Innovations in Generative AI: Weekly Update #2
Written on
Chapter 1: Introduction to the Weekly Update
Hello everyone! Welcome to the second installment of my weekly overview on the advancements in AI, which I also share through my newsletter at Lore.com.
In just a week since its launch, we've surpassed 360 subscribers, with around 700 readers engaging with the first issue. My initial target for the month was just 100 subscribers! I'm still fine-tuning the newsletter's style and layout, so your feedback is greatly appreciated.
In this edition, we'll cover the following topics:
- 💡 Most Innovative Product: Runway AI for video editing.
- 🧠Key Insights: Surprising capabilities of Jasper and trends with GPT-3 Twitter bots.
- 📰 AI News: Partnership between OpenAI and Shutterstock, release of Stable Diffusion 1.5, and more.
- 🤯 Highlight of the Week: An AI-generated music video.
- 🖼 Artist Competition Results: Over 200 submissions received!
Chapter 2: Most Innovative Product - Runway AI
This week, I explored Runway after a reader's recommendation, and it's impressive! Think of it as the Figma for video editing. Runway utilizes AI to produce stunning videos that would typically require significant time and financial resources to create.
With Runway, users can:
- Collaborate in real-time.
- Perform inpainting and convert images and text into video content.
- Eliminate backgrounds and objects from videos.
The potential for YouTube and social media creators to achieve Hollywood-style special effects at a fraction of the cost is remarkable.
Chapter 3: Key Insights from My Experience
I was pleasantly surprised by Jasper's performance. Initially, I learned about Jasper due to its recent $125 million funding round, and after testing it, I found it exceeded my expectations. It assists in crafting various content types, from blog posts to advertisements, and offers SEO suggestions that help improve Google rankings—making it a valuable resource.
This tool has the potential to significantly boost productivity, enabling users to produce 3-5 times more content or simply complete the same amount of work more efficiently. I'm planning to experiment with it for blog content and a small advertising test next week, and I'll share my outcomes with you.
Additionally, I stumbled upon a humorous YouTube video by Fireship while researching GPT-3 bots. He created a Twitter bot powered by GPT-3, which produced amusing tweets that garnered significant engagement. If you're interested in trying it, check out the GitHub repository linked in the video.
However, it's important to be cautious; Fireship noted that this activity violated OpenAI's terms of service, leading to the suspension of the bot's activity. This raises concerns about the authenticity of accounts on social media, as the influence of AI-generated content continues to grow.
Chapter 4: Community Concerns with Stability.ai
Stable Diffusion 1.5 has finally launched, yet there are ongoing issues concerning Stability.ai, the company behind the development of Stable Diffusion alongside Runway. The much-anticipated release faced delays, which many speculate could be linked to commercial motivations or governmental pressure.
There has been growing concern within the community, especially following a controversial Reddit post highlighting Stability.ai's perceived takeover of community platforms without notice. Recent communications from the company have further fueled skepticism, as they blamed Hugging Face for the early release of Stable Diffusion 1.5.
Trust within the community is crucial, and the current sentiment appears to be shifting towards skepticism. I hope they can address these issues, as failure to do so could jeopardize their standing in the AI space.
Chapter 5: AI News Highlights
In partnership news, OpenAI has collaborated with Shutterstock to allow users to generate and license AI-created content, ensuring that artists receive compensation for their contributions. This partnership clarifies the recent ban on sales from AI artists.
Additionally, Stable Diffusion 1.5 was unofficially released by Hugging Face, and Scale recently held an AI conference in San Francisco. Eric Schmidt, former Google CEO, has been appointed to the National Security Commission on AI, a move that has sparked some criticism.
Chapter 6: Highlight of the Week
Check out this remarkable video created by Dreaming Tulpa, utilizing Stable Diffusion and Deforum. It offers a glimpse into the future of AI in music production—imagine fully realized music videos crafted by AI enthusiasts in just a few months!
Chapter 7: Artist Competition Results
This week, I introduced a new approach by involving the AI art community in a contest to create cover art. We received over 200 submissions from talented artists, and the results were impressive.
Winner: Proxima Centauri b
Runners-Up:
- "I Am" by Shaktidreams
- Modern Occult Piece by Lust
- Fantastical Stolen Kiss in the Night by AmliArt
Honorable Mentions:
- Works by various talented artists, including LunaMage AI Art, Rob Dawkins, and many others.
I wish I could acknowledge every participant, but the newsletter is getting lengthy!
Closing Thoughts
That's all for this week. I have some exciting updates planned for next week, so stay tuned!
Until next time,
Nathan Lands
💌 Support the Newsletter
Share this newsletter and follow me on Twitter @NathanLands.
Consider signing up for our affiliate partner's services to help support this newsletter in the long run. Thank you for your continued support!