#2: 🎅 Merry Christm(AI)s! Education will never be the same, GPT-4 rumors, and more on AI art
AI is having a massive impact on education. How are teachers responding? 🎓 Read this week's issue of Not A Bot, the newsletter about AI that was *definitely not* written by AI.
Greetings, fellow humans. This is Not A Bot - the newsletter about AI that was definitely not written by AI. I’m Haroon, founder of AI For Anyone, and I’ll be sharing with you the latest news, tools, and resources from the AI space.
Merry Christmas (for those who celebrate) and Happy Holidays! 🎄
Here’s an AI-generated Christmas carol to mark the occasion:
Feeling festive yet? ⛄
Okay good. Let’s get into it…
🧠 AI in education - who’s doing the thinking?
Few questions in AI are more hotly debated than how the technology should be used in classrooms. This topic contentious one for some time now, but has reached scorching 🔥🔥🔥 levels in the past few months.
The emergence of OpenAI’s GPT-3 marked one of the biggest leaps in tech that we’ve ever seen, and it caught entire professions off-guard.
Many teachers, frankly, didn’t expect to deal with the mass use of AI in the classroom until much later on.
Stephen Marche, a former Shakespeare professor, explains his (incorrect) prediction in The Atlantic:
Going by my experience as a former Shakespeare professor, I figure it will take 10 years for academia to face this new reality: two years for the students to figure out the tech, three more years for the professors to recognize that students are using the tech, and then five years for university administrators to decide what, if anything, to do about it.
Teachers are already some of the most overworked and underpaid professionals in the world, and now, they have to deal with their profession being revolutionized virtually overnight.
Kevin Bryan, Associate Professor of Strategic Management at the University of Toronto, makes the bold statement that take-home exams and homework are now pointless, since students can use tools like ChatGPT to generate responses in a matter of seconds:
But it’s not all doom and gloom.
Some folks, like Ben Thompson of The Stratchery, are more optimistic about the role that AI is going to play in education, and believe it can even improve the learning experience.
In his piece “AI Homework”, he suggests it could be an opportunity to learn how to recognize correct and incorrect AI-generated outputs, a valuable skillset to have today:
In the case of AI, don’t ban it for students — or anyone else for that matter; leverage it to create an educational model that starts with the assumption that content is free and the real skill is editing it into something true or beautiful; only then will it be valuable and reliable.
Some tech-savvy trailblazing teachers, like Ethan Mollick, are turning the tables by leveraging tools like GPT-3 to make their jobs easier:
Regardless of which side of the debate you’re on, a few thing are for certain:
Technologies like GPT-3 are here to stay
These technologies will only get better over time (GPT-4 is right around the corner)
Students will inevitably use these technologies
So the question isn’t “do we let students use AI in school?”, but “how can we best leverage AI to improve educational outcomes?”.
As Gary Vaynerchuk said in a video about ChatGPT in the classroom, “Don’t fear innovation, figure out how to take advantage of it.”
Teachers - we’re rooting for you. 🧡
PS: would you be interested in attending a “Generative AI 101” workshop, where we cover the basics of generative AI + ethical discussions over the course of 60 minute?
Let me know in the below Twitter thread and we’ll see if we can partner with the Mark Cuban Foundation again for a sequel to our record-breaking “Learn AI in 60 Minutes” virtual workshop:
🗞️ In other AI news
OpenAI releases Point-E, a 3D version of the DALL-E digital image AI (TechCrunch)
AI summary: OpenAI has open sourced Point-E, a machine learning system that creates a 3D object from a text prompt. Point-E generates point clouds, which are easier to synthesize from a computational standpoint but don't capture an object's fine-grained shape or texture. To get around this limitation, the Point-E team trained an additional AI system to convert Point-E's point clouds to meshes. Point-E is made up of two models: a text-to-image model and an image-to-3D model. The text-to-image model generates a synthetic rendered object that's fed to the image-to-3D model, which then generates a point cloud. Point-E's point clouds could be used to fabricate real-world objects, such as through 3D printing, and could potentially be used in game and animation development workflows.
GPT-4, the upcoming successor to OpenAI’s GPT-3, is supposedly 10x better than GPT-3.5 (Twitter)
AI summary: Sam Altman (CEO of OpenAI) believes that AGI (artificial general intelligence) and fusion are important, with AGI being dependent on exaflops (a measure of computational power) spent on training. He mentions the GPT-3.5 and GPT-4 language models as examples of AGI, and mentions that Google and Microsoft are heavily invested in the development of AGI. Sam suggests that the key to intelligence is the amount of exaflops spent on training, and that the current trend towards centralization in the development of AGI can be reversed by optimizing the process of gradient descent, which is used to improve the accuracy of machine learning models, and making it asynchronous. He also notes that while model configuration and training parameters do not seem to matter as much as the amount of exaflops spent on training, expert training is still important. Finally, the speaker expresses interest in the development of a "co-pilot" for text-based browsers.
AI art is having a breakout year, but it’s bringing along a host of ethical dilemmas (Emerging Tech Brew)
AI Summary: Generative AI, which creates original content or images through artificial intelligence, has gained significant attention and funding in recent months, with companies in the sector raising $1.3bn through November 2021, a 15% increase on the previous year. The technology has a range of potential applications, including in marketing, graphic design, film and retail, but has also raised ethical and legal concerns, with image platforms such as Adobe, Getty and Shutterstock taking different approaches to AI-generated content. One programmer has even sued Microsoft, GitHub and OpenAI over GitHub Copilot, a generative AI tool that used publicly available code to learn how to write its own code.
🔥 Trending Tools
Here’s a list of my five favorite AI tools & resources of the week
Anypod - A search engine for podcasts. This makes podcast knowledge much more discoverable.
Tome.ai - Use AI to create narratives. The demo video illustrates using Tome to create slides based on a prompt. I know many people (me) are excited for a world where AI generates PowerPoint slides for them.
You.com Chat - You.com, a search engine, launched a ChatGPT-like chatbot. It pulls high-quality search results and cites them automatically.
The Safe Zone - the first-ever film directed and written by AI. I’ve only seen snippets, but looks great.
AI Video Avatars - Lensa avatars are old news…it’s all about video avatars now!
🤔 Interesting Tweet of the Week
An employee at Runway, an AI-powered video editor company, re-created the White Lotus Season 2 opening credits with scenes from Home Alone.
Wow. Just…wow.
And that does it for this week’s issue.
As always, thanks for reading, and see you next time. ✌️
Haroon - (definitely) Not A Robot
🐦 @haroonchoudery on Twitter
Fantastic share!!!!