Will AI become too powerful? 😳

Hey there,

The OpenAI saga is finally coming to a conclusion and it seems we will know everything soon. With Sam getting back to OpenAI, a new board has been reinstated without Ilya Sutskever (the one board member that played a big rol in getting Sam being fired). Sam released a statement saying an independent investigation is going on to find out exactly what happened. We will be waiting for that one…

Elon Musk also commented on AI during his interesting interview with CNBC claiming the following when asked about lawsuits regarding copyright and datatraining:

Table of contents

• 🔱 What does a God like AI look like? 🤖 

• OpenAI competitor Anthropic is making use of the OpenAI chaos 🌋

• Amazon joining the AI race 🏎️

• Talk to the internet like it is a person 🤓

• 🏔️ AI creating materials? 🏔️

Updates

🔱 What is Q* and is it our new God? 🤖 

Q* is the AI breakthrough that almost killed OpenAI. It was frightening enough for lead open AI researchers to write a letter of concern to the board which likely precipitated the firing of Sam Alman. But here's the thing only a handful of people within open AI know exactly what Q* is.

People on the internet have been speculating what it might possibly be. It is speculated to be a combination of two types of algorithms, one of which is Q-learning, an algorithm for reinforcement learning. This article goes in dept on the possible things Q* might be. Be warned it is a bit complex…

If Q* is what the internet thinks it is, it would be a giant leap towards AGI, it might even be AGI. AGI is short for artificial general intelligence, which refers to the concept of a highly autonomous system that outperforms humans at most economically valuable work.

If that’s the case, Elon Musk is right and we will have a digital God…

Competitor Anthropic making use of the OpenAI chaos 🌋

Anthropic released Claude 2.1, their latest model. The new model comes with several advancements, including an industry-leading 200K token context window, significant reductions in rates of model hallucination, and a new beta feature called tool use. The 200K context window allows users to upload large bodies of content or data, enabling Claude to perform tasks such as summarization, Q&A, trend forecasting, and document comparison. The model has also seen a 2x decrease in hallucination rates and has made significant gains in honesty, with a 2x decrease in false statements compared to the previous model. The new tool use feature allows Claude to integrate with users' existing processes, products, and APIs. The release also includes improvements in developer experience, with a new Workbench product and system prompts. Claude 2.1 is available in the company's API and is powering their chat interface at claude.ai for both free and Pro tiers. The usage of the 200K token context window is reserved for Claude Pro users.

Amazon joining the AI race 🏎️

Amazon's cloud business AWS is launching a chat tool called Amazon Q, which acts as an AI assistant allowing users to ask questions specific to their companies using their data. The tool was announced during a keynote speech by AWS CEO Adam Selipsky at AWS re:Invent. Amazon Q can work with any of the models found on Amazon Bedrock, AWS’s repository of AI models, and was trained on 17 years’ worth of AWS knowledge. It is currently available only for users of Amazon Connect, AWS’s service for contact centers, and will eventually be available on other services like Amazon Supply Chain and Amazon QuickSight.

Pricing for Amazon Q in Connect starts at $40 per agent per month. According to AWS’s Connect website, users can try Amazon Q in Connect “for no charge until March 1, 2024.” It also has guardrails for Bedrock users to ensure their applications and the models they use to power them follow their data privacy and responsible AI standards.

Salesforce and Amazon Web Services (AWS) have announced a “significant” AI-focused expansion of their partnership.

The new agreement, announced Monday (Nov. 27), makes it easier for customers to manage their data across Salesforce and AWS, and infuse the latest generative artificial intelligence (AI) technologies into their applications and workflows. 

🛠️ Tool of the week 🗓️

Perplexity AI is an advanced question-answering tool designed to provide informative and nuanced responses to a wide range of queries. It combines cutting-edge AI models with sophisticated search capabilities, offering users detailed, accurate, and context-aware answers across various topics.

This tool will drastically enhance your experience in information-seeking tasks. Give it a try here.

More updates

🏔️ AI creating materials? 🏔️

A new deep learning tool discovered millions of new materials with deep learning. The use of a new deep learning tool called Graph Networks for Materials Exploration (GNoME) can predict the stability of new materials. It discoverer 2.2 million new crystals, including 380,000 stable materials that could power future technologies.

The use of this AI-guided guided approach can accurately predict materials that could be experimentally viable. In doing so to tool predicted structures for 380,000 materials that can be used by the research community. This technology can generate “recipes” for materials that will be essential for developing greener technologies and the potential for AI to revolutionize materials discovery, experimentation, and synthesis.