ScholarSphere Newsletter #12

Welcome to 12th edition of ScholarSphere

“Embrace the new era; Progress and Tradition can thrive together. ”

Welcome to our AI Newsletter—your ultimate guide to the rapidly changing world of AI in academia. If you haven't joined us yet, now's your chance! Click that button, subscribe with your email, and get ready for an exciting journey through all things AI in the academic realm!

In today's search of AI, we'll see...

• Deep Dive into AI: Expand Your Knowledge

• Mastering AI: Prompt Perfection

• Cutting-Edge AI Insights for Academia

• Spotlight on AI Tools for Academic Excellence

• Academic Frontiers: Exploring AI Innovations

• Navigating Ethical Frontiers: AI and Ethics

• Engage and Learn: AI Workshops & Seminars

Deep Dive into AI: Expand Your Knowledge

AI Watermarking
By Lev Craig, Techtarget.com

What is AI watermarking?
AI watermarking embeds a hidden signal into AI-generated content, like text or images. This watermark identifies the content as AI-made and is invisible to the naked eye but detectable by AI scanners. Ideally, the watermark shouldn't hurt the model's performance, be easy to remove, or be compatible only with certain models.

How AI watermarking works
There are two stages: encoding during training and detection after generation. During training, the model learns to embed a watermark, like a special pattern in an image or rare words in text. After training, specialized algorithms look for these watermarks to check if a piece of media was AI-generated.

The benefits of AI watermarking
AI watermarking has advantages. Social media platforms can use it to show users that content is AI-generated. It can also indicate authorship, helping creators prove their work wasn't used deceptively. Additionally, AI watermarks act like digital signatures, showing where a piece of media came from. This can be useful for legal purposes.

The limitations of current AI watermarking techniques
Current techniques are not perfect. In one case, an AI text detector was taken down due to low accuracy. Watermarks can also be easy to remove, especially in text. There's also the risk of falsely identifying human-made content as AI-generated. Other limitations include watermarks only working for specific data and difficulty in balancing watermark visibility and model accuracy. Finally, widespread use might raise privacy concerns.

For reading the full article click here.
You can also find extra teaching articles in our LinkedIn Page.

Mastering AI: Prompt Perfection

Chain of Thought (CoT) & Zero Shot CoT
By Learnprompting.org

What is Zero Shot Chain of Thought Prompting?
Zero Shot Chain of Thought (Zero-shot-CoT) prompting is a follow up to CoT prompting, which introduces an incredibly simple zero shot prompt.

From this chain of thought, they are able to extract more accurate answers.

Technically, the full Zero-shot-CoT process involves two separate prompts/completions. In the below image, the top bubble on the left generates a chain of thought, while the top bubble on the right takes in the output from the first prompt (including the first prompt itself), and extracts the answer from the chain of thought. This second prompt is a self augmented prompt.

Zero Shot Chain of Thought Results 
Zero-shot-CoT was also effective in improving results on arithmetic, commonsense, and symbolic reasoning tasks. However, unsurprisingly, it was usually not as effective as CoT prompting. An important use case for Zero-shot-CoT is when obtaining few shot examples for CoT prompting is difficult.

Ablations of Interest
Kojima, et al. (2022) experiment with a number of different Zero-shot-CoT prompts (e.g. "Let’s solve this problem by splitting it into steps." or "Let’s think about this logically."), but they find that "Let's think step by step" is most effective for their chosen tasks.

Notes 
The extraction step often must be task specific, making Zero-Shot-CoT less generalizable than it appears at first. Anecdotally, I've found that Zero-shot-CoT style prompts are sometimes effective in improving the length of completions for generative tasks. For example, consider the standard prompt Write a story about a frog and a mushroom who become friends. Appending the words Let's think step by step. to the end of this prompt leads to a much longer completion.

Conclusion 
Zero shot chain of prompting, despite its simplicity, tends to improve model performance by including step-by-step reasoning in the response. It is encouraging that this technique can be used to solve complex tasks without the necessity of providing multiple input exemplars like in chain of thought prompting.

Full getting access to our Prompt Inventory check here
Don’t forget to visit our LinkedIn Page

Cutting-Edge AI Insights for Academia

ChinaDaily: AI a new challenge to academic integrity
Guest Post: AI Meets Academia—Navigating the New Terrain 
Openaccessgovernment: AI-empowered higher education: Challenges and opportunities
Harvard Gazette: What is ‘original scholarship’ in the age of AI?

Article of the Week: Navigating the confluence of artificial intelligence and education for sustainable development in the era of industry 4.0: Challenges, opportunities, and ethical dimensions by Ammar Abulibdeh, et al. (2024)

Spotlight on AI Tools for Academic Excellence

Bypassgpt.ai: A leading undetectable AI writer, trained to mimic professional human writers and create high-quality, error-free content that can get around any detector, like Turnitin, Originality.AI, ZeroGPT, and Copyleaks.
Eightify.app: An AI YouTube tool which helps when you're swamped with too much content. It’s an AI YouTube tool which finds the key points in topics like AI, Business, Finance, News, or Health.
Leap.ai: An AI platform that allows users to easily integrate AI into their applications. It provides pre-trained models for generating images, music, and more, as well as tools for building custom AI workflows and automations.
Openread.academy: Enhance research efficiency, accelerate scientific, technological, and humanitarian advancements, and reshape the landscape of knowledge discovery
Turnitin: Assessment to preventing plagiarism, We surface the most relevant matches for academic writing by comparing against the largest content database of best-in-class scholarly content, student papers, and current as well as archived web pages, built over our more than 20-year history.