🤝 Free Copilot email course for your professional & personal growth
Learn how to get started with generative AI to stay ahead in your career
Dear curious minds,
Maybe you wonder if you missed last week’s newsletter, but I have to tell you that it was the first week without a newsletter since I sent the first issue on April 27th in 2023. A big thanks to everyone who reached out and stated that they did not receive a newsletter, this really fulled me to continue this journey. However, it was never planned to stop writing this newsletter, but the week was just too busy, especially as I launched a free email course to get started with Microsoft Copilot.
In this issue:
💡 Shared Insight
Generative AI in the Workplace: The Need for Clear Company Policies
📰 AI Update
Google’s Open Gemma 2 Outshines Meta’s Llama 3
High-Quality AI Videos: Runway Gen-3 Alpha Takes the Lead
🌟 Media Recommendation
Free Copilot Course To Get Started with Generative AI
💡 Shared Insight
Generative AI in the Workplace: The Need for Clear Company Policies
One of the most exciting advancements in AI is the rise of generative AI. This type of AI can create new content, such as text, images, and even code, based on a set of instructions or data. This has opened up a world of possibilities for businesses, allowing them to automate tasks that were once considered too complex for machines. From automating repetitive tasks to generating creative content, AI is transforming the way we work and the skills we need for work.
A recent survey by Salesforce found that three out of five workers (61%) are currently using or planning to use generative AI at work. This number is expected to grow further in the coming years, as businesses increasingly recognize the value of AI-powered tools.
Despite its benefits, the widespread adoption of AI comes with its set of challenges. One major issue is the use of unapproved AI tools. In November 2023, a Salesforce article stated that a significant number of employees are integrating AI into their workflows without formal approval or guidelines, which can lead to data security risks.
Many companies are lacking in creating an AI policy. The launch of ChatGPT happened in November 2022 and shortly after it was more than obvious that this will change how we work. Nevertheless, many companies are still lacking a policy about how their employees can and cannot use generative AI. As a result, there are many people using generative AI at work without approval, as Ethan Mollick states in his article Detecting the Secret Cyborgs.
In conclusion, while the rise of generative AI presents transformative opportunities for businesses by automating complex tasks and enhancing creativity, it also introduces significant challenges, particularly regarding the use of unapproved AI tools. As the adoption of generative AI continues to grow, it is crucial for companies to establish clear policies and guidelines to manage its use effectively. This will not only mitigate data security risks but also ensure that employees can leverage AI responsibly.
📰 AI Update
Google’s Open Gemma 2 Outshines Meta’s Llama 3
After its announcement in May, which was covered in an earlier issue, Gemma 2 is now available globally in 9 billion (9B) and 27 billion (27B) parameter sizes. Both models perform really well compared to other models of similar and also larger size, as Google states in their release article.
More insights, including more detailed performance analysis are covered in the technical report.
As already stated in earlier issues of this newsletter, it is difficult to trust the well-known benchmarks, as the questions might be covered accidentally in the training material. The so far best alternative is the Chatbot arena. There, both versions of Gemma-2 outperform every other open model in blind side by side evaluations.
The models are released under the Gemma license, which so far shows the Apache 2.0 license. This means that they can be used commercially.
My take: Gemma 2 represents a significant step forward for open models. The 27B model runs nicely with high-end consumer GPUs. The quality for many responses is at GPT-4 level, but to name it once more: Gemma 2 can run locally on my own hardware. This is truly astonishing!
High-Quality AI Videos: Runway Gen-3 Alpha Takes the Lead
Nearly all of us are still waiting to get access to Sora from OpenAI, which was announced in February and covered in an earlier issue. The recent release from Runway Gen-3 Alpha can at least to some extent cover the gap between the existing video generation tools and the Sora announcement.
Users can create 5 or 10-second long videos in the resolution 1280x768 with 24 frames per second with Gen-3 Alpha. Currently, only text inputs are supported, and image inputs are announced to be available at a later stage.
The quality of the videos shared with the announcement is impressive. Especially humans are very realistic in their actions, gestures and the emotions they express.
Twitter is full of Gen-3 videos which prove that it really works, and the videos shared by Runway are not cherry-picked. The 𝕏 thread from Javi Lopez might be a good place to start.
For generating great videos, detailed descriptions including the scene, camera style and movement, lightning and further details should be described. To get an overview of what is tested and known to work, take a look at the Gen-3 Alpha Prompting Guide published by Runway.
Creating with Gen-3 Alpha is available to all users with a paid subscription plan, which currently starts at $12 per month. The Free Plan can't use Gen-3 Alpha at this time. Furthermore, it can so far only be used on the website and not in the app.
My take: The videos generated by Runway's Gen-3 Alpha look very impressive. The big advantage of Sora is that videos of up to 60 seconds can be generated, instead of only up to 10 seconds for Gen-3 Alpha. However, Sora is so far not more than an announcement for most of us, and Gen-3 can be used by any paying Runway subscriber.
🌟 Media Recommendation
Free Copilot Course To Get Started with Generative AI
As AI continues to transform the professional world, staying ahead of the curve is essential. To help you get started with generative AI, I've created a free Microsoft Copilot email course.
This course is designed to guide you through the basics of AI, providing practical insights and hands-on experience by spending daily 15 minutes. Over the course of two weeks, you’ll receive daily emails packed with insights and practical tips, empowering you to:
Get started with Microsoft Copilot
Modify texts in length, tone and complexity
Create summaries of webpages and PDFs
Create images and music
Don't miss out on the AI revolution. Sign up for the free Copilot email course today and start using generative AI to your advantage.
Disclaimer: This newsletter is written with the aid of AI. I use AI as an assistant to generate and optimize the text. However, the amount of AI used varies depending on the topic and the content. I always curate and edit the text myself to ensure quality and accuracy. The opinions and views expressed in this newsletter are my own and do not necessarily reflect those of the sources or the AI models.