About Us

Showing posts with label Aphelele Xulu. Show all posts
Showing posts with label Aphelele Xulu. Show all posts

Wednesday, 26 April 2023

Messiah or Monster: ChatGPT in Academia

                                      

 By Aphelele Xulu and Pravina Laljeeth


The perfect hack! Well almost! The Chatbot can do miracles! At least for students and learners in the classroom. No kidding! ChatGPT has been trained to harvest relevant data from the vast internet, datasets, websites, ebooks, scientific publications, scholarly databases, and social media, allowing it to analyse, calculate, synthesize, and collate language-generated diction that is coherent, contextually relevant, and often resemblant to the text produced by humans.   

ZDNet states that the AI (artificial intelligence) chatbot model ChatGPT is a conversational language model developed by OpenAI, based on the Transformer architecture and programmed on extensive amounts of text data that generate human-like responses, output, and solutions. The real game-changer!

The language model has been trained to receive data on an input prompt and output conversational data that is grammatically correct, much like having a conversation with humans. ChatGPT is trained to augment and generate rational responses that are syntax and semantically correct. It can decode, deconstruct, merge, synthesize, and generate results such as computer codes, webpages, essays, emails, songs, and stories in seconds. The intelligence model can manipulate texts from datasets and long sentence algorithms to generate paraphrasing at its finest.

According to Entrepreneur, ChatGPT is a transformer-based neural network that provides answers and data with human writing patterns. The AI model has been programmed with a limitless amount of text data to interpret and contextualize relevancy and generate human-like responses to questions. A neural network is a method in artificial intelligence that instructs computers to process data in a way that is inspired by the human brain. It is a type of machine learning process called deep learning that uses interconnected nodes or neurons in a layered structure that resembles the human brain. ChatGPT multitasks. The portal has more than one language function, so it can simultaneously juggle translation, summarize and answer questions, streaming longer, seamless conversations.

ChatGPT can generate intelligible and comprehensible text, allowing it to communicate in longer dialogues, conveyance prompts, and solution-driven transmission. The model has been trained on an immense amount of conversational data, and has learned how to analyse and generate grammatically correct text, uses appropriate vocabulary and tone, and is coherent with the input prompt and the overall conversation mechanism. ChatGPT, like other bot technologies, works well in mechanical, technical, and e-commerce settings but cannot transmute humans' critical thinking, cognitive and intellectual domains. It is not flawless and can deliver confined results.

 

A no-brainer, it will short-circuit solutions in almost every sphere and promises to pilot the shift in the cognitive trajectory of future generations. Organic, logical, and intellectual capacity suddenly looms as an endangered species. The concern must be focused on school-going children, learners, and students where this type of technology will progressively abate their natal cerebral compass, natural intrinsic trial and error to discovery logic, explorative and experimental gateways to new and undiscovered frontiers, reasoning capacity and problem-solving skills. ChatGPT will constrain inherent and intrinsic critical and analytical thinking.

 

Higher education institutions have questioned the effects ChatGPT will have on teaching and learning, curricula, research, scientific output quality, and academic integrity. ChatGPT is regarded as disruptive technology as it can be prompted to write an entire essay and acknowledge sources to reference from its knowledge base. To mitigate academic integrity, unpacking the strengths, weaknesses, opportunities, and threats must be examined by academia. This may lead to higher education policies and protocols regarding the use of ChatGPT. Further, the assessments policy must be revisited and reworked into curricula that may include oral assessments. In the future, AI tools may serve as research assistants, conducting virtual experiments, analysing data, copywriting and editing text, and generating citations.

ChatGPT can affect teachers, students, and researchers' innate academic excellence potential and can condition academia from producing authentic honest investigation, research output and diction to mediocrity.

