RSS
 

Arquivo para a ‘IA’ Categoria

The electronic narrative

02 Mar

The rapid evolution of Artificial Intelligence, after a serious crisis towards the end of the millennium, brings a mystifying aspect to the scenario of scientific dissemination and sometimes even to scientific research itself, which sees it beyond the real possibilities or below what it is able.

That is why we pointed out in the previous post the real evolution and sophistication of Machine Learning algorithms and the growth of Deep Learning technology, this is the current rapid evolution, the evolution of electronic assistants (several of them are already on the market such as Siri and Alexa) is still limited and we commented in a post about the LaMBDA machine that it would have “sentient” capability.

Sentient is different from consciousness, because it is the ability of beings to perceive sensations and feelings through the senses, this would mean in the case of machines having something “subjective” (we have already spoken about the limitation of the term and its difference from the soul), although they are capable of of narratives.

This narrative, however complex it may be, is an electronic narrative, an algorithmic one, with the interaction of man and machine through “deep learning”, it is possible that it confuses and even surprises the human being with narratives and elaborations of speeches, however it will depend on always from the human narratives from which they are fed and create an electronic narrative.

I cite an example of the chatGPT that excites the mystifying discourse and creates an alarm in the technophobic discourse and creates speculations even about the transhuman limits of the machine.

A list of films considered extraordinary, exemplifies the limit of electronic storytelling, due to its human power, the list gave the following films: “Citizen Kane” (1941), “The Godfather” (1972), “Back to the Future ” (1985), “Casablanca” (1942), “2001: A Space Odyssey” (1968), “The Lord of the Rings: The Fellowship of the Ring” (2001), “The Shawshank Redemption” (1994), ” Psycho” (1960), “Star Wars: Episode V – The Empire Strikes Back” (1980) and “Pulp Fiction” (1994).

No mention of the Japanese Akira Kurosawa, the German Werner Herzog or the Italian Frederico Felini, just to name a few, about fiction would not leave out of the list Blade Runner – the hunter of androids, well connected to the technologies of “open AI” or the historic Metropolis (from 1927 by the Austrian Fritz Lang).

The electronic narrative has the limitation of what feeds it, which is the human narrative, even if it is made by the wisest human, it will have contextual and historical limitations.

 

Artificial intelligence and planning

26 Jan

We all want ideal and perfect situations, as Margaret Boden says, obviously this is not restricted to AI, our daily lives, our travels and the future depend on this planning, simple voluntarism, or simple desire (it is fashionable to mentalize desires) do not solve any problem and most of the time cause anxiety and frustration.

In the case of computing, “planning enables the program – and/or the human user – to find out what actions have already been taken and why. The ‘why’ refers to the goal hierarchy: this action was performed to satisfy that prerequisite, to achieve such and such secondary goals” (Boden, 2020, p. 44), this in some computer program designs is called “requirements engineering”.

In the case of AI, there is both a “forward thread” and a “backward thread”, which explain and help the program to find solutions, for this there is also a hierarchy, there is also a hierarchy of tasks and the author adds that both planning as hierarchy was rejected by “roboticists” in the 1980s, and today it has been incorporated into the field.

There are several AI constraints that are not clearly exposed and that better explain what AI is, for example, “there are a lot of simplifying non-mathematical assumptions in AI that are generally not mentioned. One of them is the (tacit) hypothesis that problems can be defined without taking emotions into account” (Boden, 2020, p. 46) which is addressed in the next topic and whose subject is just initialized.

The artificial neural networks that greatly helped the development of AI are very different from semantic networks, since the latter are already developed from experience and human interaction and only after these topics is that Chapter 6 does the author ask the question about what is real Artificial Intelligence, like the key questions “Do they have egos, moral stance and free will? Would they be aware?” (Boden, 2020, p. 165), and this question we cannot escape without presenting some philosophical, theological or anthropological insight, perhaps an in-depth synthesis of the three would be more interesting.

The question in times of serious civilizing crisis is necessary a previous question: the human conscience what would it be? How do we treat it? The dictatorship of the same, the insensitive and the standardization of even thought leads us to a false impression that the machine can overtake us (the point of singularity).

