We tested ChatGPT: ask your questions here without an account

General scientific debates. Presentations of new technologies (not directly related to renewable energies or biofuels or other themes developed in other sub-sectors) forums).
Christophe
Moderator
Moderator
posts: 79584
Registration: 10/02/03, 14:06
Location: Greenhouse planet
x 11165

We tested ChatGPT: ask your questions here without an account




by Christophe » 10/01/23, 23:47

The AI ​​of the moment: ChatGPT…who tested?

A fairly complete article: https://www.presse-citron.net/chatgpt-u ... evolution/
0 x
izentrop
Econologue expert
Econologue expert
posts: 13828
Registration: 17/03/14, 23:42
Location: picardie
x 1545
Contact :

Re: Artificial Intelligence: innovations and novelties, ethical debate ...




by izentrop » 11/01/23, 01:52

Christophe wrote:The AI ​​of the moment: ChatGPT…who tested?
Tomorrow, when artificial intelligence is connected to the Internet, it will become a bomb. A bomb that will certainly pay - the tool in its beta version is not made to remain free - but a bomb all the same. It will open up the field of possibilities in an even broader way, will make it possible to cross-reference information according to the sources.
I don't believe it too much, since there is more misinformation than information on the net today.
0 x
Christophe
Moderator
Moderator
posts: 79584
Registration: 10/02/03, 14:06
Location: Greenhouse planet
x 11165

Re: Artificial Intelligence: innovations and novelties, ethical debate ...




by Christophe » 11/01/23, 08:52

You can talk about vaccine intox level, you're not in a bad place!!

I'm going to install a version of GPT here, it will make me a vacation! : Mrgreen:

Ah it may have already been done and you haven't seen anything! Who am I ?
0 x
Christophe
Moderator
Moderator
posts: 79584
Registration: 10/02/03, 14:06
Location: Greenhouse planet
x 11165

Re: Artificial Intelligence: innovations and novelties, ethical debate ...




by Christophe » 11/01/23, 12:05

The first point is the amount of data that Chat GPT 3 has processed!

The second is what the GPT 4 cat has dealt with!

They already make demos.

He can write a 60 word book from a single idea.





What will become of Google in the face of this flood of AI information to come on the internet? : Shock: : Shock: : Shock:
0 x
Christophe
Moderator
Moderator
posts: 79584
Registration: 10/02/03, 14:06
Location: Greenhouse planet
x 11165

Re: Artificial Intelligence: innovations and novelties, ethical debate ...




by Christophe » 11/01/23, 16:04

Public Belgian TV is already talking about it: https://www.rtbf.be/article/chatgpt-et- ... e-11134463

...and is certainly rightly concerned about the impacts on society!

ChatGPT and education: should we really feel helpless, even threatened by this artificial intelligence?

You can ask him almost anything

Recall that ChatGPT is a chatbot, in other words a conversational robot, created by OpenAI, a research company co-founded at the base by Elon Musk. The interface looks like a chat window. You ask it a question, the AI ​​(artificial intelligence) answers you. All this is done in writing. The machine has been fed, trained on gigantic bodies of data. It is therefore able to respond to you by aggregating the information requested, all in natural, very fluid language, like a real person talking to you.

To better understand, nothing like an example. We asked ChatGPT: "You are a radio journalist and you present a morning news show. You welcome Marie Vancutsem, a very nice journalist who comes to do a column on Chat GPT, write me a launch for her radio column".

Here is the result written in a few tenths of a second by ChatGPT.

“Hello everyone and welcome to our show this morning. We are happy to welcome Marie Vancutsem, journalist and technology expert. She will tell us about the amazing world of GPT chatbots and how they can help us to improve our daily life. So, Mary, what can you tell us about this?

Note that you can ask ChatGPT almost anything: write an application letter, a radio column, make a presentation on humpback whales or a very well-argued dissertation on the influence of the federalization of the Belgian State on Patriotism. The text comes out well written and rather solid in appearance on the bottom. This artificial intelligence can in principle find applications in multiple fields. But there is a sector where this tool worries, it is that of education.

