Showing posts with label learning. Show all posts
Showing posts with label learning. Show all posts

Monday, 1 April 2024

National Seminar- Convergence of AI, DH, and English Studies

Convergence of AI, DH, and English Studies

Organised by DoE, MKBU

Participated in a National Seminar hosted by Smt. Sujata Binoy Gardi, Department of English, M.K. Bhavnagar University, Bhavnagar. The seminar, titled 'Convergence of AI, DH, and English Studies,' was conducted in a hybrid mode. The event commenced with a cordial welcome from the Head of the Department and Professor Dr. Dilip Barad. Distinguished speakers included Prof. (Dr.) Nigam Dave, Director of the School of Liberal Studies and Dean of International Relations at PDEU, Gandhinagar, Gujarat, delivered a thought-provoking session on the Value Neutrality of Deepfake and its Positive Implications. Dr. Richa Srishti, Associate Professor and Head of the Department of Languages at CHRIST (Deemed to be University), Lavasa, Pune, Maharashtra, explored the Role of AI in the Creative Process and Redefined Authorship. Dr. Richa Mishra, Chairperson of the Board of Studies and Head of the Department of Humanities and Social Sciences at the Institute of Technology, Nirma University, Ahmedabad, Gujarat, offered insights into Engaging with Digital Humanities: Tools, Techniques and Scope. Finally, Dr. Shobha K.N., Associate Professor of English at NTTTR, Chennai, Tamil Nadu, discussed the application of AI for Teaching and Learning.

This blog is Jheel Barad's Major takeaways from the seminar attended in online mode.



Prof. (Dr.) Nigam Dave on Value Neutrality of Deepfake and its Positive Implications


The first session started with a question: Why do we call the swapping of images or videos "deepfake" and not "Synthetic Media"? We discussed the background of Deepfake and pondered whether deepfakes could shake up global governance. If yes, what can we do to stop it? If not, what have we missed in raising awareness about Deepfake?

We talked about fake identities, a concept rooted in Indian myths. Stories like Chavan Rudhi and Sukanya, and the legend behind Chavanprash, showed how old this idea is. Examples like Ravan becoming a golden deer and Arjuna disguising as Brihanala highlighted that deepfake isn't new; it's how we interpret things that's the problem.

These examples show how things are duplicated, copied, or altered to look like the original, which is the basic idea behind Deepfake - manipulating or changing something to mislead others about someone or something.
  • Meme with text change in the original dialogue
  • Body double
  • Look-a-like celebrities
  • Mimicry artists
  • Singers singing with the voices of known singers
  • Fake products in chor bazaar


The concept of Deepfake or Synthetic Media can be utilized for positive purposes.


Just like guns don't kill people, it's the malicious intentions of people that alter technology.

Dr. Richa Mishra on Engaging with Digital Humanities: Tools, Techniques and Scope. 
In her presentation, she delved into the myriad tools, techniques, and scopes within the realm of Digital Humanities (DH) as a researcher. The session commenced with an observation on the prevalence of theoretical research over practical production in the Indian DH landscape. It provided a comprehensive background on the emergence of Digital Humanities, highlighting key figures in the field. CLICK HERE

Furthermore, the talk explored the contrasting patterns between traditional research methodologies and those employed in DH. Analogous to traditional research, DH research encompasses digitized data, born-digital content, and data revitalized in digital formats. This comparison sheds light on the evolving nature of research methodologies in the digital age.



Additionally, the session delved into various research techniques pivotal to Digital Humanities, including:

1. Text Mining & Analysis: Uncovering patterns, trends, and insights from large volumes of textual data through computational methods.
2. Data Visualization: Presenting complex data in visual formats to facilitate understanding and interpretation.
3. Geospatial Analysis: Analyzing spatial data to reveal geographical patterns, relationships, and trends.
4. Network Analysis: Studying the connections and relationships between entities within a network to discern patterns and structures.
5. Corpus Linguistics: Analyzing large collections of text (corpora) to study language patterns, usage, and evolution.
6. Text Encoding and Markup: Employing markup languages to encode and structure textual data for analysis and presentation.

These techniques underscore the interdisciplinary nature of Digital Humanities, leveraging computational tools to explore and understand humanistic inquiries in novel and insightful ways.

Furthermore, the session highlighted a plethora of tools essential for implementing these techniques:

1. Hermeneutim: An advanced tool facilitating text analysis and interpretation, aiding researchers in uncovering deeper meanings and insights within textual data.
2. Tupor.ca: A comprehensive platform offering a range of tools for textual analysis, enabling researchers to explore linguistic patterns, sentiment analysis, and more.
3. Hypothes.is: A collaborative annotation tool allowing users to annotate and discuss digital documents, fostering collaborative research and knowledge sharing.
4. Recognition: A tool specializing in optical character recognition (OCR), converting scanned documents and images into editable and searchable text, thus facilitating text mining and analysis.
and many more. The session also introduced other tools tailored to specific research needs, including data visualization platforms, geospatial analysis software, network analysis tools, and text encoding software.


