In my latest article, I've delved into the intersection of artificial intelligence and education, examining how Large Language Models (LLMs) such as the GPT series can elevate personalised learning. Drawing on my experience in the field of AI and machine learning and as a parent, I explore how these powerful tools can tailor the educational experience to align with each student's unique learning styles, pace, and strengths. The potential of LLMs in education is vast. From generating human-like text, responding to complex queries, to crafting insightful essays or even poetry, these models can effectively act as digital tutors, offering customised educational resources based on individual student preferences. Yet, this advancement isn't without its challenges. Ensuring accurate information, understanding the nuances of text comprehension, and navigating issues related to privacy and technology access are all critical considerations. Nonetheless, with thoughtful implementation, LLMs could make education more engaging and inclusive. They hold the promise of narrowing the educational achievement gap and empowering students to reach their full potential. The intersection of AI and education heralds an exciting future, one that necessitates careful stewardship and human oversight. I encourage you to read the full article here: hamsa.substack.com/p/the-daw…. I'm eager to hear your thoughts on the potential and challenges of incorporating LLMs into education. #LLM #ChatGPT

Shooting visual effects with a disposable camera and Gen-3 Alpha.

THIS IS AI!! Ok it’s getting crazy! This was made using @HeyGen_Official

Realtime emotion detection with #AI is an impressive technological advancement, but can it truly capture the complexity of human emotions? What do you think? Credit: Rondinelli Morais #innovation #artificialintelligence #machinelearning

A society that has swapped human judgement for the cold precision of AI is ill-equipped to grapple with any disruption in the system. A single glitch could send tremors through the brittle skeleton of such a society, leaving it floundering in the face of adversity… #LLMs #ArtificialIntelligence hamsa.substack.com/p/the-ech…

“Every day, we software engineers face a multitude of decisions: which architectural pattern to use, how to optimise an algorithm, or which reported bug is most critical. Often, our intuition, our 'System 1', proposes an immediate answer. But, are these quick solutions always reliable?” #SoftwareEngineering open.substack.com/pub/hamsa/…

A Cautionary Tale for the Digital Age, Inspired by E.M. Forster's 'The Machine Stops "Like an intricate network of unseen veins, these digital entities have woven themselves into our very existence, extending their influence into the farthest corners of our daily routines" Pause. Ponder. Are we trading our uniqueness for AI's convenience? Consider how deep #LLMs imprint on our thoughts. Can we stay human in an AI-shaped world? Let's navigate this era with thoughtfulness, cherishing our individuality while riding the AI wave. Question. Engage. Preserve creativity. The balance is in our hands! 👉Blog Post: hamsa.substack.com/p/the-ech…. Are we losing ourselves to the comforts of AI? Do Large Language Models homogenise our thoughts? #AI #ChatGPT4

Latest research compares #LanguageModels like GPT-3, GPT-4, and Bing’s Prometheus with RetA LLM for biomedical Q&A. RetA, with domain-specific data, shows greater accuracy & relevance, yet readability remains a challenge. Proves once again, one size does not fit all in #AI! More studies needed to harness full potential of LLMs in healthcare. #GPT4 arxiv.org/pdf/2305.17116.pdf

AI-powered knowledge sharing platform. #AI #ChatGPT4 parabrain.ai/

GPT4All LocalDocs offers a feature to engage in conversations with your personal data! - Simply drop files into a designated folder, and GPT4All will utilise them as reference context for providing responses. - It is compatible with over 40 different file types. - It provides the added benefit of citing sources in its responses. #AI #GPT4 gpt4all.io

"Japan’s government recently reaffirmed that it will not enforce copyrights on data used in AI training" #AI #GPT4 technomancers.ai/japan-goes-…

Based on a recent social experiment carried out by AI21 Labs, approximately one-third of the population cannot differentiate between a human and artificial intelligence? #AI #GPT4 venturebeat.com/ai/one-third…

mashable.com/article/chatgpt… #ChatGPT

Could it be that by assigning the term 'intelligence' to these machines, we are, in fact, doing them a disservice? #LLM #ArtificialIntelligence open.substack.com/pub/hamsa/…

A new model called Med-PaLM 2 is redefining AI's potential in medicine. It surpasses physicians' scores on medical question answering, achieving 86.5% accuracy on the MedQA dataset and outperforming previous models by 19%. Human evaluations show doctors prefer Med-PaLM 2's answers on multiple clinical utility aspects. #ArtificialIntelligence #LLMs arxiv.org/abs/2305.09617

We co-hosted an AI hackathon with @southpkcommons and asked our communities to build something great on Replit. In just 5 days, 17 teams wrote production-ready code on Replit and submitted demos of their final projects. Let’s dig into the winners and what features they used 👇

"Like an intricate network of unseen veins, these digital entities have woven themselves into our very existence, extending their influence into the farthest corners of our daily routines..." Read more here: hamsa.substack.com/p/the-ech… #ArtificialIntelligence #LLMs

"Mojo is a statically-typed language that, while extending Python, revolutionises our concept of compile-time meta-programming." #MOJO hamsa.substack.com/p/discove…