🤖 Text-to-video is here
PLUS: Exclusive preview of GPT-4's image-to-text
In today's rundown:
🚨 Sam Altman on AI: Potential pitfalls and harnessing potential
🎥 Text-to-video is here (sort of): Breakthrough for generative AI
👁️ GPT-4 for greater accessibility: Exclusive preview of Be My Eyes
🎉 Today's giveaway: 500+ best ChatGPT prompts
Read time: 5 minutes
🎉 Exciting news:
Due to the recent rapid advancements, we've realized that our bi-weekly emails aren't enough to cover all the important news.
We're thrilled to announce that we're increasing the frequency to 5 times a week, from Monday to Friday. As always, our updates will be and always will remain free.
This will allow us to cover all the significant AI updates, while maintaining our commitment to concise and digestible content.
A massive thank you to our readers for your ongoing support – this expansion wouldn't be possible without you!
Now, let's dive into today's newsletter, shall we?
🚨 Sam Altman on AI
In a recent ABC News interview, OpenAI CEO Sam Altman discussed the potential pitfalls of AI.
We found two quotes particularly impactful:
On AI taking jobs: "It is going to eliminate a lot of current jobs, that's true. We can make much better ones." [full video here]
His motivation for developing AI: "In terms of impact on our lives and improving our lives and upside, this will be the greatest technology humanity has yet developed" [full video here]
Sam did an excellent job in this interview by acknowledging AI's challenges but also highlighting AI's potential to revolutionize industries and improve countless lives.
🎥 Text-To-Video is Here (sort of)
The first-ever open-sourced text-to-video model is out, and it's, well, interesting.
The model, available Visit huggingface.co to try on Hugging Face, is trained on a whopping 1.7 billion parameters.
We think it's "interesting" because it appears as if the model was actually trained on videos from Shutterstock. A lot of videos that it produces, have Shutterstock watermarks written across them.
For example, check out this Twitter post where someone created their own Star Wars movie clip using this model.
While it's incredible that text-to-video is here, our excitement feels somewhat damped by the fact that there are Shutterstock watermarks all over it. However, this is a giant leap forward, and we can expect even better text-to-video models very soon.
👁️ GPT-4 For Greater Accessibility
Be My Eyes, an app that allows blind and low-vision individuals to seek assistance from sighted people has just launched their new "Virtual Volunteer" feature.
The feature is the first to integrate the new image-to-text generator powered by OpenAI's brand-new GPT-4.
Users can soon send images to the AI-powered Virtual Volunteer, who provides instant identification, interpretation, and conversational visual assistance for various tasks.
We slid into Be My Eyes CEO Mike Buckley's DM's on Twitter, and he generously shared some never-seen-before exclusive access to the current use of GPT-4's image-to-text model:
The images he sent us were incredible, to say the least. GPT-4 was able to detect content from a TV screen from a small corner of an image and translate it in real time to another language.
Be My Eyes remains one of our favorite AI companies, demonstrating the powerful potential of AI for the greater good. We look forward to providing ongoing coverage of their progress.
Sign up for the Virtual Volunteer waitlist on their website Visit bemyeyes.com.
☀️Sunsama (Sponsor)
Meet the Tesla of Productivity Apps.
Many of the most impactful leaders know exactly what they need to accomplish before the day starts.
They use productivity apps like Sunsama to help them organize their weekly goals and smaller tasks to become the most productive version of themselves.
🔥 Trending
ChefGPT- AI-Powered digital chef that makes meal ideas easy (Visit chefgpt.xyz)
Chat Thing- Chatbots for anything powered by data from anywhere (Visit chatthing.ai)
PromptPal- Share and discover the best AI prompts (Visit promptpal.net)
MovieBot- Create, edit, and share AI-generated movies instantly (Visit moviebot.io)
GetBotz- Set your blog on autopilot with AI (Visit getbotz.io)
Bloop- Understand codebases with GPT-4 (Visit bloop.ai)
RATH- Data, exploration, and visualization with automation (Visit kanaries.net)
OpenRead- Experience the future with AI-powered interactive papers (Visit openread.academy)
Storywizard- Create incredible children’s stories in no time using AI (Visit storywizard.ai)
Easy Peasy- Create amazing content 10x faster (Visit easy-peasy.ai)
My AskAI- Your own AI Q&A, powered by ChatGPT with no-code (Visit myaskai.com)
AI Gift for you- Search for gift ideas with AI based on personalized filters (Visit aigiftforyou.com)
Cody- ChatGPT with bonus of being able to train it on business (Visit meetcody.ai)
STORYD- Create data presentations in seconds (Visit storyd.ai)
Gitfluence- Quickly find the correct Git command (Visit gitfluence.com)
Sizzlepop- Create custom t-shirts with AI (Visit sizzlepop.ai)
AItinerary- Creates personalized travel plans in seconds using AI (Visit aitinerary.ai)
ConversAI- Respond to anything across any social media platform (Visit conversai.co)
Magic Eraser- Erase unwanted background objects in images (Visit magicstudio.com)
CanceledGPT- Finds old offensive tweets before you get canceled (Visit canceledgpt.com)
View our database of all the best AI tools mentioned in our newsletter:
Have cool resources to share? Submit a tool or reach us by replying to this email.
🐦 Tweet of the day
20 jobs that GPT-4 thinks it will replace:
20 jobs that GPT-4 will replace, written by GPT-4:
— Rowan Cheung (@rowancheung)
6:06 PM • Mar 15, 2023
That's all for now!
If you have any interesting projects or ideas, please reach out to us by responding to this email or by sending us a DM on Twitter: @rowancheung & @therundownai
As always, thanks for reading, and see you next time. 🫡
Interested in sponsoring The Rundown? Check out our sponsorship options to see if you'd be a good fit!
How was today's email?We'd love to have your feedback! |
|
