- author: Matt Wolfe
AI News Update: WWDC, Style Drop, and Face-Swapping Tool
This week in the world of AI, there were several noteworthy events that took place. The biggest one was undoubtedly Apple's Worldwide Developer Conference (WWDC), but there were also other interesting developments in the field of AI. In this article, we will cover the highlights of these events and how they impact the AI landscape.
The WWDC event was the main highlight of the week. Apple made several announcements related to AI, including the release of the New Vision Pro mixed reality headset, which will be available next year for a hefty price of $3,500. The headset boasts a high-resolution display and uses advanced technology that requires a battery and a wire to use it. The battery life is limited to two hours, which means it may not be suitable for extended use.
Meta, Facebook's parent company, released an internal memo discussing Apple's new product. In the memo, Mark Zuckerberg highlighted that while the device is impressive, it's not a game-changer since most of the constraints on laws and physics have already been explored by his team. Zuckerberg also stated that he believes the metaverse's future is fundamentally social, which clashes with Apple's demo videos that showed people using the headset alone.
Apple also made several additions to iOS 17, including machine learning features such as real-time transcription of voicemails and automatic transcription of audio messages in text messages. Furthermore, users can now turn their images into stickers and use them as reactions to text messages. Another addition is the improvement of autocorrect using AI/large language models in the typing feature.
Google made an exciting research announcement called Style Drop. It allows users to generate new images that follow a specific style of their choice with a single reference. For example, users can input an image of watercolor flowers and generate images in a similar watercolor style using prompts. This feature will be particularly useful in logo designs and website icons.
Another interesting development is the open-source face-swapping tool called Roop that allows users to replace faces in a video with another person's face. While it's fun to play with, it also has significant implications for trust in media, as it becomes increasingly challenging to distinguish between what's real and what's not. Google and Facebook are already facing pressure from the European Union to label AI-generated content, which would require individuals posting AI-generated content to mark it as such.
EU Urges Google and Facebook to Label AI-Generated Content
The European Union recently urged Google and Facebook to label AI-generated content, which means that whoever posted the content has to mark it as AI-generated. Though it can be viewed as a good idea, it seems difficult to enforce in the long term because of an ongoing race to create better AI-generated content and better tools to spot them. The potential misuse of AI technology by politicians is also a growing concern. The use of fake AI-generated images to discredit other politicians is already happening as the political season begins in the United States.
Hugging Face's 40 Billion Parameter Open Source Language Model
Hugging Face has released a language model with open-source 40 billion parameters. This language model has outperformed all of the other open-source models in testing. While the model is not yet ready for people to run on their own computers, since it requires 90 GB of GPU memory, it is available for people to test on Hugging Face's site. For now, it seems like OpenAI Playground and ChatGPT are still the go-to tools for most large language model needs.
Instagram's AI Chat Bot: Fun and Engaging Experience
A screenshot of Instagram has been leaked on Twitter, which suggests the possibility of an AI chatbot within the app. It is not yet clear whether the chatbot will be within Instagram DMs or help with writing social media posts. The leaked screenshot shows that the chatbot can answer questions, give advice, and chat with 30 AI personalities.
Daz 3D's AI-generated Characters
Daz 3D, which offers a free tool to create character models, has announced that it is adding AI to the mix. A demo video released by the company shows that users can create characters by typing in whatever they imagine, and the AI generates a unique character model. Advanced AI models like Daz 3D's can be combined with tools such as Convey to create unique one-on-one conversations in real-time within video games, creating realistic NPC characters.
Runway's Gen 2
Runway has released Gen 2, which can generate up to 90 seconds of video for free before requiring a paid plan. Anyone can demo the model at runwayml.com by clicking on "Try Now" and then selecting "Introducing Gen 2". The model generated four-second videos and demonstrated the most coherent text-to-video model. The potential combinations are endless when combined with other AI models and tools.
