This week the world saw huge leaps in the field of AI. Here are some interesting new things that happened this week. Google’s styledrop was the google’s latest AI tool and advancement in adobe firefly shook the world.
1. SnapFusion – Ability to run stable diffusion on smartphones
After google’s research team ran stable diffusion successfully on a smartphone, snapfusion team published a research paper and also the results of their research.
They claim that they can get the same results in 8 steps opposed to stable diffusion 1.5 on a PC with 50 steps, which would democratic the content creation further to those with low end GPUs or no PC to start with.
2. Google Bard new update
Google introduced a new update to Bard that would make Bard more efficient with mathematical calculations. Google claims that Bard would be 30% more efficient in mathematical calculations prior to update.
3. Adobe FireFly powered with Adobe Express
Adobe Firefly just got more powerful. With the introduction of Adobe Generative fill to Photoshop adobe already was ahead of its competitors. Generative Adobe express will allow creators to make customized text pop-specifically.
In a video adobe demonstrates how they customized cards with just a few simple prompts. This will increase productivity by 10x in my opinion. You can resize, recolor, re imagine the whole template. Amazing isn’t it?
4. Adobe Firefly can now be customized
Another adobe update that is good news for creators is that now they will be able to customize their adobe firefly model according to their art style or creation.
It is sort of like how stable diffusion models are customized according to specific generation style.
Service is only available for its enterprise customers at the moment.
5. RunwayML release Gen 2 Text to Video
Runway released gen 2 of their text to video model and it’s available for a free trial. This is just revolutionary stuff. Check their demo video here
6. Google’s StyleDrop
Google’s styledrop is another tool that would increase the productivity of any artist by 10x. Google’s styledrop helps creators to generate images in a particular style by just one reference image.
One thing that stable diffusion struggles with is text but with styledrop you will be able create stylish text by just writing simple prompts and a reference image.
7. JetPack AI for writing blog posts – Good news for bloggers
Good news for bloggers who use WordPress for their websites. Jet pack has been powered by AI to write blog posts for you.
From changing the tone of the text to correcting grammatical errors this tool can be really handy for serious bloggers or the ones that are just starting out.
For now this AI tool is free for WordPress users, while other jetpack users will get 20 free suggestions and will be charged $10 after that.
8. HuggingChat by Hugging Face
Hugging chat by hugging face is an alternative to ChatGPT. Access Hugging chat by going to this link. HuggingChat also introduced a web search feature to its chat.
9. Visual Captions by Google
Visual captions by google is another revolutionary feature that would change the way we video chat. Google demonstrated that you will be able to see images related to the speech in the video.
Suppose you are talking about New York Statue of liberty then that photo as a caption will appear in the video. If you talk about your mom, your mom pic will be shown in the corner.
I think this will be a great feature for those who can’t hear and will now be able to to interact with the others based on visual captions.
10. Tafi introduced text to 3D character models
Tafi has now come up with text to 3D characters model and you will be able to generate 3D characters by simple prompts. You can join Tafi’s beta program and get early access to text to 3D model.