​'FRIENDS,' KIDS EDITION (10 IMAGES)​

​

APRIL 22, 2024

🤖 Innovation Dispatch: Your Daily Dose of AI

Read Time: 6 Minutes

Welcome back, Innovators!

Meta's Llama 3 is available now, for free; Zuck says what is more valuable to AI than data; AI to improve by understanding our irrational side; FRIENDS, Kids Edition (10 images); plus, more headlines and tools!

🎯 Today's Innovations

• Meta's Llama 3 is available now, and free (Link below)
• QUESTION OF THE DAY
• AI Must Learn from Our Irrational Side to Get Better at What It Does
• Zuck says this is more valuable to AI than data; more headlines for April 22, 2024
• AI Word of the Day: Principal Component Analysis (PCA)
• 5 New AI Tools for Work and Everyday Life

FEATURE

Meta's Llama 3 Can Generate Images. It's a Bit Stubborn Though (See, and Try, for Yourself)

🤖

​

​Q: Will increased AI functionality in our phones kill the staying power of any current or future separate AI gadgets/hardware?​​

TIDBIT

AI Must Learn from Our Irrational Side to Get Better at What It Does​​

Researchers have developed a new technique to model the suboptimal, irrational decision-making behavior of humans and AI agents working towards unknown goals. The method can automatically infer an agent's computational constraints by observing just a few traces of their previous actions to determine their "inference budget" - how much time/effort they spend planning before acting. This inferred inference budget can then predict the agent's future behavior on similar problems.

Tested on tasks like navigation goal inference and predicting chess moves, the technique matched or outperformed existing methods for modeling bounded rationality. The interpretable inference budget correlates with factors like task difficulty and player skill.

In a nutshell, this research wants to create smarter AI assistants that can work well with people. The AI will understand when humans make mistakes or poor choices because of limited thinking abilities. Knowing why people act irrationally at times will allow the AI to help humans avoid those errors.

If the AI sees someone is about to take the wrong route based on their previous bad navigation choices, it could provide better directions. Understanding human thinking limits will make AI assistants way more helpful to work with.

​CHECK OUT MORE HERE FROM MIT​

📰 Headlines

1. ​Google Consolidates AI-Focused DeepMind Research Teams​
2. ​The future of AI gadgets is just phones​​
3. ​An AI Salvador Dalí will answer any question when called on his famous 'lobster phone'​​
4. ​How United Airlines Uses AI to Make Flying the Friendly Skies a Bit Easier​
5. ​A coffee roastery in Finland has launched an AI-generated blend. The results were surprising​​
6. ​AI Tools for Patent Drafting: LLMs Will Likely Never Write Claims as Well as Humans​​
7. ​As most AI execs scramble for more data, Mark Zuckerberg says there's actually something more 'valuable'​​
8. ​People Are Freaking Out Over This AI-Generated Video of the Mona Lisa Rapping​
9. ​iOS 18 to Integrate All AI Features On-Device, According to Mark Gurman​
10. ​Sex Offender Banned From Using AI Tools in Landmark UK Case​

AI Word of the Day: Principal Component Analysis (PCA)

Principal Component Analysis (PCA) is a statistical technique used to simplify complex data sets.

It works by reducing the number of variables used to describe a large set of data, focusing instead on a few principal variables that capture the most important information. This process is known as dimensionality reduction.

Why It Matters

PCA is crucial for making data analysis manageable and insightful. In scenarios where data involves many variables, PCA helps to uncover patterns and trends that are not immediately obvious. It reduces the complexity of data, enabling easier visualization and faster computational processing.

This is especially valuable in fields like finance, where PCA can help identify the main factors influencing stock movements, or in bioinformatics, where it's used to find patterns in genetic data.

Real-World Example

In customer segmentation, a business might collect extensive data on customer behavior, preferences, and demographics. Using PCA, the company can reduce these numerous variables to a few principal components that most significantly define customer groups.

This distilled view can highlight underlying factors that influence purchasing behaviors, enabling the business to tailor marketing strategies more effectively. By focusing on these principal components, companies can more efficiently allocate resources and craft targeted messages, improving customer engagement and boosting sales.

📚 AI TOOLS

• ​Faraway: Build Movies with AI is an AI movie creation app that allows users to generate images and control the composition, style, and characters. Users can also add voices to their creations. The app offers a simple and powerful timeline interface. Faraway is currently available for iPhone.
• ​Tiered is a career coaching app that offers users a Jump or Stay score to help them decide if they should stay in their current job or look for new opportunities. The app also provides users with their rank compared to others in their age, industry, and location. Users can also see a breakdown of their scores across five key factors: Engagement, Compensation, Marketability, Company, and Industry.
• ​Gulf Picasso is an AI image-generation tool that allows users to create images from text descriptions.
• ​Slidey is an easy presentation maker app for students on Android. It allows users to create presentations quickly and easily, even offline. Slidey offers pre-made templates and themes, one-click slide design, and the ability to add text, pictures, graphics, and tables.
• ​AI-Ebooks is a company that offers a new AI tool that can generate custom ebooks. The tool can be used to learn about a variety of topics, including history, ethics, and artificial intelligence. The article also mentions some of the company's other products, such as an AI-powered ebook generator and a tool that can help users navigate ethical dilemmas.

​Read All Innovation Dispatch Back Issues Here​

Your input fuels the content. Let me know how I did today!

​Email me your thoughts, or just tell me whether you thought it was:

Great ❤️
Average 😐
Sucked 😞

Also, let me know what you would like to see more of. Have a great rest of your weekend!