This new AI video generator called Cling, which comes out of China, has been the talk of the AI world this week. If you have a Chinese phone number, you can reportedly register for the app and use it right now.

The videos generated by Cling are typically around 5 seconds long, but there are examples of longer videos as well. One video shows a boy riding a bike, with the environment changing from desert to snowy landscapes as the video progresses. While the videos are clearly AI-generated, they are impressively realistic.

Cling also has a feature that allows you to upload an image and a template action, and it will animate the image to match the action. This has resulted in some creative and entertaining examples, such as a man dancing on the beach or people eating various foods.

Overall, the Cling AI video generator seems to be producing results that are better than many other video generators we've seen lately, though they still don't quite match the quality of Sora. It will be interesting to see how this tool develops and whether it becomes more widely accessible outside of China.

Ton Crafter is a cool AI tool that can animate between two frames. You provide it with a start image and an end image, and it will generate the animation in between.

The tool works best with cartoon-style or anime-like images, rather than real photographs. It can take a simple head turn or a character taking a step and animate the transition smoothly.

You can use Ton Crafter right now for free on Hugging Face. Just upload your start and end images, and the tool will generate the animation. It's an open-source project, so you can also download the code and run it locally on your own computer.

Some examples of Ton Crafter in action include:

• A man walking down the street with an umbrella
• A glowing orb or gem pulsing and changing
• A cartoon character's head turning slightly and blinking

Overall, Ton Crafter provides a simple but effective way to animate between two frames, making it a handy tool for creating short, looping animations without having to manually draw each frame.

Domo AI is a tool that allows users to transform regular video footage into cartoon-like animations. Here's how it works:

• Users can upload a video file to the Domo AI platform.
• The tool then processes the video, applying cartoon-style filters and effects to create an animated version of the original footage.
• This can be done for a variety of video sources, including clips from movies, TV shows, and user-generated content.
• The resulting animated videos maintain the original movement and actions, but with a whimsical, hand-drawn aesthetic.
• Domo AI even handles tasks like lip-syncing, ensuring the cartoon characters' mouths move in sync with the audio.

This tool provides an easy way to give standard videos a unique, animated look and feel. It can be used for creative projects, video essays, or simply to add some visual flair to existing footage. Domo AI makes the cartoon transformation process accessible to a wide range of users.

Proper prompter recently shared a new tool called Verse, which includes a feature called Magic Brush. This feature allows you to select a specific portion of an image and animate just that selected area.

Here are some examples of what the Magic Brush feature can do:

• Animating Harry Potter's wand, with the hand and wand moving.
• Animating Elon Musk's face, making him nod.
• Animating a rocket ship taking off, with the steam coming out.
• Animating the Hogwarts Express train, with the steam and movement of the train.

The Magic Brush feature seems to provide better results than similar tools like Runway, allowing for more natural and seamless animations of the selected areas. Users can upload an image, select the portion they want to animate, and Verse's AI will bring that selection to life.

This tool provides another powerful way for creators to add animation and movement to their images, without having to animate the entire scene. The selective nature of the Magic Brush makes it a versatile tool for a variety of use cases, from visual effects to creative projects.

This week saw some exciting developments in the world of AI-generated audio. Two notable announcements stood out:

1. 11 Labs' AI-Generated Sound Effects: 11 Labs showcased their new feature that allows users to prompt any sound effect, which the AI then generates. Examples included an "ogre saying 'stay away, puny human'" and a unique sound effect that resembled a Warcraft-style creature.

2. Stability AI's Stable Audio Model: Stability AI released an open-source model called Stable Audio, which can generate up to 47 seconds of audio samples and sound effects, including drum beats, instrument riffs, ambient sounds, and production elements. The audio quality demonstrated in the examples was quite impressive.

These advancements in AI-generated audio highlight the rapid progress being made in this field. Users can now prompt specific sound effects or audio samples, and the AI models are able to produce high-quality, realistic results. This opens up new possibilities for audio creation, sound design, and even audio post-production in various industries.

As these tools continue to evolve, we can expect to see even more impressive and versatile AI-powered audio generation capabilities in the near future.

Jensen Huang, the CEO of Nvidia, made several significant announcements during the Computex event. Here are the key highlights:

1. Earth 2: Nvidia unveiled Earth 2, a digital twin of the entire Earth designed to help better predict climate change and weather. It can do hyper-local forecasting down to tens of meters, trained on vast amounts of weather data.