Inherently, in academic libraries, there comes the ChatGPT dashboard. Chatbots are already used in many higher education institutions and academic libraries, offering 24/7 reference query-related assistance linked to library databases, resources, and discovery tools. Notably, advanced ChatGPT combined with advanced data extraction programs such as Covidence, Grammarly, and Endnote may produce a masterpiece that mimics human excellence in creating a document. Will Librarian's roles be redefined by ChatGPT? Will overall academia gradually become enslaved by the duality of AI and human intelligence? Is ChatGPT a paradox to human consciousness? Will man once again become Pavlovian, classically conditioned to new disruptive technologies? Is this the new paradigm shift of learning and discovery? Or have we become so clouded in the 5th industrial revolution with the rapid pace and expansion of AI technology that we are missing the master plan of monetizing disruptive technologies such as ChatGPT as a trade-off for human consciousness?

 

References:

1.     Open AI:  https://openai.com/blog/chatgpt

2.     ChatGPT: What Is It and How Does It Work?: 

https://www.entrepreneur.com/science-technology/chatgpt-what-is-it-and-how-does-it-work/445014

3.     What is ChatGPT and why does it matter? Here's what you need to know: https://www.zdnet.com/article/what-is-chatgpt-and-why-does-it-matter-heres-everything-you-need-to-know/

4.     Embrace it or reject it? Academics disagree about ChatGPT: https://www.universityworldnews.com/post.php?story=20230207160059558


T

Tuesday, 19 February 2019

Do’s and Don’ts of Data visualization


by Nokulunga Ziqubu and Aphelele Xulu

Data visualization is the practice of converting data from raw figures into a graphical representation such as graphs, maps, charts, and complex dashboards. Choosing the right visualization is essential, your data is only as good as your ability to understand and communicate it. The average person responds far better to visual information compared to just plain text. Whether you are buying a product or revising for an exam, visual stimulation over text translation allows the brain to consume the material with ease. Visually displayed data is easier to understand and analyse, making it faster for decision makers to find patterns, including new and hidden, and to understand even difficult concepts. Data visualization can be helpful in identifying issues and deficits, choosing the best product and business operation strategy, forecasting sales volumes and stock prices, fine-tuning project management and resource administration, and so on.



Bar charts are best used to show change over time, compare different categories, or compare parts of a whole.

References

AnyChart. (2018) What is data visualization? Definition, history, and examples. Accessed on 2018-12-11: https://hackernoon.com/what-is-data-visualization-definition-history-and-examples-e51ded6e444a
Crazy Egg. (n.d.) Potential data storytelling visuals. Accessed on 2018-12-05: https://twooctobers.com/wp-content/uploads/2018/03/CrazyEggVisual.png
HubSpot. (n.d.) 8 Dos and Don'ts for Creating Effective Infographics. Accessed on 2018-12-05: https://blog.hubspot.com/marketing/dos-and-donts-infographic-creation
HubSpot & Visage. (n.d.) Data visualization 101: How to design charts and graphs. Accessed on 2018-12-06: https://cdn2.hubspot.net/hub/53/file-863940581 pdf/Data_Visualization_101_How_to_Design_Charts_and_Graphs.pdf
Learnevents. (n.d.) Imagery vs text: Which does the brain prefer? Accessed on 2018-12-11: https://www.learnevents.com/blog/2015/09/07/imagery-vs-text-which-does-the-brain-prefer/
Severino R. (n.d.): Data visualization catalogue. Accessed on 2018-12-06: https://datavizcatalogue.com/
TwoOctobers. (n.d.) The 8 commandments of data storytelling. Accessed on 2018-12-06: https://twooctobers.com/blog/8-data-storytelling-concepts-with-examples/
Visme. (n.d.) The do’s and don’ts of chart making. Accessed on 2018-12-06: https://blog.visme.co/dos-and-donts-chart-making/
Zoss, A. (2014) Top ten dos and don’ts for charts and graphs. Accessed on 2018- 12-06: http://coalition.psesd.org/wp-content/uploads/2016/10/Mini-Session-Data-Visualization-Handout-2.pdf
Learnevents. Imagery vs text: which does the brain prefer? Accessed on 2018-12-11: https://www.learnevents.com/blog/2015/09/07/imagery-vs-text-which-does-the-brain-prefer/