🪖 OpenAI Revises Military Policy

Hello AI Enthusiast,

Soon we’ll launch the sales of the fifth edition of our Master in Prompt Engineering course, set to commence this March.

This time, we're giving a special treat to those who sign up on our waiting list. You'll not only get to join the course early but also get a special invite to a webinar by Gianluca Mauro. He'll show how we, at AI Academy, have integrated AI into our internal tools without writing a single line of code.

Two weeks ago in this newsletter, we asked you which vertical we should focus on for our new course. Thanks to your feedback, we're developing our first-ever Master in Generative AI for Marketing, specifically designed for marketers. To validate our course format, we're conducting customer discovery interviews. If you have a marketing-related role and are intrigued by this program, click here for an interview opportunity with our team. As a token of appreciation, participants will receive a special discount for the first edition of the program.

Now, let's dive into this week’s AI updates. 👇

News Bytes 🗞️

• OpenAI has updated its usage policy, removing the prohibition on the military and warfare application of its technologies. This change suggests that OpenAI is now open to serving military customers, raising questions about the company's stance on responsible AI development.

• Last week OpenAI finally launched its GPT store, allowing users to publicly share their chatbots, with over 3 million created so far. They have also introduced ChatGPT Team, a new plan aimed at smaller businesses, offering securely published GPTs for collaborative workspaces.

• Google is investing in initiatives to protect the integrity of elections by preventing abuse and improving access to accurate voting information. With two major elections coming up this year, the EU and US elections, they are refining their AI tools to anticipate and prevent potential abuse, enhancing transparency around image provenance, and integrating ChatGPT with real-time news reporting. Will it work this time?

• The Rabbit R1, a new virtual assistant device, has been launched by startup Rabbit. It features a distinctive design, is operated through a "Push-to-Talk" button, and uses a Large Action Model¹ (LAM) to carry out tasks on interfaces. The device is available for pre-order at $199 with expected shipping in March or April. Is this the beginning of hardware designed specifically for AI use?

• Amazon and BMW are collaborating to integrate a cloud-based voice assistant language model (LLM) into BMW cars, providing users with the ability to ask intricate questions about their vehicles, dependent on always-on connectivity and 5G for certain features. This initiative aligns with Amazon's 2022 integration of Alexa into BMWs, capitalizing on the growing generative AI trend.

• Volkswagen has announced that the ChatGPT chatbot will be integrated into its IDA voice assistant, making it available in VW models with electric and ICE drive. The inclusion of artificial intelligence in their vehicles shows Volkswagen's commitment to incorporating advanced technology into their product lineup.

• Microsoft has released Copilot Pro, the newest version of its advanced and fast coding tool, for $20 per month per user. It offers features like access to Copilot across devices, priority access to the latest models, and the ability to create customized Copilot GPTs.

• The US government has approved BrainSee, an AI-based software that predicts memory loss progression using clinical brain MRIs and cognitive tests, potentially enabling early treatment for some patients. San Francisco-based company Darmiyan states that BrainSee shifts the patient experience from anxiety to proactive management and could have a significant economic impact by reducing healthcare costs associated with Alzheimer's care.

• Getty Images has launched a new service called Generative AI by iStock, powered by Nvidia's technology, which aims to guard against copyright infringement by generating new and modified images. The service offers affordable and legally protected AI options for every licensed visual generated. This follows Sony, Nikon, and Canon's initiative to embed digital signatures in photos, aiming to distinguish them from increasingly sophisticated fakes.

This week’s glossary 📖

1. Large Action Model: An AI model specifically designed to understand users’ intentions and take actions by leveraging other services. For instance: if Rabbit understands you want to generate an image, it would leverage Midjourney for that task.

Educational Pill 💊

Understanding OpenAI's Policy Shift

Recently, OpenAI updated its usage policy, removing the explicit prohibition on military and warfare applications of its technologies. OpenAI clarified that while they still prohibit the use of their tools for causing harm, developing weapons, surveillance, or destruction, they are now open to national security uses that align with their mission. An example cited is their collaboration with DARPA (a branch of the U.S. Department of Defense) to develop new cybersecurity tools for securing vital open-source software.

Previously, the policy featured a bulleted list of specific prohibitions, which has now been replaced with more general guidelines. OpenAI maintains a blanket prohibition on developing and using weapons but acknowledges that the military's scope goes beyond weaponry. The update opens avenues for OpenAI to explore new business opportunities with military establishments, particularly in areas unrelated to direct warfare. This shift raises important questions about the role and responsibilities of AI companies in military collaborations.

From our Podcast 🎧

Check out our latest podcast episode of "I'm Not a Robot"! We chat with Nermeen from LEGO Group. She's super passionate about making AI fair and diverse.

In this episode, we talk about why it's important to think about who AI might leave out and how to make it right for everyone. Nermeen shares her story and some great tips for companies who want to do good things with AI.

From our Community 🤝

Did you know our students get access to our AI community? Here’s Gianluca’s answer to a student's concern about the applicability of their current learning about ChatGPT in the context of newer AI models like GEMINI, and the worry that these new models might render ChatGPT obsolete.

Join our program and enter a room full of bright people.

LOLgorithms 😂

Raise your hand if you've never used ChatGPT to look smarter. 🤫

That is the end of our newsletter.

Remember, if your company is looking to implement AI technologies, we also offer customized corporate training.