🌹 Adobe Flirts with Sora to Rewrite Video Editing

Hello AI Enthusiast,

Adobe is gearing up to possibly be the first to roll out OpenAI's Sora in Premiere Pro, basically setting the stage to revolutionize video editing This move might give its pro users amazing new ways to edit videos.

Yet, in a world where AI can almost make us a cup of coffee, it’s pretty wild that 70% of our respondents in last week’s poll said their companies do not provide AI training. One comment said it all "My company is not even encouraging employees to use AI, and many of us are advanced AI users. The company should be talking to us as we know more than the executives." Clearly, there’s a gap between the rapid pace of AI innovation and its adoption in workplace training. As providers of AI training for companies, we understand that while bureaucracy can be a barrier, embracing AI is no longer optional but essential for staying competitive.

To help bridge this gap, we remind you that there are just a few hours left to grab the early bird discount for our new course, Leading Business Growth with Generative AI.

Now on with the news! 👇

News Bytes 🗞️

• Adobe is working on a new AI technology that can turn text into video, similar to OpenAI's Sora. They are paying artists about $3 per minute for video clips to help train this technology. At the same time, Adobe is adding other AI tools to Premiere Pro, like Pika, Sora, and Runway AI, to give users more options and improve their experience.

• 💡 Our take: Integrating third-party models is a popular strategy among companies looking to improve customer experience. This method allows them to use established technologies while they develop their own, just like Adobe is doing, or avoid investing money in building a model that might soon be outdated. Meanwhile, OpenAI focuses on distributing its models through partnerships over building a product to make a quick profit. At the end of the day, Adobe stays competitive, and OpenAI expands its technology's reach. This approach is just one of the strategies we teach in our Leading Business Growth with AI course.

• While the wearable Humane AI Pin is facing problems, marked by negative reviews similar to past AI flops, and struggles amid layoffs, there's a new contender in the wearable tech arena. The Limitless Pendant is a new device that wants to become the next big device we'll use. It aims to revolutionize meeting preparation with AI features such as audio recording and real-time notes.

• Intercom’s Fin AI Copilot uses AI to quickly generate responses for customer service queries, based on past conversations and various content sources. This AI assistant helps boost productivity by retrieving information from internal articles, public websites, and more, reducing the risk of generating misinformation.

• 💡 Our take: Intercom claims they are using AI in customer support in a careful way. They used their own data to train the AI, aiming to make it effective avoiding hallucinations. They’re probably trying to avoid the issues that companies like Klarna and an American car dealership faced with their chatbots, which led to problems that could have been avoided. A similar situation is seen with Devin AI, which promised to be the first AI software engineer but has been widely criticized for underperforming, while GitHub Copilot has been successful. These examples highlight the need to understand the limits of generative AI and how to best use it. This is a key concept in our Leading Business Growth with AI course, where we teach a framework to identify where AI can truly enhance business processes effectively.

• TikTok is exploring the introduction of AI-generated virtual influencers, capable of promoting products by reading scripts provided by advertisers. The feature is still in the development phase and may undergo changes before its official rollout.

• 💡 Our take: Avatars provide cost-effective and manageable advertising solutions, yet they sometimes miss the unique style and personal touch that human advertisers bring, which can be difficult to replicate. In a related application of AI technology, Alessandro, a talented student from our academy, will showcase in a webinar how he used generative AI to create an assistant without any coding. Register for the free webinar on April 30th at 6 PM CEST.

• OpenAI has launched an advanced version of GPT-4 Turbo, now accessible through an API and to ChatGPT Plus subscribers.

• OpenAI is expanding into Asia with a new office in Tokyo, Japan to collaborate with local entities and develop AI tools catered to Japan's needs.

• Republican lawmakers criticized the Biden administration as Huawei unveiled an AI-powered laptop using an Intel chip under a preexisting license.

• Google is developing its own Arm-based CPU called Axion and updating its TPU AI chips for more powerful AI acceleration tasks, positioning itself to reduce reliance on partners like Intel and Nvidia.

• Quora's AI platform, Poe, launches "multi-bot chat" allowing interactions with various AI models like GPT-4 and Claude in one conversation.

• Texas Education Agency (TEA) plans to save $15-$20 million annually with an AI-powered grading system for state exams.

• Therabot, an AI-driven mental health app, is undergoing its first clinical trial with 210 participants. The app uses generative AI to simulate therapy sessions, aiming to complement human therapists.

• DaVinci Resolve 19, revealed at NAB 2024, introduces new AI-powered features for advanced video and audio editing, including motion tracking and noise reduction.

• xAI, led by Elon Musk, unveils Grok-1.5V, a multimodal model performing well in various domains like reasoning and document understanding.

• Amazon has appointed Andrew Ng, a prominent AI expert with prior roles at Google and Baidu, to its board of directors as of April 9.

Educational Pill 💊

Understanding Multimodal Models

Let's break down what a "multimodal model" in AI means, taking a cue from xAI’s latest creation, Grok-1.5V.

Think of a multimodal model as an all-in-one AI tool. Unlike regular AI that might just understand text, these models can handle text, images, and even understand real scenes around us. Grok-1.5V, for example, can look at documents, photos, and diagrams, making it super handy for jobs that need more than just words.

So, remember, a multimodal model is like a multitasking AI, ready to take on different kinds of information all at once.

LOLgorithms 😂

Even with all the buzz around Devin AI, it seems developers are still very much in the game. Aren’t they?

From our community 🤝

Yesterday, we hosted an extra workshop for our Generative AI Project Bootcamp students, led by Kyle, a former student who taught how to build an AI automation from the ground up. We really enjoy letting our students take the lead and learn from one another.

Here's an image of Kyle, getting excited just a few hours before the workshop began.

That's a wrap on our newsletter! Here’s a quick recap before you go:

• Generative AI Project Bootcamp: Accelerate your processes and prototype AI business ideas with your own automated AI project.

• Leading Business Growth with Generative AI: Identify and develop AI strategies for your company’s growth.

• Startup and SME offers: If your team has 4 or more members, contact us to receive a group offer on our AI courses.

• Customized Corporate Training: Equip your team with tailored sessions designed for companies diving into AI.

Catch you next week! 👋