- OpenTools' Newsletter
- Posts
- 🔥Adobe's Firefly 3 Leaves Rivals Burned
🔥Adobe's Firefly 3 Leaves Rivals Burned
Reading time: 5 minutes
Have an AI Idea and need help building it?
When you know AI should be part of your business but aren’t sure how to implement your concept, talk to AE Studio.
Elite software creators collaborate with you to turn any AI/ML idea into a reality–from NLP and custom chatbots to automated reports and beyond.
AE Studio has worked with early stage startups and Fortune 500 companies, and we’re ready to partner with your team. Computer vision, blockchain, e2e product development, you name it, we want to hear about it.
Today we will discuss-
💥Microsoft debuts new small AI model—Phi-3-mini
👩🏻🎨All about Adobe's new Firefly model
😋Coca-Cola's oddly satisfying AI
⚙️9 amazing AI tools you might not have heard of
All this and more - Let's dive in!
👩🍳What’s cooking in the newsroom?
Microsoft launches new AI model that is "as capable as" GPT-3.5, while being small enough to run on smartphones
🕵🏻♀️News - Microsoft has launched a new small language artificial intelligence model called Phi-3-mini.
Phi-3-mini is the first of three small models that Microsoft plans to release. It follows the release of Phi-2 in December, which performed at par with larger models like Llama 2.
According to Microsoft, Phi-3-mini outperforms its predecessor and can provide responses similar to those of a model ten times its size. Eric Boyd, corporate vice president of Microsoft Azure AI Platform, stated that Phi-3-mini is as capable as larger language models like GPT-3.5, despite its smaller size.
📲For context - Small AI models are specifically designed for simpler tasks. They are not only cheaper to run compared to larger models but also perform better on personal devices like phones and laptops, making them ideal for companies with limited resources.
💼Is Phi-3 for you? Phi-3 can help businesses by summarizing lengthy documents, extracting relevant insights and industry trends from market research reports, as well as performing tasks such as content creation and generating social media posts.
If you'd like to give it a try, starting today, the model will be available on Microsoft Azure, Hugging Face, and Ollama.
Just curious...Have you ever experimented with Microsoft's previous iterations: Phi-1, Phi-2, or other small AI models such as Google's Gemma 2B or Anthropic's Claude 3 Haiku? |
Adobe unveils its most advanced AI image generation model yet
☕News - Although Adobe's family of generative AI models, Firefly, hasn't received the best feedback from creatives, Adobe is aiming to improve its reputation with the release of its third-generation model, Firefly Image 3, at this week's Max London conference.
✨What's new - While Adobe is adding several new generative AI tools in its Photoshop software to provide users with more control over their designs, the most notable addition is Reference Image. This feature allows users to upload images to inspire Adobe’s AI, matching similar elements in style and color. For example, instead of repeatedly tweaking a prompt description like “a blue vintage truck with flower decals,” users can provide a reference image for Photoshop to use as a guide.
This feature could give Adobe an advantage over rival AI image generation models. However, it comes with a caveat—users must have the rights to use the images they reference. A notification will alert users to this ownership requirement when they first use the tool.
😄What's more? The latest model, now available in Photoshop (beta) and Adobe’s Firefly web app, produces more realistic imagery than its predecessors (Image 2 and Image 1). This is thanks to its ability to understand longer, more complex prompts and scenes, as well as improved lighting and text generation capabilities.
🤓Our review - For what it's worth, Image 3 does appear to be a step up from its previous versions. However, it's important to note that this assessment is based on a few outputs shared by Adobe, so it's possible that they don't give a clear picture (pun intended) of the model's performance across various scenarios. Nonetheless, here's an example of an image generated by Firefly 3:
Prompted to create a hyper-realistic eye inspired by cordyceps mushrooms.
Which AI image generator do you prefer to use? |
Coca-Cola launches an AI-powered instrument to capture the sounds of Cola
🥤News - As part of its latest marketing effort, Coca-Cola has launched Coke SoundZ, a musical instrument that allows users to create music mimicking the sounds of opening and sipping a Coke.
