Need to Know News - December 11th, 2024
In this week's Need to Know News edition:
🤖 Open AI's text-to-video tool Sora is finally here...
🤖 AI celebrity mentors you can chat with anytime...
🤖 Mastercards AI-powered plan to help you reduce abandoned checkouts...
Plus, a whole lot more!
AI In Action
OpenAI's New AI Video Tool Sora Goes Live (Finally)
OpenAI finally dropped their game-changing video creation tool. Sora turns simple text descriptions into high-quality videos. And it's now live for ChatGPT subscribers in most countries except Europe and the UK.
The pricing tiers pack different perks. ChatGPT Plus users ($20/month) can make fifty 720p videos running 5 seconds each. The beefier Pro plan ($200/month) unlocks unlimited 1080p videos stretching to 20 seconds - without watermarks.
Free users aren't left out - they can browse a feed of AI videos made by others. To keep things safe, OpenAI added visible watermarks and special metadata showing when AI makes a video. The launch hit a snag when some test artists leaked the product, claiming OpenAI used their work without pay.
OpenAI Also Makes ChatGPT Canvas Free - New Code and Graphics Features Added
Previously, only paying subscribers could use this powerful tool. It turns ChatGPT's basic chat into a side-by-side workspace. And brings other upgrades along with it. You can now write and test Python code right in Canvas. It even spots bugs and suggests fixes. Plus, it creates charts from code that you can preview and tweak instantly.
Canvas works differently than regular ChatGPT. Instead of generating new responses in chat, it shows changes directly in a right-side panel. This makes editing and collaboration smoother. Custom GPT creators can now add Canvas to their tools too, making them more powerful.
Full Story
🚀 Discover the 27 ‘Plug-and-Play’ AI Prompts Every Marketer Needs
Access All 27 Prompts Here ==>
MasterClass Launches $10/Month AI Celebrity Mentors You Can Chat With Anytime
MasterClass just made celebrity mentoring affordable through AI. Their new product, MasterClass On Call, lets you chat with AI versions of famous teachers. Some of the initial AI mentors include Mark Cuban and Richard Branson
These AI mentors sound remarkably human. They can role-play scenarios, give instant advice, and even draft emails. The company builds these AI personas using multiple models working together. Each mentor's knowledge comes from their courses, writings, and special interviews.
Early tests revealed something surprising. Users felt more comfortable asking "dumb questions" to AI mentors than real people
Meta's New AI Helper "Metamate" Aims to Revolutionize Workplace Tasks
Meta just launched an AI assistant that could change how its employees work. Built on their Llama language model, Metamate helps staff with coding, research, and writing tasks .
While it's currently for internal use only, Meta has big dreams. They want Metamate to become "the world's best enterprise assistant." The timing is smart - experts predict the AI tools market will explode from $5.1 billion to $47 billion by 2030.
Yelp's New AI Tools Turn Reviews into Smart Business Insights
Yelp is making its massive collection of reviews smarter with AI. Their new Review Insights feature uses AI to analyze what people say about businesses. It creates easy-to-read scores from 1 to 100 for things like food quality and service.
The company isn't stopping there. They've launched over 80 new features this year, with more coming soon. Users will get personalized home feeds powered by AI that show relevant businesses and trending content.
AI-Powered Chatbot ‘Reddit Answers’ is Slowly Rolling Out
Reddit is jumping into the AI game with a clever new tool called Reddit Answers. This AI chatbot helps users find exactly what they need from millions of conversations. It creates quick summaries and suggests related topics, making Reddit's vast content easier to digest.
The timing is interesting. Reddit just turned its first profit in 19 years, with AI playing a big role in their success. Their translation tools alone brought in four times more users than before. Now they're betting big on AI again with this chatbot launch.
For now, only some U.S. users can try Reddit Answers. But the company has bigger plans. They'll expand to 30 countries by 2025, using AI to break down language barriers. It's part of Reddit's strategy to keep growing after reaching 97 million daily active users - up 47% from last year.
Full Story
At the Frontier of AI
Ad Giants Merge to Supercharge AI and Tech Capabilities
Two advertising powerhouses, Omnicom and Interpublic, are joining forces in a game-changing merger. Why? To better tackle the AI revolution sweeping through advertising. In today's world, AI makes it easier than ever to create viral content that spreads fast.
The deal isn't just about saving money. It's about staying ahead in a world where data and technology rule advertising. The combined company will be bigger than their main rivals, WPP and Publicis Groupe. They'll have more muscle to invest in cutting-edge AI tools and real-time data tracking.
CEOs from both companies agree - success in advertising now demands heavy investment in AI and tech. This merger follows a broader trend, as 70% of companies now use AI in their marketing efforts
MIT's New Tool Makes AI Show Its Work, Boosting Trust in Chatbots
MIT researchers have created a clever solution to a big AI problem: How do we know if chatbots are telling the truth? Their new tool, ContextCite, acts like a fact-checker for AI. It shows exactly which sources an AI used to answer questions.
Here's how it works: When AI makes a statement, ContextCite highlights the specific information it used. If the AI makes something up (called a hallucination), the tool reveals that too. It's like catching AI in the act of doing its homework wrong .
The tool has another smart trick - it can spot "poisoned" sources where bad actors try to trick AI systems. This matters because AI is becoming a daily helper for many people. Companies like Google are already using AI to write 25% of their code , making trust more important than ever.
Mastercard Says the Checkout Revolution Will Be Optimized … by AI
Mastercard is using AI to fix a costly problem - failed online payments. When checkouts go wrong, 42% of shoppers never return to that store. That's why Mastercard built an AI-powered Payment Optimization Platform to make buying things online smoother.
Here's the clever part: Their AI analyzes massive amounts of payment data in seconds. It can predict when you'll have enough money in your account. If your payment fails, the AI knows the perfect time to try again. No more awkward declined cards.
The system acts like a smart traffic controller for payments. It guides transactions through the best routes and fills in missing information that banks need. For stores, it means more sales. For shoppers, it means fewer headaches when checking out.
AI Solves Complex Engineering Problems in Seconds, Not Hours
Johns Hopkins scientists just created an AI that's changing the game in engineering. Called DIMON, this smart system can solve massive math problems on a regular computer - no supercomputer needed.
The breakthrough? DIMON learns patterns instead of crunching endless calculations. Think of it like this: Instead of solving a puzzle piece by piece each time, it recognizes the whole picture instantly. This makes it thousands of times faster than traditional methods.
In tests with heart models, DIMON turned a week-long process into a 30-second task. That's huge for doctors trying to help patients with irregular heartbeats. But it's not just for medicine. The AI can tackle all sorts of engineering challenges - from car crash testing to bridge design.
Got 15 Seconds?
Thanks for reading.
Until next time!
The AI Marketers