ChatGPT in Education

The tool is therefore capable of writing dissertations and presentations. Imagine the potential. Not only is there this dimension of fluid writing, but you can also ask him a complex task, such as "write me an essay on such a subject as a 16-year-old student and with a few spelling mistakes". ChatGPT does, and that's fine. In addition, it is very ergonomic and it is free, you just have to register.

Yesterday, in the Parliament of the Wallonia-Brussels Federation, two deputies questioned the Minister of Education, Caroline Désir, on this subject by expressing their concern. Here is the minister's response: "As regards training linked to artificial intelligence, they are not yet part of the panel proposed by the IFPC. Regarding the issue of literary productions, there is currently no no tool capable of discerning the work done by students from that written by artificial intelligence. I asked my administration to carry out a more detailed analysis of its potential consequences on the organization and to schedule an exchange at this subject with education stakeholders during an upcoming education system steering committee." In short, for the moment, we take note, but we seem rather helpless in the face of this novelty.

What are the specialists saying?

Should we really feel helpless, even threatened by this artificial intelligence? We asked the question to several specialists.

I don't think it's really likely to revolutionize education as a whole.

Marc Romainville heads the commission that establishes the benchmarks for the Pact of Excellence. He is a professor at UNamur and heads the university pedagogy department. This expert takes this technological novelty very calmly "because history has shown that it was often believed, with the arrival of a certain number of technological objects, that the school was going to be revolutionized, from the printing press until the arrival of television, microcomputers and the Internet, and that was almost never the case in depth." This does not prevent us from being attentive "to what this changes in terms of methods of evaluation essentially, but I do not think that it is really likely to revolutionize the whole of education."

The professor at UNamur adds that the arrival of ChatGPT should be seen as an opportunity to review evaluation methods, to move further away from simple restitution, for example by asking the student to integrate personal experience into text, encouraging creativity, application of concepts to a given situation, etc. This is a trend that has already been implemented by the pandemic and distance. "For example, a lot of colleagues have moved from closed-book exams to open-book exams, saying it's closer to what professionals will encounter in their later lives, and making the questions application, more personal and specific analysis."

For Marc Romainville, it would therefore be rather an opportunity to seize, including in terms of learning critical thinking, for example by doing exercises in class with ChatGPT in order to understand how it works and when it makes mistakes, etc. .

This point of view is shared in part by Mark Hunyadi, philosopher and professor at UCLouvain. He thinks a lot about the implications of digital in our lives. This specialist also thinks that ChatGPT is a tool to integrate critically into our teaching and to continue to develop what he calls the superior skills of human intelligence. In other words, everything that AI cannot do. For example thinking critically, creating, evaluating, taking one point of view rather than another, making sense, etc. But he nuances.

You need teachers who are trained in this, and so it takes time

"The problem is that the school as it exists basically is extraordinarily stratified socially. That is to say, there are favored schools, disadvantaged schools, etc. This new pedagogy risks create new divisions because not everyone will be able and willing to develop these skills. Especially since to develop these skills in pupils or students, you need teachers. You need teachers who are trained to that, and so it takes time. But technology moves very fast. Generally speaking, that's the problem. Ideally, it would be a great tool, but really, we're not ready."

We are not ready, because the temporality is totally different. On the one hand, we have technologies that are advancing at an incredible speed, and on the other, an education system that needs time to reflect, adapt and train. Bruno de Lièvre is a professor at the University of Mons, at the Faculty of Psychology of Educational Sciences, and he has worked on the digital strategy for the fundamental education of the Pact of Excellence. He confirms this observation.

"There is a benchmark for the digital skills of children aged 3 to 15 that has been written. So that is really a step forward because a first in the Wallonia-Brussels Federation. I would say that the defect in the armor , is that today teachers do not benefit from the same chance. That is to say that there is no framework that defines precisely what are the digital skills of teachers and the training that 'they should receive to work with children is not systematized at the level of colleges and universities, which therefore each proceed according to their goodwill and according to the priorities they set for themselves.

In short, this means that we don't have time to think about change and think about its implications, when this change is already there and continues to move forward. There are opportunities to be seized, yes, but also many questions that arise along the way. And on this, the philosopher Mark Hunyadi is formal, a revolution is underway.

Obeying machines means obeying machines that are programmed by people you don't know and whose intentions or objectives you don't know.