According to a press release from the brand, the sounds generated by the app - "phsst," "fizz," "clink," "glug," "ahhh" - are aimed at simulating the sensory experience of enjoying a Coca-Cola product. Simply put, it's likely to trigger a satisfying feeling in our brains.
🤤Still curious? Here's more: The app – available for free on the App Store and Google Play – includes five toggles representing each of the cola-derived sounds, positioned above an "AI generate" button, indicating intentional mixing of these sounds through an algorithm.
In addition to the digital instrument app, Coca-Cola has launched a physical version of Coke SoundZ, shaped like a glass bottle of Coke Zero. It is being distributed to select DJs, music producers, and creators to inspire them to create their own tracks.
Similar to the app, the physical instrument offers a variety of additional sound settings, allowing it to produce various synth tones and providing musicians with a wider range of creative options.
🤭All about the experience? Well, this most definitely isn't the best use of AI, but after writing this, I'm craving a Coke. And if you're reaching for one too, then hey, at least their campaign worked!
🙆🏻♀️What else is happening?
Drake uses AI to mimic 2Pac and Snoop Dogg on new Kendrick Lamar diss track “Taylor Made Freestyle”
China acquired recently banned Nvidia chips in Super Micro, Dell servers, tenders show
Apple acquires French startup behind AI and computer vision technology
Chatbots providing 'unintentional' misinformation ahead of EU elections
👩🏻💻Prompt engineering tips
Multi-question prompts - When crafting prompts for ChatGPT, pose multiple interconnected questions within a single prompt to guide the model's response in a specific direction. Being interconnected, these questions help the model talk about different parts of the topic in a well-organized and detailed way.
Ask for comparisons - When crafting prompts for any chatbot, ask the model to compare different aspects of a topic such as similarities and differences, strengths and weaknesses, benefits and drawbacks. This approach helps get detailed insights.
Refine length - The length of your prompt could play a vital role in a chatbot's response. The prompt needs to be effective and to the point. It is also very important to strike a balance between brevity and providing enough context. Think of it this way - too many ingredients can overpower a dish, while too few might leave it lacking.
👩🏼🚒Discover mind-blowing AI tools
IMAI - An influencer marketing platform that empowers brands to discover and manage influencers ($395/month)
Avian.io - A data automation platform for marketing data, integrating seamlessly with Google Sheets, Looker Studio, or Data Warehouses ($79/month)
OpenLead - An AI-powered lead generation tool that helps businesses generate custom lead lists based on specific criteria ($99/month)
Hotjar AI - A tool that helps website owners understand user behavior and optimize their websites ($32/ month)
Piktochart AI - An infographic generator that uses artificial intelligence to quickly create high-quality infographics ($14/month)
AIPRM - A prompt management tool and community-driven prompt library ($10/month)
Portaly - Allows users to create a customizable link-in-bio landing page for their website (Free)
Tailwind - A marketing tool for social media automation and optimization ($14.99)
SurgeGraph - An SEO toolset that uses 50+ SERP data points to help generate SEO-optimized content ($48.99 /month)
😯Unusual AI use cases
Aquarium monitoring - AI has been used to transform aquarium management by employing cameras and sensors to monitor conditions. These systems autonomously adjust parameters, such as lighting and water temperature, ensuring an optimal environment for marine life and simplifying the task of aquarium maintenance.
Waste management - Robotics equipped with AI vision systems can identify and sort recyclables from general waste, significantly enhancing recycling efficiency. This innovative approach ensures more accurate sorting, reducing environmental impact and promoting sustainable waste practices.
Prosthetics - AI has also been used to develop prosthetic devices that adapt to users' movements quickly. Thus, providing more natural and intuitive functionality for individuals with limb loss. These devices analyze user input in real-time, adjusting functionalities seamlessly, promoting mobility, comfort, and a higher quality of life for individuals overcoming physical challenges.
Can you help us, please?Be our mentor and let us know your take on today's newsletter. Share your thoughts on which AI tools you would like to see featured in the OpenTools newsletter or just drop a Hi. YES! We personally read all the messages. |