The session culminated with an exploration of the diverse scopes available for research within Digital Humanities (DH), as in photo



Recorded Sessions


Dr. Richa Srishti on The Role of AI in the Creative Process and Redefined Authorship
The session commenced with fundamental inquiries: Can AI truly embody the essence of an author? Will AI revolutionize the very concept of authorship? Let us delve into the roots of these terms, beginning with the etymology of 'Artificial' and 'Intelligence'.

The term 'Artificial' traces its origins to the Latin word 'Artificium', derived from 'ars' meaning 'arts or skills', and 'facere' signifying 'to make or create'. When merged, they epitomize the notion of crafting or creating with skill. On the other hand, 'Intelligence' stems from 'Intelligentia', originating from 'inter' meaning 'between or among', and 'legere' connoting 'to choose or read'. Collectively, they encompass the capacity to understand and discern. the term "author," derived from 'auctor', which embodies the concept of a creator, someone who fashions something out of nothingness.


In her address, Ma'am delved into the profound implications of AI's learning capabilities, noting that AI is progressively absorbing and adapting to human commands and behaviors, even learning from our mistakes when they are pointed out. Drawing parallels with historical figures like Einstein and Ada Lovelace, she emphasized the potential for AI to be regarded as genuine authors, akin to their human counterparts. Moreover, she underscored the collaborative potential between humans and AI in creative endeavors, citing various online AI tools tailored for collaborative writing experiences.

Show casing various examples of poems and images, challenging the audience to discern between those created by humans and those generated by AI. She recommended a range of online AI tools such as Verse by Verse, ChatGPT, and AI Dungeon for generating poetry, prose, and images collaboratively.

 

One term that resonated strongly in her discourse was "Scriptor," as coined by Ronald Barths. According to Barths, a Scriptor possesses the ability to rearrange existing text in novel ways, a notion that aligns with AI's capacity to generate content based on learned patterns and inputs.

The session emphasized collaboration between human creativity and technology's capabilities, following insights from Salman Rushdie and Kevin Kelly. Rushdie noted AI's challenge to unoriginal writers, suggesting it doesn't threaten all human creativity. Kelly reframed the narrative from competition to collaboration, advocating a "race with robots" mindset. This approach encourages leveraging AI's strengths to enhance human innovation, fostering a synergistic relationship rather than a competitive one.


Dr. Shobha K.N. on AI for Teaching and Learning.
The speaker began by tracing the evolution of technology in education, highlighting milestones from the printing press to the World Wide Web. She emphasized that AI, despite its recent surge in popularity, has been under development for a significant period. The journey of technological advancement in education, starting from early computing to machine learning and Deep Learning, was illustrated. Drawing an analogy from sports, she illustrated how evolving technology offers educators new opportunities akin to athletes setting new records. Three exemplary educators—Benjamin Bloom with Bloom's Taxonomy, Sal Khan with Khan Academy, and Sam Altman, CEO of OpenAI—were cited for providing frameworks for AI-integrated teaching, spanning mathematics to literature. 



The speaker advocated for teacher training in AI through courses available on platforms like Coursera. Emphasizing the need for educators to familiarize themselves with AI, she recommended various AI tools for lesson planning, including Pictory, InvideoAI, Tome, SlideAI, Gamma, Eduaideai, Twee, Magicschool.Ai, Alayna.Ai, Preplexity, Curipod, and autoclassmate. Positioning AI as an opportunity that democratizes technology, she highlighted its potential for personalized teaching and learning, heralding a transformative shift in education.



Recorded Sessions


Thank you for visiting! I trust you found this information valuable and insightful.
at No comments:
Labels: , , , , ,

Sunday, 13 November 2022

Gian Course E- Literature- Learning Outcome (Part-2)

Electronic Literature and Artificial Intelligence (AI): Theory and Practice of Digital Storytelling

Recently I attended an Online course on the Gian platform on ‘Electronic Literature and Artificial Intelligence (AI): Theory and Practice of Digital Storytelling’ hosted by Prof. M. Rizwan Khan, The Department of English Aligarh University, Aligrah, U.P. in a virtual mode. This blog deals with my learning out of attending this interesting course.
Day-3

Lecture 5: Prof. Paola Carbone [Foreign faculty, Department of Humanities IULM University, Milan] on Locative Narratives: Definition of “Locative narrative” as a way to write with the Physical World, to read within the Physical World and give Place and History a voice, Examples
And How to project Locative Narratives. Locative Narratives and the Meta-verse

Locative narrative is a way to write within the physical world, to read within the physical world and to give place and history a voice”; “to write with place, object and absence as well as textuality.