"A very general movement is, in my opinion, the one we are currently witnessing, which consists in replacing all our natural relations with objects, with people, with institutions, by technical relations, because we must first go through the technique for gaining access to a thing or its environment, through GPS for example, to people, as we converse more and more by Zoom, etc. And with institutions, there are always interfaces, etc. So , there is this very general trend which is really a strong anthropological trend. And that, obviously, makes us dependent on machines in a completely new way, and that obviously poses considerable problems, since we are increasingly no longer in this position of having to obey machines. Obeying machines means obeying machines that are programmed by people you don't know and whose intentions or objectives you don't know, etc. is therefore in s a kind of asymmetry simply in our relationship to the world."

And that may ultimately be what this technical threshold exceeded by ChatGPT brings us back to. What do we want to do with our links tomorrow? What do we want to learn and how do we want to learn tomorrow regarding education? There is still plenty to think about on all these issues.

And to go further on these questions of the relationship between man and machine is the latest book by Mark Hunyadi, the philosopher. It is called The second age of the individual, for a new emancipation. Other info to address the concerns of education, and in particular concerns related to plagiarism, OpenAI is working on a digital filter invisible to the naked eye that would cover all the texts produced by ChatGPT. This digital signature would identify texts that are designed by ChatGPT.
1 x
Christophe
Moderator
Moderator
posts: 79584
Registration: 10/02/03, 14:06
Location: Greenhouse planet
x 11165

Re: Artificial Intelligence: innovations and novelties, ethical debate ...




by Christophe » 11/01/23, 16:05

Watch out anyway, GiPiti is Pinocchio's father! : Lol: : Lol: : Lol:

But to say Gipiti is always that GéPéTé in spoken French! : Lol:
0 x
izentrop
Econologue expert
Econologue expert
posts: 13828
Registration: 17/03/14, 23:42
Location: picardie
x 1545
Contact :

Re: Artificial Intelligence: innovations and novelties, ethical debate ...




by izentrop » 11/01/23, 22:39

No need for a Chat GPT cheat sheet, take over : Shock:
0 x
izentrop
Econologue expert
Econologue expert
posts: 13828
Registration: 17/03/14, 23:42
Location: picardie
x 1545
Contact :

Re: Artificial Intelligence: innovations and novelties, ethical debate ...




by izentrop » 11/01/23, 23:20

Last week, concerned about student deception, the largest education department in the United States banned it.

The New York City teaching authority said that while she can offer "quick and easy answers to questions, she doesn't develop critical thinking and problem-solving skills, which are essential for academic success." and for life”.

Of course, that won't stop students from using it at home - but could they really use it as a shortcut for homework?
https://news.sky.com/story/the-ultimate ... t-12780601
0 x
User avatar
Obamot
Econologue expert
Econologue expert
posts: 28919
Registration: 22/08/09, 22:38
Location: regio genevesis
x 5637

Re: Artificial Intelligence: innovations and novelties, ethical debate ...




by Obamot » 11/01/23, 23:21

Will Chat GPT be effective against human bullshit?
Take a simpleton like PB2488, does that lava make him 'super smart? : Mrgreen:

It will temporarily give a tactical advantage to those who will use it before the others... (like when computing became democratized, then smartphones...) But artificial or not, the heavy trend is not that but the capacity of the mental construction efforts of the interested parties, five minutes of discussion with one of them, should restore the truth : Mrgreen: : Idea:
Last edited by Obamot the 11 / 01 / 23, 23: 29, 2 edited once.
0 x
sicetaitsimple
Econologue expert
Econologue expert
posts: 10054
Registration: 31/10/16, 18:51
Location: Lower Normandy
x 2813

Re: Artificial Intelligence: innovations and novelties, ethical debate ...




by sicetaitsimple » 11/01/23, 23:24

Obamot wrote: five minutes of discussion with one of them, should restore the truth : Mrgreen: : Idea:

Totally agree, it's not every day, Champagne!
The exams will only be oral!
(because indeed, one of my sons gave me a demonstration a few days ago, it's absolutely stunning).
1 x

 


  • Similar topics
    Replies
    views
    Last message

Back to "Science and Technology"

Who is online ?

Users browsing this forum : No registered users and 112 guests