#20: Easy access to AI with Clipdrop and Midjourney Alpha
This week we’ll discus the rise of gen-ai tools and how to pick the best ones via Clipdrop, the new interface of Midjourney and we’ll look into a research paper that let’s the Mona Lisa speak. Let’s go!
Gen-AI for everyone to use at Clipdrop.co
There is an enormous number of generative AI tools, and every day, countless new tools emerge. However, this creates a clouded mess of tools where it's difficult to find the ones that are both good and easy to use. Clipdrop.co solves this problem by offering a website with a set of 11+ generative AI tools that are easily accessible for anyone to use. For developers, it even provides a separate overview of 13 easy-to-use APIs, complete with documentation.
Here's a small selection of the most useful tools on Clipdrop.co:
For instance, start using the Generative Fill feature to add to images you've already generated or taken with your phone. It's incredibly easy and accessible for anyone to start using today!
Simply:
Create an account
Go to Generative fill
Upload an image you want to alter
Select the area you want to change
Type in a simple prompt like: "fire and horns"
Press enter and wait...
A centralised place for your Midjourney account
Heavy midjourney users can now use the image generation ai without discord via the new alpha environment. Via the website alpha.midjourney.com you can now start creating new prompts, find the archive of all creations you’ve made over time and rate the images.
Where you previously had to make a discord server with multiple channels, you now have a centralised place as your MJ creation hub.
While it’s still in development, it’s only available for users with a subscription and that at least have generated 100+ images via Midjourney in Discord.
In addition to the centralised hub, Midjourney has also launched a new "infinite styles" feature to help you generate creative prompts. By adding "--sref random" to the end of your prompt, a random style will be applied to your image.
A talking Mona Lisa with Microsoft VASA-1
There have been numerous advancements in deepfake technologies in recent months and years. However, Microsoft has recently made a significant leap forward with the demonstration of Microsoft VASA-1, a deep fake technology that is remarkably user-friendly and nearly indistinguishable from real humans.
What sets this technology apart is its ability to create convincing deepfakes using just a single image and a single audio file of your choice. Below is an example of this groundbreaking new tech:
For anyone interested, the entire research paper can be found at this link. If you're wondering where you can try this technology for yourself, unfortunately it is still being kept behind closed doors at Microsoft. They believe this technology is too potentially dangerous and impactful to release to the general public at this time.