Can You Really Trust The Information Given By Chatgpt?

Just out of curiosity, I tried to test ChatGPT what information it really knows and the results were not really satisfactory.

Many times that I have tried, I usually only get the answer:

* ********* *** *** *********, *** as ** ** ******** *****, ** responses *** ********* ***** ** * mixture ** ******** ****, **** ******* by ***** ********, *** ******** ********* data. * ***'* **** ****** ****** to *********** ********* ** *** ******* to ****** *** ********.

**, * ***** **** * ***** simple ********:

IPVM Image

* **** ** *** ********** ** 2016, ***** ** **** ****:

IPVM Image

****, *** *********** *** ******** *****, to *** **** ****** ******* **** give:

IPVM Image

* ***** ** ****** ** **** careful ** ******* ***** ** **** ChatGPT, **** ** *** *****?

Reactions:

(1)

John Honovich

IPVM

*** ** **** ************* *********! *********.

** *** *** ****, **'* ********** amazing *** **** ******* **** ****. On *** ***** ****, *****'* ** way ******* ** ***** **** *** digest ** ****** *** **** ****** of ********* ** *** ***** ***** all ***** ** ***** ***********.

Reactions:

(1)

bashis mcw

In reply to John Honovich

* ********* ***, *** *** ******* are ******** ****** ************.

*******, * **** ** ******** * pseudo * **** ** ********* ** python, *** ** *** **** ***** well, *** **** ****** ** **** the ** *** ******* ** **** the *********** **** ********.

** **** *******, ** ********, ** what * ***** "**** *******" ******** like ******* ***** ** ** ** blindly ***** *** ******* ******* ******** the *******.

Reactions:

Undisclosed #1

** **** *******, ** ********, ** what * ***** "**** *******" ******** like ******* ***** ** ** ** blindly ***** *** ******* ******* ******** the *******.

***** * ********* ***** **** ******* at *** ******, * ***** ******* that *** ********* ******* **** ****** currently ***** (***** ** ***** *** viewpoint/exposure/research) ** **** *** ****** (** more *** **** **** ** ********) will***** ******* **** ******** **(******* **** ** *** ********* ** GPT) ** **** ****** *** ******** themselves ** **** - ***** ** their *** ********** ****** ******(*).

******* ********* ****** *** ** ****** on *** ****** ************ ** ****** nor ******** (*****, ** ****** **** pointed ***, ********* ***** ***** ** to *** ******* **** ******* ** relies **).

*** **** **** **** **** - and *** ** ***** ** *** bottom ** *** ******* ****:

**** ******** *******. ******* *** ******* inaccurate *********** ***** ******, ******, ** facts.******* *** ** *******

Reactions:

(1)

Undisclosed #1

In reply to Undisclosed #1

IPVM Image

Reactions:

(1)

bashis mcw

In reply to Undisclosed #1

* ***** ** ** ********* ** question ****** *******, ********** ** ** from * **** ***** ********* ** GhatGPT. * ****, *** **** ***** has *** *** ******* **** **-******* by ***, **** ******* *** *** the **** *** **** ********?

* **** ******** *** ****** ** mine, **** * ********** *** **** one ***** * *** ** ******** transfer ** **** ***** *** (********* Shift ******), **** ** **** **** "you ****** ** ****!" *** * asked ***, ** ***** ** ******* "you ****** ******** **** ** ***** square ****, *** **** ** *** sinus ****". ***** * ***** ********** at *** ****, ** * ****** him **** ** ** ********* ******** to ******** **** **** ****** **** modulating ** (********* **********) ** ******** data, ***** ********* ** ***.

**** * ****, *** *** *********, whatever ** ** ******, ******* ** a ******, *** ** *** *** question ***** ******* ****** ******* **** on ****.

Reactions:

(3)

(2)

Undisclosed Integrator #2

** ********** ** **** ******* ** most ****** *** ********** - **** it *********** *** ** *** ****** out **** ** ** **** **. For *******, *********** *** ***********.