2. Nvidia Aces: Nvidia showcased its suite of digital human technologies, enabling real-time path-traced subsurface scattering to simulate the way light interacts with skin, giving it a soft and translucent appearance.

3. GPU Performance and Efficiency: Nvidia demonstrated that its GPU compute power is far exceeding Moore's Law, while the power consumption has been dropping significantly, enabling more efficient AI processing.

4. GPU Roadmap: Nvidia outlined its GPU roadmap, with the upcoming Blackwell, Reuben, and future generations, planning to release a new GPU every year to drive continuous advancements.

5. Project G Assist: Nvidia introduced Project G Assist, an AI-powered assistant that can help gamers by answering questions and providing guidance while they are playing video games.

6. Nvidia's Market Position: Nvidia briefly surpassed Apple to become the second-largest company in the world, highlighting the growing importance of its GPU technology in the AI era.

These announcements showcase Nvidia's continued leadership in the field of AI, from its advancements in digital twins and digital humans to its roadmap for even more powerful and efficient GPU hardware. The company's focus on driving AI innovation is evident across its product portfolio and future plans.

AMD made some major announcements at Computex, including their next-generation laptop processor, the Ryzen AI 300 series. This chip features AMD's XDNA 2 NPU, which they claim has 5 times more compute capacity and twice the power efficiency compared to the previous generation. The Ryzen AI 300 will be coming to some of the co-pilot PCs starting in July 2024.

Intel also unveiled their Lunar Lake client processor architecture, continuing to grow the AI-powered PC category. They showcased their "AI Playground" which includes an image generator using stable diffusion models, as well as an "Answer" section that provides a ChatGPT-like large language model running locally on the user's computer.

The key takeaway is that all the major chip manufacturers - Nvidia, AMD, Intel, and Qualcomm - are focused on developing hardware specifically optimized for AI processing. This reflects the increasing importance of AI capabilities in consumer and enterprise computing. The new chips and technologies announced at Computex are aimed at enabling more efficient and powerful AI applications on a wide range of devices.

Cisco's focus at their recent Cisco Live event was on improving "digital resilience" - the ability of companies to handle issues that may arise in the digital world, such as hacks, cybersecurity threats, and data integrity problems.

Cisco is using AI to help enterprises better monitor and manage their digital infrastructure. They have developed a tool called ThousandEyes, which uses AI to keep an eye on a company's entire digital environment, alerting them to problems and helping them quickly identify the source.

While consumers may not directly use Cisco's technologies, the enterprise companies that provide the services and tools we use likely rely on Cisco's infrastructure. By enhancing digital resilience through AI, Cisco aims to improve the overall security and reliability of the digital systems we all depend on.

In addition to developing its own AI-powered tools, Cisco announced a $1 billion global AI investment fund to support the growth of innovative AI solutions in this space. The company recognizes the vital role AI will play in ensuring the safety and stability of our digital world going forward.

Apple's upcoming Worldwide Developers Conference (WWDC) is expected to be a major event for AI announcements. According to reports, the tech giant is planning to unveil a range of new AI features and capabilities across its product lineup.

One of the key expectations is the introduction of a revamped "Apple Intelligence" platform, which will likely replace the current Siri artificial intelligence. The new system is expected to offer significant improvements in natural language processing, task completion, and integration with Apple's ecosystem.

Additionally, Apple is rumored to be integrating more advanced AI capabilities into its core products, such as the iPhone, iPad, and Mac. This could include features like improved image recognition, enhanced voice commands, and more intelligent personal assistant functionalities.

The company is also expected to showcase advancements in its augmented reality (AR) and mixed reality (MR) technologies, which are likely to leverage AI for tasks like object recognition, scene understanding, and seamless integration with digital content.

Furthermore, Apple may unveil new developer tools and APIs that will enable third-party app creators to leverage the company's AI capabilities within their own applications. This could lead to a surge of AI-powered experiences across the Apple ecosystem.

Overall, the expectations for Apple's WWDC event are high, with the potential for significant AI-driven innovations that could shape the future of the company's products and services. As the tech industry continues to prioritize AI development, Apple's announcements will be closely watched by both consumers and industry analysts alike.

