- Prompt Hunt Newsletter
- Posts
- The Infinite Conversation
The Infinite Conversation
Here's your daily briefing:
There's lots of research out there that shows the benefits of stories and storytelling for helping us to internalize and remember the things we learn. With this cool new app you can use AI to generate a story about any topic:
🌃 Who doesn't love stories?
Stories can be a fun way to learn new things. And guess what? Reading them can be super addictive too :)
You guys loved the MVP, so I am officially launching it.
🌎 whatonearth.xyz
What did you learn today? Share your screenshots with us!
— naklecha (@naklecha)
11:38 AM • Oct 31, 2022
Our attempt:

"Any sufficiently advanced technology is indistinguishable from magic." -Arthur C. Clark. We couldn't agree more. We've been knee-deep in this space for a while now and this image-to-video demo from Runway still feels like magic:
Introducing Frame Interpolation
Generate seamless video from still images with just one click.
Available now: bit.ly/3E373m1
— Runway (@runwayml)
1:00 PM • Nov 2, 2022
We haven't had a chance to play around with it yet but this AI writer plugin for Notion seems very cool, especially if you're already a fan of Notion (we are):
We did not have an AI-generated infinite conversation between Bavarian director Werner Herzog and Slovenian philosopher Slavoj Žižek on our bingo card for "methods of raising awareness about AI and media literacy," but hey, the world's a wild place innit?

From the "About" page:
This project aims to raise awareness about the ease of using tools for synthesizing a real voice. Right now, any motivated fool can do this with a laptop in their bedroom. This changes our relationship with the media we consume online and raises questions about the importance of authoritative sources, breach of trust and gullibility.
Will this technology lead to a massive proliferation of sub-optimal-quality content? Should we simply distrust anything we see online? As new tools are developed to help identify generated content, I recommend maintaining a skeptical stance, particularly when the source/channel of information doesn't seem reliable and when the claims seem preposterous or outrageous.
Ultimately, I don't see this as a technical problem, but as a human one. We all share a duty to educate the coming generations about the new paradigm while focusing on forming compassionate individuals who would not misuse these awesome powers.
Today in "things we don't quite understand but seem like a big deal":
Using HyperTree Proof Search we created a new neural theorem solver that was able to solve 5x more International Math Olympiad problems than any previous AI system & best previous state-of-the-art systems on miniF2F & Metamath.
More in our new post ⬇️
— Meta AI (@MetaAI)
1:00 PM • Nov 3, 2022
Another generative avatar maker dropped. As far as we can tell, this is the first of these proprietary tools we've seen that's totally free:
Coqui is building a Swiss-army knife for generative AI:
Generative AI 🤖🎨 has fundamentally disrupted content creation 🔥
Now AI can:
✅ Generate expressive voices
✅ Create video from text
✅ Compose beautiful music
✅ Write long, coherent textsPut it all together and you get the future of media 🐸🚀
1/🧵
— coqui (@coqui_ai)
6:52 PM • Nov 2, 2022
🐙 Notes for Prompt Engineering
If you are tracking the ML space closely, you may have noticed all the interest in prompt engineering. This repo contains a useful collection of resources/notes on prompt engineering.
github.com/sw-yx/prompt-e…
— elvis (@omarsar0)
12:49 PM • Oct 18, 2022
This very cool generative tool just dropped by Tal Stramer, with backing/advising by esteemed giga-brain Andre Karpathy:
✨Excited to share a project I've been working on with advising from @karpathy . Introducing stableboost.ai , a practical tool to generate images and videos with AI! 🧵
— Tal Stramer (@tall)
9:04 PM • Nov 2, 2022
We think a lot about how AI can be used to actually improve our lives (as opposed to just our bottom line), but we didn't think to imagine how it might be used to save the lives of people who are thinking of taking their own:
From the piece:
Recent research conducted by Ms. Kusuma and a group of scientists from the Black Dog Institute and the Centre for Big Data Research in Health investigated the evidence supporting machine learning models’ ability to predict potential suicidal behaviors and thoughts. They evaluated the efficacy of 54 machine learning algorithms that were previously created by researchers to predict suicide-related outcomes of ideation, attempt, and death.
The meta-analysis, published in the Journal of Psychiatric Research, found that machine learning models outperformed conventional risk prediction models in predicting suicide-related outcomes, which had traditionally performed poorly.
That's it for today! We'll see you tomorrow : )

"homer simpson riding a bicycle down a tree lined road in the style of early 1900's photograph"



