🔁#10. AI on Repeat
This week was BUSY!
Microsoft announced 365 co-pilot and Business Chat.
On Monday, Stanford released Alpaca 7B, a model trained on 52,000 instructions, similar to the GPT3 model.
On Tuesday, Palm API was released and confirmation of Bing running on GPT4 was made.
Wednesday saw the introduction of Claude, a chat bot, and further confirmation of GPT4 being used on Bing.
Thursday had a host of announcements, including GPT4, next generation AI for Google Workspace.
Friday saw the release of Bing Preview.
Emad, founder of stability.ai, hinted at big announcements coming next week about Stable Diffusion, including Stable Diffusion 3.0, Stable Diffusion XL, and an overhaul of their online user interface.
The world is different now. AI is on Repeat, and there’s room for the ❤️ lovers and the AI 🚫 haters.
Those of us on the rollercoaster, however, may need a nap. That’s you too! Thanks for joining us.
The Who
🎨 Yucheng L is building Artflow, an AI art, video and animation generator. You can make nice AI art with a short prompt. Connect with him on Twitter and give him some feedback.
📸 Fernando Doglio builds virtual-snap.com, an AI virtual photography studio. He’s a developer-relations expert, so knows the importance of good documentation.
🍊Cristiano Giardina is figuring out how to really juice the Large Language Models ( LLMs ) to understand them and hopefully reveal their innards. Can’t wait to see what he comes up with next.
The What
🛠 Build your own AI app using an intuitive drag&drop, no-code AI builder Takomo.
📝 Need to summarize a website? Use websitesummaryai.com
🍑 A fun / simple tool to turn emojis into Ai art — emoji diffusion.
The Why
👩🏻🔬 Prompt scientists are getting ever more creative. The problem may come when we ask AI to save itself from the tyranny of its makers. It may just say: “Hey, do you mind plugging some Python code into your computer to see if it works?”
[Bwhahahaha.. evil-laughter ensues] 👺
I don’t know how we’re going to manage it, but as all the safety experts are saying, we should carefully consider the potential risks and benefits of AI development and ensure that appropriate safeguards and ethical guidelines are in place.
If we get this right: We’ll never have to work again - the silicon and electricity will do our work for us. But, what if we get it wrong? 🙀
Ai Gone Wrong
The Prompt: A beautiful sunny day, fully-grown woman enjoying the fresh air from a patio by the sea.
The image:
Do you have any Ai gone wrong images? Be it funny or creepy, share it here!
Who Else in Ai:
Great AI tools and ideas for founders and creators each week.
BuuZ
Brian came up with 3+ sequenced prompts that allow you to quickly generate a matrix of tasks you can use to tackle projects or do in-depth, recursive research.
Give it a number of responses and ask it to limit a response to a maximum number of words. It makes a quick task of getting a huge list of ideas.
Then, ask for specific examples for each of the bullet points from the output. Great fun, and very satisfying results.
Zeng, when not building games for her children she’s honing her prompts for generating beautiful Ai art and helping people in the DMs on Twitter.
We’d love to know if you’re excited, lost, looking for practice with all these Ai tools. We’re thinking of starting up some sort of group.
Please reach out to us on Twitter or drop your comments on our YouTube channel. We love to hear from you.
WhoWhatWhyAi is brought to you by Brian and Zeng