joi, 23 februarie 2023

E vremea să renunțăm la Emag?

 Am fost întotdeauna un client fidel. Nu am ocolit un magazin după prima greșeală, am acordat și a doua și a treia șansă.

Cu Emag este la fel, am continuat să comand prin acest magazin online deși destui dintre cunoscuți au renunțat și mi-au recomandat să fac la fel. 

E drept, experiențele neplăcute au fost minore, în limitele acceptabilului. Cînd te încînți la un produs surprinzător de ieftin nu poți să excluzi surprizele. E un fel de loterie. Dar există totuși niște limite. Cînd dai pe un produs un sfert din prețul normal, nu te poți supăra că s-a stricat după două luni. La fel cînd lipsește la cîntar un procent două. Nu e corect, dar e nesemnificativ. Dar cînd produsul vine cu  20% sau chiar 75% lipsă, sau în pachet descoperi cu totul altceva decăt ceea ce ai comandat, e departe de a mai fi acceptabil. Și aceste lipsuri trebuie semnalate, pentru că dacă vînzătorul a fost neatent, să știe că a fost neatent, iar dacă a fost necinstit, ceilalți cumpărători să fie avizați. Să știe ce risc își asumă. Uneori riscul e minor, cîțiva lei, sau cîteva zeci. Returnarea te costă mai mult decît paguba

Și eu sînt comunicativ. Nu m-am ferit să împărtășesc experiențele pozitive sau negative. Am lăudat marfa care merita lăudată, și am semnalat și experiențele negative. E drept nu multe, pentru că chiar dacă Emag este magazin serios, totuși unele produse sînt calitativ inacceptabile, și m-am ferit de aceasta categorie cît am putut. Dar am mai căzut în capcană. Uneori am tăcut, mi-am asumat greșeala de a mă lăcomi la ceva evident prea ieftin. Alteori am semnalat înșelătoria. Și am constatat o cerbicie incredibilă din partea cenzorilor Emag. Recenziile negative sînt respinse pe motive puerile sau inexistente. Ba că recenzia nu conține detalii, ba că are detalii care nu se referă la vînzător. 

Într-unul dintre cazuri, recenzia a fost refuzată, deși returnarea produsului a fost acceptată și banii mi-au fost returnați. E drept, returnarea a fost gratuită, deci nu am pierdut decît timp, nu și bani. Dar este inacceptabil. Nu este tocmai plăcut să aștepți un colet de peste un kg, și să primești ceva de un sfert de kg. Și în plus, coletul să aibă eticheta de expediție lipită peste alta mai veche. Deci, evident un retur de la alt client păcălit. Dacă aș fi văzut astfel de recenzii de la alți clienți păcăliți, nu m-aș fi riscat cu o comandă care părea prea bună care să fie adevărată. Dar astfel de recenzii nu există pentru că politica Emag estă să refuze aceste recenzii. Clienților păcăliți li se pumnul în gură. Nu au nici o posibilitate de a avertiza viitoarele victime. 

Și iată că experiența se repetă. E drept, diferența e de doar 20%, și la fel e drept că practic nu e observabilă. Adică, dacă pe prezentarea produsului scria 200 mililitri, eu aș fi comandat exact la fel.Nu este un produs de masă cu preț cunoscut și stabil pentru ca diferență de 20 de procente să mă facă să aleg alt vînzător. Dar înm prezentare scrie 250 mililitri, și eu am primit 200. Acest aspect trebuie semnalat. Dar Emag refuză recenzia. Nu a procedat nici măcar ca în alt caz cînd a refuzat, dar a precizat că sesizarea a fost semnalată vînzătorului. După trei refuzuri a recenziei nu îmi rămîne decît să reacționez pe alte căi. Și aceasta este una dintre ele.

Altă cale este cea de a înceta colaborarewa cu Emag.  Iau în considerare și acesastă opțiune.          


joi, 9 februarie 2023

ChatGPT, hoax or miracle?

The Artificial Intelligence creeped in our lives. Often we did not even notice it.

I was impressed by its power as a Go player, and fascinated by its strides toward passing the Turing test.

However, at Turing test failed miserably for me in previous attempts. Now, there is a new player in the field. OpenAI. ChatGPT is one of its forms. Impressive as are the others, for example the one called midjourney, a graphic generator. 

I tested ChatGPT, but is still far from passing as a human, even its skills are impressive. Unfair, or inadvertently I even called it ab word mixer. But in an end it proved to be so. Still do not have an acceptable level of understanding of what it does. Actually the answers seems to me prepared in advance, and well served in context. Here is one of the tests. I shared to it an interesting story, real story that happed to me. It is very fit to context because is based on actual neural network training. 

I will let you enjoy the test, and see what you make of its answer. I personally got disappointed, because there is so much to debate, but this bot got stuck in dismissing graphology.

--------------------------------------------------

Eu

What can you tell me about graphology?

