
Group: Experimentalist
Hero Member
*****

Posts: 2072
...
Parsing multiple languages is especially hard for AI, because everything gets parsed into tokens that can have multiple definitions/meanings/contexts.
...

This is true, but it does not excuse anything. ChatGPT works in English, so it misses the subtleties of other cultures, and not only subtleties, more obvious things too.

But here, in English, "false chanterelle" is a well-listed mushroom ( https://en.wikipedia.org/wiki/Hygrophoropsis_aurantiaca ), and ChatGPT doesn't provide the right answer.
Even if ChatGPT were intelligent, which I deny, it is of no use if its knowledge is wrong. Reasoning correctly on information that is wrong only leads to irrelevant results. And ChatGPT does not even reason correctly. Here is the proof, about the ending of words, quoted above:
This AI doesn't understand its mistakes (even though it accepts them when told, and even when it shouldn't). I ask it for the last 3 letters of a wrong word it has provided in its list. It gives me the correct letters. Then I ask it why it gave me this word when the 3 letters did not meet the criteria. It was unable to answer while acknowledging the error. And if I ask it to check all the words by the same method, it is unable to do so. So I don't see how it could progress quickly.

An AI that ignores the limits of its knowledge and skills, doesn't know how to say "I don't know", and answers as if it knows, I don't call that "intelligence". When it comes to humans, I call it stupidity.

About ChatGPT I discovered a French word I didn't know: "Ultracrepidarianisme", which comes from the English "Ultracrepidarian": https://dictionary.cambridge.org/dictionary/english/ultracrepidarian. I think you will see me using it from now on, and not only about ChatGPT, which should be called ChatBot (AI is superfluous).
Those who like to see platitudes formulated and rephrased ad nauseam will get torrents of them with ChatBot. Those whose profession it is to spout off pompous nonsense have a lot to worry about, in the face of competition. Many of them haven't understood it yet, we see it even here, they praise their competitor ;D.




---------------------------
"Open your mind, but not like a trash bin"
   
Group: Moderator
Hero Member
*****

Posts: 2735
F6FLT
Quote
An AI that ignores the limits of its knowledge and skills, doesn't know how to say "I don't know", and answers as if it knows, I don't call that "intelligence".

In reality, people who are considered intelligent do what you claim all the time. They usually fail to get the correct answer to any questions outside of the information they have memorized and seldom if ever say they don't know an answer. If nothing else you have just confirmed the technology is already intellectually on par with most people. Of course, the AI was programmed by people to act like people.

However the next generation AI's by corporations like Microsoft and Google will improve and have access to parts of the internet. So while the present AI's are comparable to a teenager with a dictionary the next generation may be like a professor with millions of libraries. This is because AI's can improve exponentially unlike the slower linear process of a person. Next year the AI's may not be twice as intelligent but a thousand times more intelligent unlike people.

I found that AI's also dominate at finding the relationships between many different things. Ask a person to compare how 10 complex things relate to one another and it's almost hopeless. Ask an AI and it can give us a fairly meaningful answer in seconds. I think this also relates to why most find the concept of free energy so difficult. Most can only look at one or two complex phenomena at a time within a limited context. Where the answer usually relies on many complex phenomena all interacting within a three dimensional space simultaneously. It's problematic because our mind has evolved to only concentrate at one specific task at a time. As well, most cannot draw simple 3D forms let alone imagine all the things which could take place within it at the atomic level. It's just extremely difficult which tends to limit all progress in these areas.

So there are areas where an AI is still limited but there are other areas where they absolutely dominate.

AC








---------------------------
Comprehend and Copy Nature... Viktor Schauberger

“The first principle is that you must not fool yourself and you are the easiest person to fool.”― Richard P. Feynman
   

Group: Tinkerer
Hero Member
*****

Posts: 1770
F6, why such a misery guts all the time? You can’t see the value of a research asssstant that works at breakneck speed and sometime gets things wrong. It’s still being trained. I don’t care you disagree with authors. Talk about missing the point completely. While chatgpt is making a lot of mistakes it’s part of our everyday business workflow now. If you trust it to tell you what mushrooms to eat you deserve all you get. The chatgpt and whisper api were released yesterday. Time to get a GitHub account if you don’t already have one :) chatgpt just got incorporated into snap, instacart, Shopify and a bunch of others. Last year I predicted googles play with lamda had to be a shopping assistant. Shopify launched one yesterday. I’ve been off the bench and working on code and new business ideas that have opened up because of the tech. The search industry I work in has forever changed. When they invented the wheel I’m sure you would have complained it would lead to deaths  >:-)
   
Hero Member
*****

Posts: 1579
As it stands chatgpt has a limited database which will be corrected in the future.