**** ** ** ***** ** ******** information, ** ***** ********** ** ** effort ** ****** **** *** ****'* request. ** ***'** ******* * **** or ***** *****, ****'* *** * problem. ** ***'** ****** ** ***** a ******* **** ******* ** ***** a ***** ********, *** ****** ***** it *****. * **** ******** **** many ****** ****** **** *** **** version, *** **** ***** ** *** way **** *** *****, * ***'* think ********* *** ** ******* *******.

** ***** ** ******, ******* *** a ******* (*** ***** ********* *****) amount ** ***********, **** * ****** of **** *** *********** *** ********. It *****'* **** ************. **** **** somewhat ********** **** ** ***** ******** searches, *** **** ******* *** ****. SEO ** * **** ***, ***** used **** **** **********.

**** * ***** "**** *******" ******** like ******* ***** **

***** *** ********* * *** ** people *** **** ************* **** ******* can *** ***'* **. ******* ** not * ********* ****** ** *****. The **** *** ** **** *** Wikipedia *** ******. **** ** **** to **** ** **** **************? ***'* know *** **** ***. ** *** is **** **** **** ********* *** Google, **** ****** **** ***** *** things **** *** **** *** **** people **** ***** ****** **** *** false. ***** ***** **** ******** ****** out **** ** **** ******, ***** inexperienced ***** **** ****** ****** ** all ** ****** ***** ** ***** it ********.

Reactions:

(2)

bashis mcw

In reply to Undisclosed Integrator #2

******* *****, ** ***** **** ** to ************** ** ** ******* ***** the *******, ******** ** ***** ** Google, *******, ********. * ***** ** need ** *********** *** *** ******** and ******** ** ****.

Reactions:

Undisclosed Integrator #2

In reply to bashis mcw

*************, **** ** ** **** ****** if ** *** ********/********* **** ***** for **, **...

Reactions:

Undisclosed Manufacturer #3

In reply to Undisclosed Integrator #2

* **** ******* ** ***** **** in * ******** * ****'* ****. I ****** ** ***** * ********** script ** ******** **** * ******* Machine ****** ******* ** ****** *******. It *** **** ** ***** *** code ******* ** ******* ** ***** the *******. ** ********, ** **** me **** ************* **** ******, *** how ** ******* *** ********** ******* from ******. ** *** *******. ** was ********* * ****** ****** ****, couldn't ****** ** ***, *** ******* it *** * ****. **** ** 10 ******* * *** ******* ****... It *** ****** ******* **** ** can ********.

Reactions:

(1)

(2)

Undisclosed #1

In reply to bashis mcw

* ****, *** **** ***** *** not *** ******* **** **-******* ** now, **** ******* *** *** *** fact *** **** ********?

* ****-**** ******** **** ****** ** force *****.

**** ***, * ******* **** *** emerging ******** *** *** ***** ********* - *** ******** ***** **** ***** things **** **** *** ***** *** access ** **** ****** - ******* they *** ********** *** **** ******** all *** *****.

**** ****** ***** **** *** ****** it ** ****** ******** *** **** that **** **** ***** *** ***** should ** ****** **** ******** ****, all *** ****.

* ***** ** *** *********** *** humans ***** **** *** ********, *** I ***** **** *** **** ****, with ****** ********* ****, **** ** can ******** **** ****** ******** *** innate ******** ** ****** *******.

Reactions:

(1)

Brian Karas

Pelican Zero

* ***** ** ****** ** **** careful ** ******* ***** ** **** ChatGPT, **** ** *** *****?

**** *** ****** **** **** *** case *** *****, *** ** ****** it *****'* *** ** **** ******** because ** ** *** ** **** or ***** ** *** ***** ****** of ****.

* **** **** *** ** ** extreme ******************, *** **** *** **** more **** ************ ** ******** **** they *** ** *** **** ** real ********/************.

**** ** *** ****** ** ********* or ******** ** *******. ** ** all ********* **** ****-****** *******. ** you *** ** ** ***** * poem, ** ****** ** *******, ** other ****** **** ***** ** *** really * ***** ****** ** ***** pretty *******. *** *** ** *** output **** *** ********** ** ********** qualities, *** **'* ****** *************.