Locative narratives are writing with space Flaneur, Street art/writing, Performance art and Land art.

Land art artists were interested in the combination of body, line, surface, site and materials and it opened up a perspective of experimenting with place and space through what Stiles described as ‘an amplification of the process over the product’– a shift from the representational object to further modes of action / presentation of experience.

geographical space = canvas

Immersive aesthetic experience
1. The place is de-familiarized in order to see differently
2. The genius loci becomes a framework for re-experience

Media artists started to explore the possibility of turning these principles into a digital artistic experience
Real world spaces are augmented with artistic contents – primarily audio and/ or textual – and mediated by mobile devices. E.G: Google maps

Locative mobile social networks – LMSN: to coordinate sociability in the city
location-based mobile games – LBMG
narration of places: enhancing the value of places through new technologies
site-specific fictional stories, stories written just for that particular environment

The user is set into a communitas, commonly referring either to an unstructured community in which persons are equal and are allowed to share a common experience, Communitas is characteristic of people experiencing liminality together, and more specifically in this case a space between organic and inorganic.
It offers to other users my sense and my knowledge of the place, my awareness of other people's behaviors, in other words my story.

The experience of public spaces especially in urban areas mainly consists of transit, a transition from one place to another... mostly a solitary experience.

Private space within public space

How can digital media draw one into an awareness of place?
narration of places: enhancing the value of places through new technologies
site-specific fictional stories, stories written just for that particular environment

Locative media is an instance of 'unframed' media practice, unframed in the sense of unbound from the desktop, detached from the singular screen and thus a fixed spectatorial perspective.

To read is to recognize that a critical engagement requires a range of cognitive and bodily activities, only one of which is reading in the sense of the visual processing of linguistic signs.

Reading involves seeing, moving, listening, touching = it is a challenge to the hegemony of words.


Examples

The design of a Locative Narrative
an idea on how to structure your contents, of what you want to say, and how you want to communicate it. This is called storytelling.
You need to activate your attention.
Before the scripting, the storyteller must :
1. survey the potential attractors distinguishing them between main and secondary >> they will become episodes of the core or satellite
2. select the myths, which will define the paradigms on which the narrative will be built visit the place
3 define the characters
4.choose the narrative typologies of the story: the choice will depend on:
the context: quantity and quality of the POI and eventual additional materials (documentation, archives, etc.)

users: if universal, it will have to refer to all typologies; if you have in mind a target, you have to adapt the choice to this, economic availability.

Vertical narration: a main plot must be identified without secondary subplots and a route that moves between the objective is the sensory immersion of the visitor in the narrated context, to be achieved through a detailed script of the movements of the user and the character.

Horizontal narration: once the main plot is established, the visitor has all the material at his disposal and can explore it as he likes and assemble it as he likes during the visit.

Example:


Lecture 6: Prof. Mohd. Rizwan Khan [Host Faculty, Department of English Aligarh Muslim University, Aligarh. U.P., India] on AI and the Discipline of Humanities

Humanities: The humanities include the study of all languages and literatures, the arts, history, and philosophy. It is a critique of human conditions.


AI education is valuable not only in the fields of Computer Science and Engineering but also in humanities. It will become an essential component in education like mathematics, language and Science. The English department has started a course in digital Humanities.
  • Literature
  • Electronic Literature
  • AI Generated Literature
  • Cyborg/ Robot Literature

Use of AI
Art historians
Historians
Archaeologists
AI in Art, Music, Dance
AI is seen as the sole generator or a collator.
Artwork can be generated.
AI song contest.
AI in film and media
Role of VFX, movie editing and creating trailers.
The film ‘Her’ depicts the story of a man who falls in love with a Virtual assistant.
‘Coded bias’- facial recognition treats dark-skinned faces unfairly’.


DAY-4
Lecture 7: Prof. Paola Carbone [Foreign faculty, Department of Humanities IULM University, Milan] on Podcast: What it is, Typologies and Examples and Discussion

Podcasts are an increasingly successful form of communication.

Three determining factors
  • The podcast is currently very fashionable and therefore the more people talk about it, the more they want to do it.
  • The on-demand characteristic of podcasts makes them suitable for multitasking listening, i.e. while doing other activities.
  • The podcast does not have interaction as its peculiarity and it is this 'lack' that makes it such an intimate and profound content.

Podcasting communicates in one direction:
  • you construct your content
  • you reason it out
  • you write it down
The word podcast comes from the combination of iPod and broadcasting.

It first appeared in an article published by Ben Hammersley for "The Guardian" on 12 February 2004.

In 2005, the New Oxford American Dictionary declared 'podcast' to be word of the year.


The term podcasts refers to original audio content, usually of an episodic serial nature, that is made available on demand over the internet.

