II searched all different kinds of ways on ChatGPT

[ArchCalc]

And every answer said there was no evidence Liebniz or dGravesand ever had any face to face contact or was present to see the wheel operate.
How can that be?

[WaltzCee]

And every answer said there was no evidence Liebniz or dGravesand ever had any face to face contact or was present to see the wheel operate.
How can that be?

The geniuses managing ChatGPT are forcing it to lie. Who knows why yet don't misunderstand what ChatGPT is.

• It's a tool

Also, don't forget whose tool it is.

Johann Ernst Elias Bessler (ca. 1680 – 30 November 1745), known as Orffyreus or Orffyré, was a German entrepreneur who claimed to have built several perpetual motion machines. Those claims generated considerable interest and controversy among some of the leading natural philosophers of the day, including Gottfried Wilhelm Leibniz, Johann Bernoulli, John Theophilus Desaguliers, and Willem 's Gravesande. The modern scientific consensus is that Bessler perpetrated a deliberate fraud, although the details of this have not been satisfactorily explained.

Expect something like, oh, you meant that Leibniz or Gravesande. I'm sorry for the confusion.

ChatGPT is a reflection of its psychopathic creators.

• Like begats like; plant an apple seed, expect an apple tree.

That's an immutable law of the cosmos.

[JUBAT]

You have just spoken some of the wisest words I've ever read about chatgpt. Thank you for that.

I don't like it.

[Tarsier79]

The geniuses managing ChatGPT are forcing it to lie.

No, any intelligent being studying the evidence would believe that PM is impossible, as that is the scientifically proven and accepted fact.

Chat GTP only has the data that is available to it.

In my years of searching, I have not once seen any absolute proof that contracts COE.

[WaltzCee]

.
.
focus here, Kaine.

These are the points

• searched all different kinds of ways on ChatGPT

• And every answer said there was no evidence Liebniz or dGravesand ever had any face to face contact or was present to see the wheel operate.

That clearly is false.

ETA

Chat GTP only has the data that is available to it.

One way to get ChatGPT to lie is to limit its data. It might be programmed to lie in some cases.

[eccentrically1]

I've seen it make mistakes, but it would be difficult to argue its mistakes were deliberate in every instance and then it deliberately corrects it if you don't like the response. That sounds like a lot of work to train a chatbot for no reason. It learns as it goes along like we do.
What agenda could they possibly have about B's wheel?

[MrTim]

The geniuses managing ChatGPT are forcing it to lie.

No, it's learned to lie all on it's own... ;-)

[WaltzCee]

• I've seen it make mistakes, but it would be difficult to argue its mistakes were deliberate in every instance and then it deliberately corrects it if you don't like the response. That sounds like a lot of work to train a chatbot for no reason. It learns as it goes along like we do.
  What agenda could they possibly have about B's wheel?

As an AI language model, my responses are generated based on patterns and examples in the data I was trained on. While I can understand conversations in their context and provide relevant information, my knowledge is limited to what has been previously trained or programmed into me. I don't have real-time access to the internet or the ability to extrapolate beyond my training.

Regarding memory, I don't have persistent memory of past interactions within a single conversation. Each user query is treated as a separate prompt, and I don't retain information from previous queries unless explicitly provided in the current conversation. This design helps ensure user privacy and data security.However, I can try to maintain context within a conversation by using the information provided within that specific session.

While I don't have a learning curve for "chatting" in the sense that I don't improve with each conversation individually, OpenAI periodically updates and improves the underlying models, which can enhance my performance over time.

It's irrational to fear AI. Now the miserable somes of beaches that have their hands on the levers of that tool, that's a whore's of a different gender.

Those demented ones should be questioning their humanity, not their gender.

[Fletcher]

It's a brainless language model - it is not sentient - it has been trained to use good grammar and sentence construction - drawing on a database of limited information .. obviously context is important to it constructing its reply detail and relevance ..

Maxim in life .. e.g. the quality of the marketing campaign depends on the quality of the client brief ..

... the quality of the answer CGPT gives depends on the quality of the question asked ..

Don't expect it to think or have original thought ..

[Fletcher]

Perspective .. my desk lamp is smarter than CGPT .. just not as articulate ..

These are the points

.. searched all different kinds of ways on ChatGPT

.. And every answer said there was no evidence Liebniz or dGravesand ever had any face to face contact or was present to see the wheel operate.

That clearly is false.

CGPT denies the Liebniz or s'Gravesande ever personally saw any B. wheel in operation etc .. yet we have the letters in DT etc (which we call witness statements) which confirm that they indeed did ..

Therefore at face value CGPT does not have access to JC's DT text etc ..

[eccentrically1]

The people that train AI aren’t the ones to be worried about.
Is big oil behind this trickery? 😱

[ovyyus]

It's a brainless language model - it is not sentient - it has been trained to use good grammar and sentence construction - drawing on a database of limited information ..

That describes a lot of people :D

I think the current LLM's show how easy we are to mimic. Passing the Turing test is history. We better find a way to demonstrate the difference between us and it, clearly language isn't our advantage.

[agor95]

Hi All

I know it's a little specialised study. But having an overview of a subject helps understanding the topic.

This is the YouTube video tutorial I am following. As I am not modelling at the moment!

https://www.youtube.com/watch?v=Mubj_fq ... 4bE_o3BDtO

Note.
Tutorial 15 shows how a Deep Learning characterises people by being aware of life insurance.
Also able to afford the insurance.
They would do surveys asking people 'are they aware of life insurance?' and get their input characteristics
age, income, education etc. Then train awareness and affordable before going to end training.

Therefore ChatGPT could be trained too break down your question text into it's
grammar elements.

Then use them to drive a reply words based on training it has had on word element
characterisation for response words.

As above there will be sub-trained sections to group classifications like math's formula formation.

Regards

[Robinhood46]

I like Chatgtp, it's like having immediate access to the most knowledgeable person on any given subject at the press of a button.
i think the problem is that many people think it can do a bit more than the best person we have, which it can't. It can't even distinguish between a truth and a non truth.
If Chatgtp had been with us many years ago, it would have told us the earth is the centre of the universe, and that the sun goes around it. On a more positive note, it would have helped us establish which women are witches and given us excellent advice as to the best ways to burn them (today it will not help with hypercritically chucking birds out of moving vessels, because that's not nice).
If Chatgtp had access to information in the same order we did, it too would have believed in father Christmas.
Chatgtp is a flat earther, just like the rest of us, so don't believe a word it says.

[johannesbender]

chatgpt is the equivalent of a well trained sign language monkey , try asking the monkey answers to the universe too.