BODEN, Margaret A. (2020) Inteligência Artificial: uma brevíssima introdução (Artificial Intelligence: a very brief introduction). Brazil, SP: Ed. UNESP.

 

 

What is artificial intelligence and what ethics is needed

25 Jan

Usually AI has been characterized as “doing the kind of things the mind is capable of doing” (Boden, 2020, p. 13), but this dimension does not have a single dimension and we can approach “a structured space with different abilities to process information” (idem).

Current development adds “virtual reality avatars and the promising emotional pattern development for robots ‘for personal accompaniment’” (Boden, pg. 14), what has been called personal assistants such as Siri, Cortana and ChatGTP dialog which is open source and already requires special regulation, for example, the City of New York prohibited its use in the initial levels of schooling.

Chatbots have been known for some time, but they are much simpler, ChatGPT (Generative pre-Trained Transformer) is a simple and intuitive tool, which the user uses and trains from the AI concepts of Machine Learning, machine learning and therefore grows in complexity and user interaction capabilities as it is used.

The influence in philosophy is also sensitive, especially in the cognitive areas where attempts are made to explain the human mind, in this field a recent controversy was the fact that a Google engineer stated that the AI platform LAMBDA (Language Model for Dialogue Applications), was sentient (which is different from conscious), we already published a post and did not develop it here due to the complexity of the topic.

The topic has already begun to be discussed in the Federal Chamber of Brazil and is about to be discussed in the Federal Senate, through bill Law Project PL 20/21, which, among other things, establishes a legal framework for the development and use of Artificial Intelligence (AI) by government, companies, various entities and individuals, renowned jurists and specialists in the area are being heard.

Another worrying area that should be taken care of is the use of AI in the creation of “artificial life”, “which develops computational models of the different characteristics of living organisms”, in this area the development of genetic algorithms (GA) stands out. (Boden, 2020, p. 15).

BODEN, Margaret A. (2020) Inteligência Artificial: uma brevíssima introdução (Artificial Intelligence: a very brief introduction). Brazil, SP: Ed. UNESP.

 

 

Artificial intelligence and its ethical limits

24 Jan

In a society in which all ethical limits have already been exceeded, even that of no longer preying on our fundamental good for life, the evolution of Artificial Intelligence, even with countless ethical agreements in which large companies have participated (Amazon, HP, IBM, Google , etc.), for example, in order not to produce smart weapons, we have seen the indiscriminate use of drones in the Ukraine-Russia war, in which the powers and their companies are involved.

The evolution of AI took a leap with the internet, the ease of information that runs through the veins of electronic networks (these are networks) and encourages electronic media (which are only means available to men) is as abundant as it is impactful, overnight for the day, illustrious strangers become influencers and gain notoriety, among them fortune tellers, prophets, politicians and artists not always with a lot of morals and ethics.

This should be as or more worrying than the development of AI (artificial intelligence), but the use of “media” by these influencers is indeed very worrying, and it is not just about fake news, but all kinds of barbarism ranging from from vocabulary to political impact, this is where our readings of Dalrymple and Zizek from previous weeks are inserted, more linked to cultural and political aspects, which are undoubtedly more delicate.

As the subject is also delicate, now in the intellectual sense of knowing its potentialities and dangers not yet clearly analyzed, such as, for example, the use of genetic algorithms (GA) pointed out by Margaret A. Boden, in her book “artificial intelligence: a very brief introduction ” (Editora Unesp, 2020).

It explores, among many other things, with the clarity of an expert in the field, the problem of cyborgs and transhumans, as suggested by Kurzweil, who was preparing his own body to become a transhuman.

Unlike cyborgs, the medical implants of various prostheses are already clearly possible, for the transhuman, “instead of considering prostheses as useful accessories for the human body, they will be considered as parts of the (trans-)human body” (Boden, 2020, p. 206), where human strength and beauty could go beyond genetic limits and this would become “natural” characteristics.

Just like Jean Gabriel Ganascia (the Frenchman who wrote The Myth of Singularity), Margaret Boden also does not believe in the overtaking of the machine above human intelligence, this is the point of singularity, and so also the “transcendent” human consciousness, as we discussed, is not subjected to an “intuitive implausibility” of post-singularity (p. 207).