RADIO VS PODCAST
  • Podcasts and radio are not the same thing. Not only are they based on two different technologies, but they also present two different types of content.
  • A podcast is not a web radio. Web radio is streamed by users through an internet connection. A podcast, on the other hand, can also be listened to offline, after being downloaded via an Internet network;
  • Radio is interactive, podcast is not
  • A difference and advantage of the podcast compared to radio is also the availability of time.
Intimacy: the radio tends to address, through the radio speaker, an indistinct mass of potential receivers while in the podcast you really have the feeling that the narrator is addressing the single person wearing his headphones.

Podcast range in time from 10 minutes to an hour long (for example for Crime Podcast)

CATEGORIES OF PODCASTS
  • Interviews + Panel Discussion
  • Free talk
  • Scripted fiction
  • Documentary + educational
  • Scripted non-fiction
  • News-recap

Example

Identify your podcast goals- To generate leads, To share an important message and To have fun
The only requirement is passion

Find two stories: same topic but with two possible endings or perspectives.
Find a story you like and then find another that somehow matches that story. This is because each episode of Shadow Lines is basically made up of two stories. Once the first story has been selected, the objective of the second story is clear. It has to be somehow similar to the first story (not only by similarity but also by contrast).

There are two types of stories:
  • Exceptional stories, out of the ordinary, able to arouse strong emotions and leave us stunned.
  • Ordinary and common stories, which precisely because of their "simplicity" allow us to identify with and relive, through the story, the events and themes that are somehow part of us.


DAY- 5
Lecture 8: Prof. Paola Carbone [Foreign faculty, Department of Humanities IULM University, Milan] on AI and Literature: Ontological Issues with examples

Ontology: the branch of metaphysics dealing with the nature of being.
In AI, an ontology is a specification of the meanings of the symbols in an information system.

Machine learning
Machine Learning is based on algorithms designed to perform a/one task.
Machine learning is able to find values not perceived by the human eye, but useful for making future predictions or behaviors based on algorithms..

Deep learning
Deep Learning works on a set of techniques that allows the system to automatically discover the representations needed for feature detection classification from raw data. Such representations are often hidden to our human comprehension. The data used by an AI to identify an image, for example, is very different from what humans would use. Through the neural network an image is analyzed and transformed into smaller representations (feature maps) that the computer can recognize so precisely to be used to identify content. The same features maps would appear as scribbles or random lines to the human eye.

Digital culture vs AI:
Digital culture defines flow of content (immaterial) distributed across various intersections of media human behaviors determined by virtual reality in its relationship with a real-world environment.

AI acts on the playground of reality since it acquires data from it in order to do things, to carry out actions, to perform tasks in the real world.

We should not consider AI as a tool (a hammer, for example) because an AI processes and interprets information. It is not even an environment (matrix) since it inhabits products and services that surround us (see internet of things).

AI vs Hyperreal:
AI emulates and determines (rational) behaviors rather than their simulation
“Hyperreal” simulates, it is a simulacrum or a “real” without origin or reality


Definition of AI:
(…) the artificial intelligence problem is taken to be that of making a machine behave in ways that would be called intelligent if a human were so behaving.” [1955, John McCarthy, Marvin Minsky, Nathaniel Rochester, and Claude Shannon]

Luciano Floridi: “ … were a human to behave in that way, that behavior would be called intelligent.”



Fears of AI:
  • Question of an advanced aligned AI system and human values.
  • A machine lacks those specific peculiarities of human intelligence that are generalization and abstraction
  • Is the AI aware of the fact that it is playing chess? No, it is simply applying an algorithm
  • Likewise, also emotions and doubts are essential human distinctiveness that a machine, trained to proceed algorithmically, cannot experience or understand.
  • Today research is trying to go farther and control emotions. Next-generation AI aims to capture these moments with webcams so as to adapt responses to emotions.
  • Machines must have the ability to understand emotions and to articulate responses in terms of both content and facial expressions, tone of voice, and body management = communication.

  1. Understanding AI means being aware of the risks
  2. transparency problems > who does what
  3. inequality, accountability problems: human bias processes on algorithms (exclusion of minorities)
  4. Manipulative problems echo chamber

This 'advanced' form of computer-assisted processing is still working on the idea that a screenplay, like a literary text, is a formal structure, but the neural networks allowed the generation of a text that is new in so far as it is a priori unpredictable and based on hidden learning features.

The text generated could only be written by that specific AI and from that particular dataset.

I hope this blog is useful. Thanks for visiting.
at No comments:
Labels: , , , , , , , , , , , ,
Subscribe to: Posts (Atom)

My Learnings from the National Workshop on Academic Writing (2026)

The National Workshop on Academic Writing  I recently had the opportunity to participate in the National Workshop on Academic Writing (2026)...