Applied AI Weekly News

Justin Grammens

Welcome readers to another issue of the Applied AI Weekly Newsletter! If you haven't already, I'd like to invite you to follow us on LinkedIn and Twitter

A quick update this week.

I posted this in the video section, but if you missed the Applied AI meetup, this past week, you for sure want to watch the video. Great presentation by Deborah Carver. Enjoy!

Be on the lookout for the next episode of the Conversations on Applied AI Podcast! We have now hit more than 10,000 Downloads and are gaining new subscribers every day. If you would like to speak with me on the podcast, reach out via our contact form.

Finally, I've got an upcoming event on September 14th at Club E where I'll be on a panel talking about how to navigate the AI landscape. Please register today and I look forward to seeing you there!

News

A Jargon-Free Explanation of How AI Large Language Models Work

Today, almost everyone has heard about LLMs, and tens of millions of people have tried them out. But not very many people understand how they work. This is probably one of THE BEST articles I've read to break down the internals of what's going on with Large Language Models (LLMs). Regardless of your experience (technical or not), I highly recommended you take the time to read this and take notes. Enjoy!

Man Beats Machine at Go in Human Victory Over AI

"In a surprising reversal of the 2016 computer victory that was seen as a milestone in the rise of artificial intelligence, an amateur player has comprehensively defeated a top-ranked AI system at the board game Go. Kellin Pelrine, an American player who is one level below the top amateur ranking, beat the machine by taking advantage of a previously unknown flaw that had been identified by another computer. This victory shows that AI still has flaws and that humans can still compete with machines at the highest levels of Go."

OpenAI Adds 'Huge Set' of ChatGPT Updates

OpenAI has made a lot of updates to its ChatGPT chatbot recently. They've added suggested prompts, multiple file uploads, and the ability to generate different creative text formats. These updates make ChatGPT more powerful and versatile, and they're likely to make it even more popular with developers and users.

I think the suggested prompts are a really helpful addition. They can help users get started with their conversations and prevent them from getting stuck. The multiple file uploads are also great for generating text from images, audio, and video. And the creative text formats are just plain cool. They open up a whole new world of possibilities for ChatGPT users.

I'm excited to see how developers and users use ChatGPT to create new and innovative applications. I think it has the potential to be a really powerful tool for a variety of tasks.

In the News: Manjeet Rege on How ChatGPT Could Help or Hurt Students

Manjeet Rege, software engineering and data science professor at the University of St. Thomas School of Engineering, recently spoke with The Chronicle of Higher Education about the uncertainty around how artificial intelligence may shape teaching and learning.

Sponsor

Leverage Artificial Intelligence to Bring Efficiency and Automation to Your Business

At Recursive Awesome, we help business leaders leverage Artificial Intelligence to bring efficiency and automation to their business. We are experts in not only strategy but implementation using a wide array of tools such as ChatGPT, Bard, Llama2, and many others from Microsoft, AWS, and Google. Contact us today!

Business

The ChatGPT Surge

ChatGPT is a new AI chatbot that is quickly gaining popularity. It is powered by OpenAI's GPT-3 language model, which makes it capable of generating human-quality text. ChatGPT is being used for a variety of tasks, including customer service, marketing, and creative writing. It is still under development, but it has the potential to be a powerful tool for a variety of applications.

Zoom Knots Itself a Legal Tangle Over Use of Customer Data for Training AI Models

Zoom is in hot water for its data mining practices. The company collects a lot of data about its users, including their meeting recordings, chat logs, and even their personal contact information. Zoom uses this data to train its AI models, but it's not clear if it's doing so in a compliant way with the GDPR and ePrivacy laws. This could lead to legal trouble for Zoom, and it could also erode user trust in the platform.

Frontier Model Forum

OpenAI, Google, Microsoft, and Anthropic have announced the formation of the Frontier Model Forum, a new industry body to promote the safe and responsible development of frontier AI models. The Forum aims to advance AI safety research, identify best practices and standards, and facilitate information sharing between policymakers and industry.

I think this is a great initiative. Frontier AI models are incredibly powerful, and it's important that we take steps to ensure that they are used responsibly. The Forum is a great way to bring together experts from different fields to work on this important issue.

Big Data

AI Algorithm Discovers 'Potentially Hazardous' Asteroid 600 Feet Wide in a 1st for Astronomy

An AI algorithm has discovered its first potentially hazardous asteroid, a 600-foot-wide (180 meters) space rock that will pass Earth within 140,000 miles (225,000 kilometers). The asteroid, named 2022 SF289, was found by the HelioLinc3D algorithm, which was developed by researchers at the University of Washington. The discovery is a major breakthrough for AI-powered asteroid detection, and it shows that these algorithms can be used to find potentially dangerous asteroids that would otherwise be missed by traditional methods.

Development

Nvidia AI Image Generator Fits on a Floppy Disk and Takes 4 Minutes to Train

Nvidia has developed a new AI image generator called Perfusion that is very efficient. It can be trained on a floppy disk in just 4 minutes, which is much faster than other AI image generators. Perfusion is also very accurate and can be fine-tuned to produce images that are more closely aligned with a user's text prompt. This makes Perfusion a powerful tool for creating personalized images, such as those used in marketing or advertising.

TinyML: The ‘Mini-Me’ of AI

Being a huge builder and supporter of TinyML for years, this article hit home. As the article summarizes, "TinyML could supercharge IoT, allowing IoT devices to take intelligent decisions in the field, literally. However, there is still the challenge of squeezing machine intelligence on the head of a pin, literally."

Events

Skills to Develop and Avoid When Building AI-Generated Content

The latest Applied AI Meetup hosted by Lab651 and sponsored by Recursive Awesome was truly informational. Take a look and contact Applied AI if you wish to sponsors, present or attend a future event!

Content generated from AI tools can run the gambit from high-quality and on-brand to poorly disguised copyright violations. Learn the most effective ways to build prompts, edit AI content, and ensure high-quality outputs from both text and image generators.

Our speaker this month is Deborah Carver. Deborah helps businesses ensure they are making sound investments in content and marketing technology. She also writes about demystifying publishing technologies and illuminating the gaps between the tech industry and the media. If it's SEO, UX, any kind of content recommendation algorithm, new tools, and workflows for content... you name it, she is figuring out how creatives can use and better understand technology.