Reactions:

(5)

Undisclosed Manufacturer #7

In reply to Brian Karas

**** ** *** ****** ** ********* or ******** ** *******. ** ** all ********* **** ****-****** *******. ** you *** ** ** ***** * poem, ** ****** ** *******, ** other ****** **** ***** ** *** really * ***** ****** ** ***** pretty *******. *** *** ** *** output **** *** ********** ** ********** qualities, *** **'* ****** *************.

** *** **** ******* ** *** college ****** *** *** *** **** other ******** **** ************* **** *******. He ***** **** ***** ** **** wrong *******, **** ** **** **** seemingly ** ***** ** *******.

** *** ******* ** *** *** writing * ***** ** * ****** figure *** ***** ******* *** **** person ****, ******* ********* **** **** died ** * ****** **** (*****) and **** ** *** **** ** that ****'* ******* ** **** **** died ** *** ***** ** * family ****** (**** *****) *** **** he **** ** **** *** ***** as **** ** **** ** **** multiple ***** ******** **** **** ****** died, **** ** ***** **** ****. He ***** ** **** ******** ****** that ***** **** **** *** ****** of ***** ***** ******* *** ******'*. Every ****** ******** ** ***** **** gave *** ******* ******.

** ***** ***** ****** *** *** it *** ****** *** ***** ******* or ***** **** *********** **** ****. This *** *** *** **** **** and ** *** *** *** **** one *** *********** ******* **** **** far ******* **** *******.

****** ** ** ***** **** ** fine, *** ******* ****** ** ******* with **** ***** *****.

Reactions:

(1)

Brian Karas

In reply to Undisclosed Manufacturer #7

Pelican Zero

** ***** ***** ****** *** *** it *** ****** *** ***** ******* or ***** **** *********** **** ****.

******* **** *** ****** ** ************ than **** *** ** ****** *******.

**** ** * ******* *************, *** when *** *** * ******** ** an *** ** *****'* *** ** find * **** ** ***** **** is ******** ****** ** *** **** of ***** **** ***** ** **** question, ** * ****** ****** *****. Instead, ** ****** **** ******** **** into ****** (********** *****, ** **** pairs/groups), **** ** ***** ***** ***** that **** ** ** **** ******* linked ** ***** ******, *** **** it ******** ********** ******** **** ********* resembling * ************* ******* ****** ******.

*** ************** *** ******* ********* ****** we *** ****** **** **** ** the ****** ** **** **** ** word ******** *** ****-**-*** **** ******. I ***** ****** *** ******* **** because ** *** **** **** ** is ***** **** ** **** *** word ************ **** *** **** *******, and **** ****** ****** ** *** training ****, ***** ****** *** ** being *** *** ****** *******. ** some **** ** **, *** ****** in * ****** *** **** ** a ********** ***, ** **** ***** sense.

Reactions:

(3)

(1)

Undisclosed Integrator #2

In reply to Brian Karas

* *** *********** ****** ** *** AI ***** ** ** *** ****** + ******* ********** ** *** *****, but * ******** **** ** ***** probably ** ************** ***** ** *** the ***** * **** ****** "***, I've ***** **** ** ****" ** "That ******* ***'* ********* ***** ***".

** ***** ** ****** ** **** could ******** **** ** **** ******* of ***** ******** **** ***** *** attention ** *********. *** *******, ** response ** * ********, * ********** do *, *, *** *. ** response ** ******* ********, * ** A, *, *** *. ******** ****** is ***** *** ** ***** *******. If * ***'* ******** *********, * say ****.

* ***** **** ** **'* ******, LLMs **** ****** **** ************* ********. For ** ** ******* **** ********* really ******, **** **** ** **** it ** **** **** ***** ***** on ****. *** ******** **** ******* with **** **** **** **********, *** they're *** **** ** *********, * high-powered ************, **** *** ****.