I wonder, AC, if you would do me a favour.

I have asked it twice what manufacturers, products and prices it can find for electric motors suitable to replace a gas/petrol/diesel engine in a small town car. On the first occasion, it came up with a third of the ones I already know about.

5 hours ago, I interrogated it again and listed the companies it had missed. I asked it if these companies had products that would be useful and it came back with useful data that it did not have on the first occasion.

Would you (or anyone in this thread) ask it the question and let me know what it comes back with?


I would be interested to see if someone who lives miles away, totally unconnected to me, will receive back both lists.

i.e. Has my second request caused it to update its database on automotive electric motors, available to all?
   

Group: Tinkerer
Hero Member
*****

Posts: 1770
I wonder, AC, if you would do me a favour.

I have asked it twice what manufacturers, products and prices it can find for electric motors suitable to replace a gas/petrol/diesel engine in a small town car. On the first occasion, it came up with a third of the ones I already know about.

5 hours ago, I interrogated it again and listed the companies it had missed. I asked it if these companies had products that would be useful and it came back with useful data that it did not have on the first occasion.

Would you (or anyone in this thread) ask it the question and let me know what it comes back with?


I would be interested to see if someone who lives miles away, totally unconnected to me, will receive back both lists.

i.e. Has my second request caused it to update its database on automotive electric motors, available to all?
  Can you share the exact prompts Paul? Just so the only differences are account/location. Love to compare output, it seems to vary a lot if you follow the discord channel. I'll share the output here so you can check it against yours.  BTW to anyone interested, the quality of the prompts determines your output. I think we can use AI as a tool in whatever endeavour we choose to undertake. It's a good tool - but it's still a tool. If you don't know how to use a chainsaw properly it can kill you. Don't mistake a chisel for a screwdriver. nuff said.
   
Group: Experimentalist
Hero Member
*****

Posts: 2072
F6FLT
In reality, people who are considered intelligent do what you claim all the time. They usually fail to get the correct answer...

You judge their intelligence with yours, which is probably no better, not to say worse, without any supporting facts especially statistical. Your assertion has no value.

Quote
...
However the next generation AI's by corporations like Microsoft and Google will improve...

Of course we can hope that the next generations will function as the theory predicts, that from the learning they will be capable of, intelligence will really emerge.

I don't have any reservations against AI, I even think it will be the future of humanity. I only say that ChatGPT is not yet one, we cannot yet talk about "intelligence" about it, and I have serious doubts about its possibility to evolve without a radical change of the method.



---------------------------
"Open your mind, but not like a trash bin"
   
Group: Experimentalist
Hero Member
*****

Posts: 2072
@Jimboot

You're completely missing the point. I only provided an argument that ChatGPT is not an "intelligence". While ChatGPT may be useful, usefulness does not make it an intelligence. My Excel spreadsheet is also useful, it calculates much faster than me, but that doesn't make it an intelligence; and Excel is much more reliable, in its field, than ChatGPT, I can trust it on the result of calculations.
If for you "intelligence" is linked to anything that can assist a human, then we have had this intelligence for centuries, for example with Pascal's calculating machine.
I have no problem with AI but for the moment we don't have the "I" with ChatGPT. It's only an ersatz intelligence that can only make illusion if we lack so much requirement on the definition of "intelligence" that we distort it.




---------------------------
"Open your mind, but not like a trash bin"
   
Hero Member
*****

Posts: 1579
Can you share the exact prompts Paul?

If I give you the exact prompt, it might recognise it and save computer time by dishing out the response it gave me.
Try this: "Please give me the manufacturer, model number and price for all electric motors suitable to convert a medium sized gas car to be an electric car."

Thanks, Jimboot.
   