As AI technology keeps progressing, it becomes even more vital to regulate and watch out for its potential misuse. However, it also opens up endless possibilities for creators to make something revolutionary with the help of AI.
Enhancing Creativity with AI Tools
In today's world, various AI tools and technologies are emerging that can significantly enhance our creativity and productivity. Let's explore a few examples of how AI can help us create better content, product images, and even code.
AI-Powered Race Car Drifting
Have you ever wanted to watch a realistic race car drifting around a track? With Generative AI, you can create visually astounding videos in less than a minute. The technology takes some prompting before yielding the desired result. But once you get the hang of it, you can create fantastic multimedia content. If you are curious to see what it looks like, many examples of such videos can be found on Twitter.
Flair AI for Enhanced Product Shots
Creating professional product images can be time-consuming and costly. Flair AI is a cutting-edge tool that blends AI technology with physical products to achieve amazing results in product photography. With Flair AI, you can choose different poses, backgrounds, and skin tones to make your products more relatable to your target audience.
Baidu AI Cloud Coding Assistant
Are you tired of manually typing and completing code lines? Baidu AI Cloud Coding Assistant uses AI to auto-fill code lines by analyzing the context. While it is not perfect yet, it will save a lot of time and energy to developers updating large-code repositories.
Jetpack AI Writing Assistant
Writing good blog posts is a challenge, but with the Jetpack AI assistant for WordPress, creating excellent content has never been easier. The AI-assistant generates blog posts in response to specific prompts and converses with the user in a chatbot interface. It even has basic editing features for spelling and grammar. This AI writing assistant is a great tool for bloggers who aim to increase their productivity.
AI News Roundup
In this AI news roundup, we cover the latest updates in the world of AI. From WordPress updates to the recent launch of stability.ai's new feature called uncrop, there is a lot to cover.
Wordpress AI Assistant Update
If you use WordPress and desire to install its AI assistant for yourself, you will get 20 free requests. Afterward, it's 10 dollars per month to use. However, if you're hosting your website on wordpress.com, the AI assistant is available for free of charge on all wordpress.com sites for a limited time.
Bard Update and Google's Announcement
Bard got an update this week, and Google announced that it's getting better at logic and reasoning. While there is not much to cover in Google's announcement, these updates indicate that Bard and Google's AI capabilities are always improving.
Uncrop by Stability.ai
Stability.ai just launched a new feature called uncrop. This feature is similar to Photoshop's generative fill, where you can take an original image and expand its borders. Uncrop automatically uses AI to generate what is in that outer area of the image.
This feature is currently available to use for free by visiting clipdrop.co/oncrop and dragging and dropping an image. By expanding the image on both width and height, you can see what it generates for you. The AI-generated images added to the original image are impressive, and it's worth trying out this feature.
Moreover, uncrop is not only suitable for AI-generated images but also works with real-life images. It will try to add onto those real images, making it a versatile tool that photographers and graphic designers can find useful.
AI News Roundup: Final Thoughts
Overall, despite the Big Apple event that garnered media attention, many AI tools, and research updates were overshadowed by it. However, the AI industry is buzzing with new developments and tools that we should take note of.
If you're interested in staying in the loop about the latest AI updates, head over to futuretools.io. We update the news page every day with more news than what we share in these videos, and we're adding all the cool tools we come across.
To receive a weekly TLDR of AI news, join our free newsletter by clicking this button on futuretools.io.
Thanks for tuning in, and we appreciate your support. See you in the next video!The events of this week illustrate the rapid advancements in ai technology that are changing the way we interact with the world. while some developments are exciting, others raise concerns about privacy and trust in media. as ai continues to evolve, we must tread carefully and ensure that we use this technology ethically and responsibly.
Ai technology is revolutionizing the way we create content, code, and capture product images. the examples above hint at the vast possibilities of what one could achieve with these ai-powered tools. many other innovative ai applications are sure to come up in the future, and with them, our creative boundaries will continue to expand.