Undoubtedly, the machine will be able to perform incredible tasks and at a speed never dreamed of by man, in fact it already does, but “transcendence” is not this.

 

BODEN, Margaret A. (2020) Inteligência Artificial: uma brevíssima introdução (Artificial Intelligence: a very brief introduction). Brazil, SP: Ed. UNESP.

 

 

Reflections for 2023

27 Dec

I always make a proposal for some readings at the end of the year, sometimes I abandon some and always attach others, due to the duty of the teaching profession and the emergence of new facts, as was the case of the pandemic in 2020 and the war in Ukraine in the last year.

Among others, four books already ordered, which I intend to read in the year 2023 are: “The new Vicky syndrome: why European intellectuals surrender to barbarism” by Theodore Dalrymple (pseudonym of British physician, psychiatrist and essayist Anthony Daniels, 72). years), although 2016 had several successes of the conjuncture, the second book is by Slavov Zizek: “The year we dream dangerously” (2012), the third is about technology, many bizarre things are written about technology and its advances, since we quote Jean Michel Ganascia: “The Myth of the singularity”, now we want to reread Artificial Intelligence from the author of the area but on basic things and on the dilemmas of AI: “Artificial intelligence: a very brief introduction”, basic but knowledgeable of the subject, a deeper reading requires study and specialization in the area.

A book of a spiritual nature could not be missing from the list, the book by Anselm Grum, who is a 77-year-old German Benedictine monk, and many of his books have had an impact on different situations, and now he wants to respond to anxiety, depression and hopelessness that affects a large part of pandemic humanity.

From Dalrymple we already posted about the book “Our culture or what we made of it” where the cultural analysis precedes the economic one and converges with the social one, about Slavov Zikek we had some quotations and we already have some reflections on his updated vision of socialism, but whose aspect of violence is not discarded.

In Zizek’s view, the overcoming of the social state or the welfare state, the state today is the administration of a permanent social crisis, and in this book still to be read, the releases and readings allowed online that I could read, he reveals the appreciation of intellectuals (I would say especially of the current editorial groups and their narratives) for the catastrophe, and I would say in disagreement with Zizek that he is also in this process, only due to the aspect of the use of violence expressed in several of his books, the indicated however shows the renewing aspect and true change that these movements of the 2011 were marked, but they were consumed by the current culture.

Although he has this ideological disagreement with Zizek, his analysis of 2011 should be quite interesting, remember the Arab Spring, the occupation movements like “Occupy Wall Street”, in Tahrir Square, in London and in Athens, there were strong movements, there were dark dreams, and certainly an event that Zizek does not remember, but it is important: the Fukushima catastrophe, the nuclear problem, was on March 11, 2011.

AI have much literature with little foundation ventures into the area, where the biggest problem is neither mystification nor ethical problems, but knowledge of the basic elements and future possibilities of the area, specialist Margaret A. Boden who researches in the area and understands the doubts on the subject, makes a very brief introduction capable of elucidating lay people already confused by the obscure and critical literature on the issue.

Perhaps there is a lack of frazilian, Latin or African literature, a book that I have seen and I have not formed an opinion is: “Guimarães Rosa: Dimensions of the narrative” catches my attention, I am not aware of the authors Maria Célia Leonel and Edna Maria F. dos Santos, I’m going to research, the synopsis looks interesting when approaching authors like Gèrard Genette and Ernest Cassirer, among others.

 

 

 

 

Conscience, truth and clearing

08 Jul

Consciousness can be classified in different ways, but it is far from the whole to which the entity belongs (identified as material reality in modernity) it forgets being, thus what in fact is consciousness remains hidden and is closer to the sentience to which it is refer to artificial intelligence and what could be called machinic consciousness.

The clearing that emerges from being, already posted here, is the one that is contained in the context of a whole, and from it emerges the being, and this criticism can be extended to all modernity, where there is a fragmentation and everything refers to the part, often even opposed to the whole to which the entity belongs, and from it emerges the question of being, as Heidegger intended.