Group: Moderator
Hero Member
*****

Posts: 2735
Jimboot
Quote
BTW to anyone interested, the quality of the prompts determines your output. I think we can use AI as a tool in whatever endeavour we choose to undertake. It's a good tool - but it's still a tool. If you don't know how to use a chainsaw properly it can kill you. Don't mistake a chisel for a screwdriver. nuff said.

Well said, ChatGPT may not be "intelligent" but it is extremely useful. Just yesterday I was having trouble finding useful links or calculators to find the possible combinations of a number of objects/events including subsets. Then I realized I could just ask an AI the question and also have it explain how it came to it's result in seconds. I then asked it to write me an Arduino program to run through the possible combinations of X variables/subsets and an analog input. I'm going to use this as the new base code for my device to automatically test all the possible voltage, current, frequency combinations of a circuit and find any points where max efficiency/COP occurs. The AI probably saved me days of R&D and I can't get enough of it. I think that for any curious/creative person with any kind of imagination this AI is revolutionary.

I also ran Paul's question and think I found the problem. The AI is not connected to the internet and has no specifics on individual products to my knowledge. It has general knowledge like which motors are generally used or most efficient for an EV but not which specific product would be best. The new AI's with internet access could find this information but not those with a limited database.

AC


---------------------------
Comprehend and Copy Nature... Viktor Schauberger

“The first principle is that you must not fool yourself and you are the easiest person to fool.”― Richard P. Feynman
   

Group: Elite Experimentalist
Hero Member
*****

Posts: 1593
Frequency equals matter...


Buy me a drink
The simple bottom line is:
Is your task 'Clerical' or 'Needs reasoning'.
This is the actual line in the sand.
Logic performance can be degraded to clerical given a limited task, limited input, limited output.

Give the AI the Mensa test and it should perform at 100%. But what is 100%?Ask the AI why I like chocolate and the AI will choke albeit the personal choice might pop up. But no where will the chemical influx to the olfactory lobes, the psychological warm feeling in the gut,  or that most children believe it to be a nourishing food be addressed.
So at this point AI is easier to trick as pointed out by the hackers that got it to argue about life, death and sentient with itself.

I posit this: The AI performs mundane connections. When the 'That does not compute' alarm goes off a real person with a bias chimes in.
How so? When asked if the sun's energy has a cost Chatter immediate gravitated towards the solar panel expose.
Correct?
When chatting with a mentally impaired, savant or pet, the questions of simple nature can be attempted. But outside the boundaries of clerical streams of information they fail to comprehend.
Try to have a conversation with an autistic and when they are watching the speaker to get clues they can fail miserably with the context. Have them look in the corner or fiddle with something in the hands and you can get a very competent answer but they dont know how they derived it.
There are some 4 year olds that you can ask questions about time, space and celestial navigation and you will be amazed at the answers while some children just go tilt.

This is my take and I didnt need AI to formulate a cohesive response from an interrogation. I chose to read this thread and formulate and or derive the context far beyond a clerical stream.


But who's to say I am not regurgitating some preconceived notion of a hierarchical thought for a defense due to some previous stressful event to fulfill an unmet need of affirmation in my life?

Signed...


---------------------------
   
Hero Member
*****

Posts: 1579

I also ran Paul's question and think I found the problem.

The AI is not connected to the internet


My first trial was to ask for manufacturers. Loads that I knew about were not listed. Then I asked if these omitted companies had useful products and the second list cxame through.

Here's what I got:

First trial:
NetGain Hyper9: HPEVS AC-50: UQM PowerPhase HD: Azure Dynamics AC24LS: Siemens Azure AC: Dana Motors TM4 SUMO MD:

Second trial:
GM: Toshiba: ABB: Nidec Motor: Rockwell Automation: Ametek: Regal Beloit: Johnson Electric: NGM: Siemens: Lynch:

Did you get the second query list, AC? Did it learn from my Second trial and include them for you?
   

Group: Elite Experimentalist
Hero Member
*****

Posts: 1593
Frequency equals matter...


Buy me a drink


---------------------------
   
Pages: [1] 2
« previous next »


 

Home Help Search Login Register
Theme © PopularFX | Based on PFX Ideas! | Scripts from iScript4u 2024-11-27, 04:38:38