Reactions:

(1)

Undisclosed Integrator #2

In reply to Undisclosed Manufacturer #7

****** ** ** ***** **** ** fine

* **** **** ****** **** *** success **** ****, *************** ***** *** **** **** **'* giving ***.

**** * ******** - * *** of ********* ** *** ******** ********* ************* **** *******. **** **** for ***** ******** ** **** ** all *** ****** ***** *** *****. Sometimes **'* **** *** **********, ********* people **** ***'* **** ******. ********* on ***** *** *** ******** **** came ****, ** ***** ***** *** some *********** ********** ****.

**** ******* ********* ***** ** **** does *** **** **** **'* ******. Little **********/**** ******* *** ** ****, but ** *** *** ** ***** complicated **** **** *** ******* *** database ************ ******* * **** ************* of *** ** ****** *** **** or ****** ******** *** **** ********... please ***'*. ***'* ******* ****/***** **** and ***'* ******* *** ***-********* ****.

Reactions:

(2)

bashis mcw

In reply to Undisclosed Manufacturer #7

** *** **** ******* ** *** college ****** *** *** *** **** other ******** **** ************* **** *******. He ***** **** ***** ** **** wrong *******, **** ** **** **** seemingly ** ***** ** *******.

**** **** ** ********** ** ****, and *** ****** *** * *** ChatGPT ** *** **** ** **** it ***** *** ********* *** *******. Well, * ****** *** *** ****** ChatGPT ***** *** **** ** **** care ***** ** ******* ****** ****, but ** *** * ******** ******* that ***** ** ****** *** ******* trust *** *******.

Reactions:

(1)

Undisclosed Manufacturer #4

*** ****** *** ******* ***** *** output ***** ** *******.

*** ******** ** * ***** ******** model (***) **** ******* *** ** seen ** * **** ** "*************", in *** ***** **** ** ********* text ***** ** ******** *** *********** associations ** *** ******* **** *** training ****, ****** **** **** ************* or *********.

**** **** ******* ************* *** ****** independently ****** *** ******** ** ************ of *** *********** **** ********.

Reactions:

(5)

(1)

Undisclosed Integrator #5

***. ** *** ***** *** ***** and ***********.

***** ****, ***.

******* * ******** *** * ****** you *** ********, **********.

******* **** ***** ************ *** ** resume, ** ** **** ***!

******* ******** *********** ********** ** **** them ***** ******, ***.

****** ** ******* *** * ******* plan ** ******* ** * *********** or ***********, *** ** *** ***** enough **** ***** *** ******* *** modify *** *******.

***** * ********* ******* ** ******** trends. ****. *** **** ** ******* to **** ** *******.

****** ***** ** ** ***** *** need ** ********** *** ***********. ** yes, ** *** ********** *** ** works.

Reactions:

(2)

Undisclosed Integrator #6

******* **** *** ********** **** **** its *********** *** ** ****** ********** those *********** ****** ** ***** ****** its ****** ** *** ****.

***** **** ******* ** ** ** a ***** ********* ***** ************ *** it ** ***, *** ****** * trained ********* **** *** **** *** vast ******* ** **** *** *** ALL ****. ** *** ***** ********* knowing **** ** ******* ******* ***** be *********** ******** ** * **** amateur, ** ** *** **** ***** before ************ ******** ** ***** ********* information?

*'** **** ***** ******* *** **** time ** ********** ** ******** ********, it *** ***** ************* ******* ********** exceedingly ******* **** **** ******* ******* changes. ** *** ******* ****** ** code *** ****** **** *** ** may *** ****, *** ************ **** are **** ********** *** **** ******* minimum **** ** ******* ****** ** countless *****.

******* ** *** ********* ** *** Internet ** ** ***'* **** ***** up *** **'* ***** **** *** a ****** ** **'* *** ******* so ** ***'* ******* ******.

**** *** **** *** **** ** good ** *** ****. *** *** buy * ****** ********* ****** *** in ***'* **** *** * ****** carpenter.

Reactions:

(2)

Arne Furre