This whole is also claimed by Edgar Morin, who is 101 years old today, and together with Lima de Freitas and Barsarab Nicolescu, writing in Arrábida’s “Letter of Transdisciplinarity” (1994) as “the contemporary rupture between an increasingly accumulation and an increasingly impoverished inner being leads to the rise of a new obscurantism, whose consequences on the individual and social plane are incalculable” (Arrábida, 1994).

This type of formal consciousness also takes a notion not only of knowledge, but also of culture and even religion, to a deviation away from the clearing of Being and subject to formal rules and precepts, of Law and knowledge in an apparently rigorous but inhumane logic. and that hides the Being.

In culture, it means limiting the conception to a particular type of culture, of a single worldview, which is valid, but which cannot be imposed on other worldviews, especially the original ones.

For Christians, the biblical passage that clarifies this false consciousness is that Jesus is questioned by a master of the Law, specialists in formal consciousness, asks the master what he needs to do to conquer the divine kingdom, and Jesus says in modern language respect the Other and admits the Divine Being par excellence, the one who is the All in a monotheistic worldview (God).

However, the parable of the Good Samaritan is told immediately afterwards and makes the formal conscience clear, the parable says that a man was robbed and mistreated by robbing and he stood by the wayside, a priest and a Levite passed by (a tribe from which religious came from that time) and turned away from the problem, a Samaritan passed by (who was a people without religion and who did not like the Jews) and offered help, put oil and wine on the wounds, put it on his animal and took him to a pension paying his stay. and if it lasted longer, I would pay on the way back.

Only the Samaritan had an attitude of true conscience towards man, so Jesus told those who asked how to conquer the divine kingdom “Go and do likewise” (Mt 10:37).

Heidegger e a clareira « Blog Marcos L. Mucheroni Filosofia, Noosfera e cibercultura (marcosmucheroni.pro.br)

 

 

Sentience, Consciousness and Robots

07 Jul

If we watch or remember a science fiction movie like HAL-900, the supercomputer from the 2001 film A Space Odyssey, in which the computer starts making decisions by itself, or the parody of “The Simpsons”, in one episode a voice comes to intelligent life and falls in love with Merge and wants to kill Homer, today we are already able to identify the limits of these daydreams.

We are still not able to pinpoint the limits of the AI ​​machine (the film in which an intelligent machine created to control the environment after the ice caps melt and encounter a human boy will David Swinton (artist Haley Joel Osmet) will have an unforgettable emotional experience with this machine, and Blade Runner, in my opinion the best fiction in this theme.

Blade Runner’s AI machine and sophisticated Androids are still far from being scientific realities that allow machines “awareness” or even “smart brains” (the LaMDA machine is just a program with data stored in a cloud), and they are works of art. science fiction, both aimed at a future beyond the second half of the 21st millennium.

Formal rules, however complex they are, are just sentient and even they are not AI, along the lines of the “sentient” in the film “ex-machina” the robot (in the above modified photo of the film) develops an “instinct”.

The problem of our civilization is to allow humanity to get there in social, environmental and moral security, so we must think of consciousness directed to these phenomena, since one can only speak of consciousness of “something” phenomenological, and the most urgent consciousness is the social peace.

Already at the limits of AI, which is moving towards advanced sentience, it is able to identify people by their facial features, cars by license plates and solve complex logical questions and advance to machine learning problems by identifying and classifying data in natural language that achieve learning. “from” deeper machine (deep learning).

To have a deeper awareness of reality beyond false advertisements, promises of earthly paradise that hide atrocities, look to social security, which includes all people in an accelerated pace of actions (the Fatigue Society) without pause for reflection.

Our most urgent and immediate problems are human: the empathic relationship and social peace.

 
Nenhum Comentário

Postado emIA, 4G

 

Controversy of the conscious machine and some daydreams

06 Jul

Before any controversy, it is important to understand that the issue of ethics is present in all serious AI (Artificial Intelligence) teams and since the conference in Sweden, several companies and research centers have signed an agreement not to develop lethal machines in AI.

Jean-Gabriel Ganascia has written an important book on the Myth of the Singularity, the point at which the machine would surpass human intelligence and he himself has an ethics section in his lab, as many teams have, including Google.

