ARTHUR C. CLARK - THE THREE LAWS

From "Hazards of Prophecy, The Failure of Imagination"

1. When a distinguished but elderly scientist states that something is possible, he is almost certainly right. When he states that something is impossible, he is very probably wrong.

2. The only way of discovering the limits of the possible is to venture a little way past them into the impossible.

3. Any sufficiently advanced technology is indistinguishable from magic.

"As three laws were good enough for Newton, I have modestly decided to stop there".  Arthur C. Clarke

Isaac Asimov's Corollary to Clarke's First Law: "When, however, the lay public rallies round an idea that is denounced by distinguished but elderly scientists and supports that idea with great fervor and emotion – the distinguished but elderly scientists are then, after all, probably right.

GENERATIVE PRE-TRAINED TRANSFORMER

The Generative Pre-Trained Transformer (GPT) system is able to transform one thing into another without losing the MEANING, just like you can!

Look at the following sentence:

• She poured water from the Pitcher to the Cup until IT was Full

What is IT referring to in the sentence?

Look at this sentence:

• She poured water from the Pitcher to the Cup until IT was Empty

What is IT referring to in the sentence?

Even though the sentence uses IT with a different meaning you do not get lost.

Through your Intelligence, you are able to Transform the meaning correctly.

The CHAT GPT system through its Artificial Intelligence is able to do exactly the same transformation. It is able to Transform the sentence in a variety of different ways and generate a variety of different outputs without losing the meaning.  This is the first time in human history that a machine could perform this trick.

ATTENTION IS ALL YOU NEED

The paper from Google that introduced a groundbreaking architecture known as the Transformer, as in Generative Pretrained Transformer or ChatGPT, which revolutionized the field of sequence modeling. This architecture relies heavily on the concept of self-attention, allowing it to capture dependencies between different positions in the input sequence. Watch the Video or read the paper here...

WHAT IS GENERATIVE AI & HOW DOES IT WORK?

Generative artificial intelligence is a relatively new form of AI that, unlike its predecessors, can create new content by extrapolating from its training data. Its extraordinary ability to produce human-like writing, images, audio, and video have captured the world’s imagination since the first generative AI consumer chatbot was released to the public in the fall of 2022.,, more information here

WHAT IS A TRANSFORMER IN AI & MACHINE LEARNING

A Transformer is a type of neural network architecture.

The key innovation of the transformer is the self-attention mechanism, which allows the model to attend to different parts of the input sequence when making predictions. This is in contrast to traditional recurrent neural networks (RNNs) which process the input sequence sequentially, and convolutional neural networks (CNNs) which apply filters across the entire input.

SYNTHETIC DATA: ANTHROPIC'S CONSTITUTIONAL AI

Also from fine-tuning to pretraining, OpenAI’s Superalignment, tips, types, and open examples. Read more...

VERIFY STEP by STEP - IMPROVING LANGUAGE MODEL REASONING WITH PROCESS SUPERVISION

This video provides a detailed analysis of the "Step by Step" research paper from OpenAI. The authors explore how large language models perform complex multi-step reasoning tasks and discuss how these models can be made more reliable. Two key methods are compared in the study: outcome supervision and process supervision. 

STREAMING LLM GIVES LANGUAGE MODELS UNLIMITED CONTEXT

Large language models (LLM) are renowned for their ability to process long text sequences. However, when dealing with lengthy articles, books, or prolonged chat sessions, these models often reach their context limit.  Current solutions to this problem are either computationally demanding, memory-intensive, or imprecise. 

A breakthrough solution is StreamingLLM, developed by a collaborative team of researchers ...more information here

ARTIST USE TECH WEAPONS AGAINST AI COPYCATS

Artists under siege by artificial intelligence (AI) that studies their work, then replicates their styles, have teamed with university researchers to stymy such copycat activity.

The artists turned to free software called Glaze created by researchers at the University of Chicago. Glaze essentially outthinks AI models when it comes to how they train, tweaking pixels in ways indiscernible by human viewers but which make a digitized piece of art appear dramatically different to AI. Read more here...

GOOGLE GEMINI

Google launches Gemini, an AI model capable of out performing humans in multitasking language comprehension. Gemini, is a multimodal artificial intelligence platform that can process and generate text, code, images, audio and video from different data sources. 

Gemini is not an application but a platform for bringing the latest artificial intelligence model to existing services, from Bard, Google’s chat competitor to ChatGPT, to the search engine, service managers, Android cell phones and large-scale data centers. more here... and here...

YOUR UNDIVIDED ATTENTION

In the podcast, Your Undivided Attention, co-hosts Tristan Harris and Aza Raskin explore the unprecedented power of emerging technologies: how they fit into our lives, and how they fit into a humane future.

Join us every other Thursday as we confront challenges and explore solutions with a wide range of thought leaders and change-makers — like Audrey Tang on digital democracy, neurotechnology with Nita Farahany, getting beyond dystopia with Yuval Noah Harari, and Esther Perel on Artificial Intimacy: the other AI. more here...

The Rabbit r1 - Personal AI Assistant

Unveiled at the Consumer Electronics Show (CES) 2024, the Rabbit r1 is a new form of personal AI device - for $199 https://www.rabbit.tech/ designed in collaboration with teenage engineering.

Lumiere

Lumiere is a text-to-video diffusion model designed for synthesizing videos that portray realistic, diverse and coherent motion -- a pivotal challenge in video synthesis. Lumiere introduces a Space-Time U-Net architecture that generates the entire temporal duration of the video at once, through a single pass in the model. https://lumiere-video.github.io/

Vector Databases - Pinecone

Dive into the world of vector databases with thisin-depth tutorial on Pinecone. Discover how to efficiently handle high-dimensional data, understand unstructured data, and harness the power of vector embeddings for AI-driven applications. https://www.datacamp.com/tutorial/mastering-vector-databases-with-pinecone-tutorial

LangChain explained - The hottest new Python framework

LangChain is a Python framework for developing applications powered by language models. This video takes a look at LangChain, to see how it works, and what you can do with it!. https://youtu.be/RoR4XJw8wIc?si=pthNKc4Q9htNh08j