When a hacker developed a tool to extract data from Microsoft's new "recall" feature, it raised concerns about privacy and data protection. In response, Microsoft has made several updates to address these issues:

• The recall feature will now be turned off by default, requiring users to specifically enable it.
• Proof of presence will be required to view the timeline and search the recall data.
• Additional data protection measures will be added, including just-in-time decryption and local storage of snapshots (not in the cloud).
• Users will have more control to pause, filter, and delete what is saved in the recall feature.

Microsoft is clearly trying to address the "creepy factor" and ensure users have more transparency and control over their data.

Google is also exploring a similar "memory" feature for Chromebooks, and they too are aiming to eliminate the potential privacy concerns around such a feature.

The key takeaway is that as these AI-powered productivity features become more prevalent, tech companies are having to carefully balance the benefits with robust privacy safeguards. Developers will need to be proactive in addressing potential misuse or exploitation of these technologies.

The proposed California bill, SB 1047 (Safe and Secure Innovation for Frontier Artificial Intelligence Models), is raising concerns among AI innovators. The key points of contention are:

1. Frontier Model Division: The bill creates a "Frontier Model Division" responsible for setting safety standards for AI models. This division would be funded through fees and fines levied on AI developers.

2. Liability for AI Developers: The bill requires anyone training a "covered AI model" (models with more than 10^26 floating-point operations) to certify under penalty of perjury that their model will not be used to enable a hazardous capability in the future, including by others. Developers who don't submit this certification must provide annual assurances to the Frontier Model Division, again under penalty of perjury, that they will mitigate risks of the model or any model built on top of it.

3. Concerns for Open-Source Development: This liability clause makes it difficult for open-source model developers to anticipate all potential future uses of their models, potentially stifling innovation.

Prominent figures in the AI community, such as Clem from Hugging Face, Andrew Ng, and Yan LeCun, have spoken out against this bill, arguing that it would seriously hinder the advancement of AI technology. The main concern is that holding developers responsible for the unpredictable future use of their models is unreasonable and could discourage the development and sharing of open-source AI models.

The proponents of the bill argue that it is necessary to ensure the safe and responsible development of powerful AI models. However, the AI community believes that this approach may do more harm than good, potentially slowing down progress in a field that is rapidly evolving and has the potential to bring significant benefits to society.

This week saw a flurry of exciting AI-related announcements and developments:

Tribeca Film Festival to Screen AI-Generated Short Films

The Tribeca Film Festival is set to showcase AI-generated short films created using Anthropic's Sora system. It will be interesting to see the audience's reaction - whether they embrace these AI-powered creations or if they are met with skepticism like at previous events.

11 Labs Releases AI-Generated Sound Effects

11 Labs unveiled a new feature that allows users to generate custom sound effects using AI. Examples shared on Twitter demonstrate the system's ability to produce convincing sound effects like an ogre's voice or a train horn.

Stability AI Releases Stable Audio

Stability AI introduced Stable Audio, an open-source model for generating audio samples and sound effects. The model can produce a variety of sounds, from synthesizer riffs to ambient nature sounds, showcasing the rapid progress in AI-powered audio generation.

Nvidia, AMD, Intel, and Qualcomm Announce AI-Focused Chips

At the Computex event, major chip manufacturers unveiled new processors designed specifically for accelerating AI workloads. These include Nvidia's Hopper GPUs, AMD's Ryzen AI 300 series, Intel's Lunar Lake, and Qualcomm's Snapdragon X Elite and X Plus chips. The focus on AI-optimized hardware underscores the growing importance of AI in consumer and enterprise computing.

Microsoft and Google Explore "Recall" Features for PCs

Following the discovery of a security vulnerability in Microsoft's "Recall" feature for co-pilot PCs, both Microsoft and Google announced plans to implement similar functionality on Windows and Chromebooks. These features aim to provide a history of user activity, but with added security measures to address privacy concerns.

California Bill Proposes Strict Regulations on AI Models

A proposed bill in California, SB 1047, has raised concerns among AI developers. The bill would require certification and liability assurances for powerful AI models, potentially stifling innovation in the field. Prominent figures in the AI community have voiced their opposition to the bill's provisions.

These developments highlight the rapid pace of AI innovation, as well as the emerging challenges around regulation, security, and the societal impact of these transformative technologies.