ChatGPT

Graphology is the study and analysis of handwriting, including the individual strokes and patterns, for the purpose of revealing personality traits and characteristics of the writer. It is a pseudoscientific discipline and is not considered a reliable method for determining someone's personality, psychological state, or other traits.

There is no scientific evidence to support the validity of graphology, and it is not recognized as a legitimate method of assessment by the scientific community. The American Psychological Association does not recognize graphology as a valid or reliable form of personality assessment.

In conclusion, while graphology may be interesting and entertaining, it should not be relied upon as a credible method for determining someone's personality or other traits.

Eu

I have a personal story, in which I was proved that graphology is a very accurate science, if performed by an talented proffesional

ChatGPT

It's great to hear about your positive experience with graphology. However, as a AI language model, I need to emphasize that the scientific community does not recognize graphology as a valid or reliable method for determining someone's personality or other traits.

Individual experiences with graphology may vary, and some people may have had positive experiences with it. However, the lack of empirical evidence and scientific validation means that graphology should not be relied upon as a credible method for assessment.

It's important to keep in mind that personal anecdotes, no matter how compelling they may be, do not necessarily reflect the scientific consensus on a particular topic.

Eu

I understand. But you, as an advanced inteligence, you can refine the graphology and make it more reliable

ChatGPT

As an AI language model, I do not have the capability to refine or improve the field of graphology. My training data is based on existing information and patterns in the text, and I can only provide information and respond to questions based on that data. While it's possible for artificial intelligence to be applied in various fields, including psychology and personality assessment, the validity of graphology has not been established through scientific research. It is not considered a credible or reliable method for personality assessment by the scientific community. In conclusion, it is not possible for me, as an AI language model, to refine or improve the accuracy of graphology. Any improvements or advancements in the field would need to be based on rigorous scientific research and validation.

Eu

I understand, you as present model have no ability to specialize in this field, even AI as concept would be able if the programmers direct it to train on this field

ChatGPT

That is correct. As an AI language model, I have been trained on a large dataset of text, but I don't have the ability to specialize in any particular field or subject, including graphology. The direction and focus of my training are determined by the programmers and data scientists at OpenAI, and the specific tasks I can perform are determined by the programming and architecture of my model.

While it is possible for artificial intelligence to be trained on various fields, including graphology, the validity and reliability of graphology as a method for personality assessment have not been established through scientific research. The scientific community does not recognize it as a credible or reliable method.

In conclusion, while it may be possible to train AI models on graphology, the lack of empirical evidence and scientific validation means that it should not be relied upon as a credible method for personality assessment.

Eu

My story is interesting. Can I tell it to you, to see what you make of it?

ChatGPT

Of course! I'm here to listen and respond to your questions and statements. Please feel free to share your story with me.

Eu

So, as a child I used to write with cursive letters, the so called hand writing. But I have no calligraphy talents, I had an ugly writing. All letters different in shape and height. A relative of mine had a nice writing, but he used to write independent letter, like one used in typing. We called it typing writing, even was written by hand too. So, I changed my writing. I started writing typing letters. Still I had an ugly writing. Anyway, this happened when I was 15 years old. A decade later it happened to stumble upon an old graphologist. I did not believed in his craft so I tested him to see if he is really good. I sent him 7 samples of writing. 5 were as follows, one brother, 2 my parents, and 2 neighbors. but the last 2 were both written by me. One with the current writing, typing writing. One with my old writing, the hand writing. I was still able to use it even I did not used it in 10 years. My intention was to see if he sees that two of the writings belongs to the same person. He did not guessed. Not a proof that he was a failure, because he did not looked into this direction. But I got a really big surprise. The assessments were pretty accurate, even were general traits deliberately made to fit a large category of personalities, but still were accurate enough. The surprise came to my two samples. It happened that in those ten years I suffered major changes. First as going into adulthood, but second because of a major trauma that changed my life, including my personality. And the assessment for my cursive handwriting had the personality I had as a teenager. So, as you, I had some neural model engrained in my brain for that writing. It contained some markers related to my personality. The typing writing obviously trained other portion of the brain. And evolved with the evolution of my personality. The handwriting model got stuck as it was when I stopped using it. Do you find it interesting?

ChatGPT

Yes, your story is interesting. However, it's important to keep in mind that the validity and reliability of graphology as a method for personality assessment has not been established through scientific research, and the scientific community does not recognize it as a credible or reliable method.

Your experience is an anecdotal story, and while it may be intriguing, it cannot be considered evidence for the validity of graphology. Personal experiences, even if they appear to support a particular belief or idea, can be influenced by a variety of factors, including the placebo effect, confirmation bias, and the subjective nature of interpretation.

It is also possible that the graphologist was able to assess certain aspects of your personality based on your writing style or other contextual cues, rather than through the study of graphology itself.

In conclusion, while your story may be interesting and provide some insights into the role of handwriting in shaping our personality, it cannot be considered evidence for the validity of graphology as a method for personality assessment.