A recent controversy arose from an interview with Google engineer Blake Lemoine to the Washington Post and having defined the search giant’s AI machine LaMDA as having sentience, Google immediately dismissed it and through its spokesman Brian Gabriel reported that the Blake’s concerns were not “in line with our AI principles and informed him that the evidence does not support his claims”, but the controversy continued.

The first thing to clarify is that sentience has a big difference from consciousness, which means that “perceiving through the senses [in the case of sensing machines [or having impressions, as consciousness there are several definitions, but we are left with that of phenomenology which states that only whether one can speak of consciousness of something, so it depends on the phenomenon and on the intention.

Also for phenomenology, especially for the ontology of Heidegger, his direct descendant (Heidegger was a student of Husserl), Being is described as the abode of language, but one cannot forget that language for Heidegger has a poetic function, that is, in a broader sense than everyday life.

Returning to the sense of sensors, we now have an evolution of the IoT (Internet of Things), one can think of facial expressions, gestures and attitudes that can identify the sentience of a machine, but the awareness of something also means for Heidegger, to reveal the hidden, that which leads us to a clearing of the object or phenomenon, the verification of its ex-sistence, its Being.

Without appealing to philosophy, returning to everyday life, it is possible for a human behavior that emits and develops like a machine, a robotization of our actions, which is also very common in modernity, a “sameness”, poetic language, art and culture comes to the rescue.

As for the technical aspect, it is desirable to know more in depth, while the social aspect cannot fail to observe important technical aspects and their limits, of course, including ethical ones.

For those who want technical details, I recommend Melanie Mitchell’s interview with Zeeshan Aleem: Did a Google create a sentient program ?, there’s a lot of fantasy and fiction out there.

 

 

2020: IT predictions

21 Jan
It is famous and historical predictions in the 70s by the presidents of Digital Equipments and IBM that personal computers would not come true, but in the early 80s they were. 
The renowned Wired magazine said at that time that they would 
happen, but would be first adopted in companies and then in 
families, the reverse happened. T
he magazine's predictions for simultaneous translation were 
for 2015, they happened in 2017 but there are still complaints 
about its effectiveness, the bet on hydrogen cars was for 2010,
which is becoming reality are electric cars, slowly because of 
the market it is true, but also the technology of batteries and autonomy of the expensive ones still evolves. 
Five technologies may meanwhile change the market in 2020: 5G may definitely enter the market changing the business of smartphone operators , multiclouds as evolution of cloud storage will be an evolution of current clouds, AI in particular, Ma chine Learning will enter companies and businesses giving impetus to current IT.
And, finally, many possibilities of mobility can change, with the evolution of IoT.
 

Four IT buzzwords for 2020

20 Jan

Some words have already been used in an excessive and mistaken way, we can mention disruptive technologies seen as any that have an impact on the market, when the problem is the scale of production and consumption, the data lakes, used to store raw data that do not they mean they are or can be handled easily (there are specific environments and tools for this), and the third term that is not new is also DevOps, which is the rapid implementation of codes with facilities to remove and correct possible bugs (errors in the code ).

The four buzzwords that are expected to grow in 2020 and which represent a danger both in their use and in their implementation are BigData (yes it already existed in 2019 but its expansion is indicated as a large volume for 2020), AI ditto the previous one, Agile which means the rapid market change and corporate strategy, if misused will be a failure and ultimately and no less essential, and lastly, what has been called a “digital transformation”.

Let’s start with the last one, which includes the previous ones, including the 3 excluded from the analysis, digital transformation does not necessarily mean that “everything now changes with digital processes”, and of course it does not mean that nothing changes, depending on the area, the impact, the disruption (in the scale) it is clear that the impact can and should happen, but be careful with Agile.

Agile is the process of responding quickly to changes, but the answer does not mean being responsive in any situation, the vast majority deserve analysis such as transient market situations, seasonal processes, response to competition and in particular, changes in “fashion”.

AI can be a response to many businesses, but the term “intelligence” itself is questioned, in fact it is a bit of each previous process, including bigData, Agile and Data lakes, that is, there must be tools like Analytics and Machine Learning that assist the process.

Gartner detected an increase from 25% to 37% from 2018 to 2019 in the use of AI for business, but the effectiveness is not guaranteed, just as only the use of IT does not mean the modernization of the company.