Meta’s ImageBind makes AI more human

PLUS: Can Students Pass Using AI Tools At University?

Good morning! Today we're covering Meta’s announcement of ImageBind AI, which mimics human perception for connecting diverse data types, how universities are addressing challenges presented by AI tools like ChatGPT, balancing tech adoption with misinformation concerns and the Start-up Anthropic, which is creating an ethical chatbot with a built-in constitution.

Let's get started!

Top News & Why It Matters

| Meta's Open-Source ImageBind AI Aims To Mimic Human Perception

Meta has introduced ImageBind, an open-source AI tool that can predict connections between diverse data types like text, images, audio, and motion. It brings machine learning closer to human learning, using multiple sensory experiences to infer information as our brains do.

Unlike image generators like DALL-E, ImageBind doesn't have to be trained on every possibility before linking different data types. With potential applications in VR, mixed reality, and the metaverse, it could assist virtual game developers, content creators, and even improve accessibility for people with vision or hearing disabilities. engadget

Why does this matter?

ImageBind's functionality could lead to the development of AI that can understand and process more complex and diverse data, making it increasingly versatile and invaluable in various industries.

| ChatGPT: Can Students Pass Using AI Tools At University?

Universities are exploring ways to address the challenges presented by AI applications like ChatGPT. Academics at the University of Bath found that while the tool performs well in answering multiple-choice questions, it struggles with more complex queries that require critical thinking. However, their concern is that less informed users may be easily fooled by the authentic-looking yet inaccurate responses ChatGPT generates.

The University of Bath is working on policies regarding AI tools that are expected to be implemented in September. Education specialists believe that embracing the tech is crucial for students to develop much-needed AI skills for future workplaces, despite possible issues with misinformation. bbc

Why does this matter?

As AI tools like ChatGPT become more prevalent, educational institutions will need to adapt their policies and teaching methods to ensure students can leverage the technology effectively and ethically. This may necessitate investing in more sophisticated AI-assessment tools for grading or monitoring usage.

| A Radical Plan To Make AI Good, Not Evil

Anthropic, a startup founded by former OpenAI researchers, aims to address ethical concerns around AI models by creating a chatbot with a built-in "constitution." The constitution outlines ethical principles that the chatbot, dubbed Claude, should follow when generating responses, making it less likely to produce toxic or unwanted outputs. The guidelines are derived from sources like the United Nations and principles proposed by other AI companies, including Google DeepMind.

Claude's constitution primarily focuses on values such as freedom, equality, brotherhood, personal security, and respect. Adrienne So, a professor at the University of Washington, agrees that Anthropics' approach will work best for companies with large models and abundant computing power, but she thinks that other methods will also prove useful, such as increased transparency in training data and the values assigned to AI models. wired

Why does this matter?

By introducing AI systems with ethical principles built in, there may be a stronger push for greater accountability and transparency in the development of artificial intelligence. The adoption of such practices could help in establishing an industry standard and encourage a more ethical approach to AI creation.

Mastering Artificial Intelligence

| Our Top Picks

✅ Oscar Personal Bedtime Stories: An app makes bedtime special with personalized stories featuring children as the main characters. (website)

✅ ValidatorAI: AI-powered business validator for startup feedback and criticism. (website)

✅ Automata: AI tool repurposes marketing content into multiple formats. (website)

✅ MakeLogoAI: AI logo generator for quick, unique project logos. (website)

✅ Architechtures: AI web tool for faster home design decisions. (website)

| Lunch & Learn

📝Text: Fundamentals of TinyML (link)

🎧 Audio: AI Hustlers - Bloomberg's large language model, Twitter open source code, E3 Cancelled, Sony Q Lite (link)

📽️ Video: Sentdex (link)

What Else Is Happening

| More News

🗞️ How to use DALL-E 2 to turn your ideas into AI-generated art. (source)

🗞️ You Trained the Chatbot to Do You Job. Why Didn’t You Get Paid? (source)

🗞️ Microsoft Work Trend Index: AI will work alongside employees. (source)

🗞️ White House addresses AI’s risks and rewards as security experts voice concerns about malicious use. (source)

🗞️ IBM's Watson returns as an AI development studio. (source)

🗞️ OpenAI’s new tool attempts to explain language models’ behaviors. (source)

🗞️ Microsoft 365 Copilot expands availability through a new early access program. (source)

🗞️ Revitalising European democracy: AI-supported civic tech on the rise. (source)

🗞️ AI Is Coming for Your Web Browser. Here’s How to Use It. (source)

🗞️ 70% of employees are happy to delegate work to AI, according to new Microsoft report. (source)

🗞️ Training machines to learn more like humans do. (source)

🗞️ AI can ease ‘drudgery’ of repetitive tasks at work, says Microsoft report. (source)

🗞️ Spotify cracks down on AI-generated music streaming fraud. (source)

| Best of AI Social Media

💬 Snoop Dogg criticizes streaming, AI, and writers strike. (Twitter)

💬 AI creates images of skydiving animals (Twitter)

💬 AI-generated art commissions scam exposed (Twitter)

💬 "10 AI Chrome Extensions Revolutionize Work Efficiency" (Twitter)

💬 "Happy Birthday Celebration for AI Genshin Impact Collei" (Twitter)

💬 What are some of your favorite ChatGPT prompts that are useful? I'll share mine. (Reddit)

💬 Created using ChatGPT (Reddit)

💬 It's a bit too difficult for me to draw (Reddit)

💬 I used ChatGPT to automatically generate and map travel itineraries for any city / country in the world (Reddit)

💬 Should we just allow students to use AI? (Reddit)

| Research Review

Logic For Explainable Ai (Full Paper)

Adnan Darwiche

∑ Summary: Explainable AI aims to understand learned classifiers' decisions through necessary and sufficient conditions, minimal sufficient conditions, and minimal necessary conditions.

!? Practical implications: This research could lead to more transparent AI systems, which may result in increased trust and widespread adoption of AI technologies in industry and society.

Professional Certification Benchmark Dataset: The First 500 Jobs For Large Language Models (Full Paper)

David Noever, Matt Ciolino

∑ Summary: The research tests large language models' employable skills through professional certification surveys, comparing GPT-3 and Turbo-GPT3.5's performances in various fields.

!? Practical implications: This research could potentially reshape the job market by automating tasks in various sectors, such as IT, customer service, and counselling, leading to a higher demand for AI developers and requiring current professionals to augment their skills with AI-powered systems.

Written by Isaac R. Ward, Casey Clifton, and Alex Brogan.

Send us your feedback at [email protected]. We’re ready to listen.

Join the conversation

or to participate.