Andreas Fraunberger

#AI: Tools of the Trade - Where Artificial Intelligence Can Make a Real Difference #MPTS2024 Chair: Ben Baldwin, Founder - Chuffed Productions Prof Florian Block, R&D Lead, AI & Immersive - dock10 Governance and ethics retrofitted to technology. Now seeing flip. How train models with considering source to create more ethical products. Adobe creative suite consider training ethics. Smaller creatives on YouTube could innovate without same considerations. Misinformation. Now everyone can produce! Once treated misinformation and copyright, could be solved. What do these mean for us as a culture? Audience changes values. Internships. Will generative AI miss the springboard? Market competition. If using AI in business and making more and better than competition. Work with audience needs. UGC out of control. Kids: everything over 60-90 seconds is too long. Invoke AI. How Photoshop would’ve looked if it started with AI. Carole Pigeard, SVP of Global Sales - Moments Lab (ex Newsbridge) AI analysis to create metadata. But what do you do with that metadata? Newscast can automatically segment or social media, so immediately available. Automatic summaries reduce time to get summary. Next 5 years: scale of use. Are all team members creatives? Need strategy. Alongside cloud and SaaS. More verticalised AI. Archivist can create more social engagement with archive content because AI automating. And more archive-dependent content, e.g. documentaries, could be created. Aaron Bhugobaun, Technical Production Operations Manager - Moonbug Entertainment Children’s content quality expectations. Need to consider ethics even if metadata. Archive doesn’t always have all assets available. AI gets archive in order quickly what would be very time-consuming otherwise. Experimenters could benefit from AI more of the short term as not always considering ethics first. Governance is slow to react. With regulation, AI is changing very rapidly. Loop. Create metadata with AI to train metadata. Changing viewing behaviour to shorter term content. Need enterprise agreements in business for protection. Eg Adobe Firefly.

12

2 Kommentare

Efficiency and Avatars - the name for the new episode of the Making the Media Podcast, featuring Miro Ambra from TRACO SYSTEMS a.s. In the episode we discuss AI in news, including AI presenters, the importance of web-based tools to drive forward journalistic collaboration, and where the Cloud fits (and as important, where it does not) in production. Check out the episode here: https://pod.fo/e/270def #avid #ai #artificialintelligence #podcast

20

1 Kommentar

The ZDF Digital team recently faced a challenge: crafting a high-quality animated intro trailer for a Terra X documentary about “time” in just three days. Using Adobe Firefly, they efficiently bypassed stock libraries, maximizing creative control while maintaining journalistic integrity and meeting the deadline. Ethical AI use remains a priority, aligning with our principles @Adobe. The results speak for themselves: streamlined production, enhanced visuals, and a glimpse into the future of digital media. https://adobe.ly/3WV43BN     #AdobeFirefly #AdobeCreativeCloud

23

#digital #twins is probably the best non-gimmicky #technology that has emerged in the last years. With the advancements in #artificialintelligence, #computing capabilities and #computer #vision, this is one key area that companies should #invest in. #strategic #planning can then take place at a lower #cost and with more faithful representation of a #realistic setup. Digital twins is the ultimate replacement of traditional #simulation techniques so far employed around the world. https://lnkd.in/gaF2SFT8

2

Do you want to feel your virtual reality VR180 images? A new dimension for immersive media opens up. Learn more in our research at ACM VRST 2024. #VirtualReality #ImmersiveMedia #Haptics #ACMVRST2024 #Research #VR180 #SpatialPhotos #SpatialVideos

4

The #AI Pact seeks the industry’s voluntary commitment to anticipate the EU AI Act and to start implementing its requirements ahead of the legal deadline. Over 100 companies have signed including Amazon (European Core), Google, Microsoft and OpenAI, however The Independent notes that Anthropic, Apple and Meta have not signed. The European Commission states that some signatories may not appear immediately. The AI Pact: Two pillars: ▪️Pillar I: Collaborative community within the AI Pact network to share best practices and internal policies. ▪️Pillar II: Foster early implementation of EU AI Act features Three core actions: ▪️Adopting an AI governance strategy to foster the uptake of AI in the organisation and work towards future compliance with the AI Act ▪️Identifying and mapping AI systems likely to be categorised as high-risk under the AI Act ▪️Promoting AI awareness and literacy among staff, ensuring ethical and responsible AI development Additionally recommended tailored activites, e.g.: ▪️Human oversight ▪️Mitigating risks ▪️Transparently labelling certain types of AI-generated content, such as deepfakes These sound sensible.

2

2 Kommentare

Introducing Blum Drop: The Exciting New Game on Blum.io At Blum.io, we’re always looking for innovative ways to engage and inspire our community. Today, we’re thrilled to introduce a brand-new feature that blends fun with functionality — meet Blum Drop! This exciting new game is designed not only to entertain but also to provide a unique experience that aligns with our core mission of data engagement and insight generation. What is Blum Drop? Blum Drop is a dynamic and interactive game that combines elements of strategy and data-driven decision-making. Inspired by the need for creative problem-solving in data management, Blum Drop challenges players to navigate through various data scenarios and make strategic decisions to achieve high scores. In Blum Drop, players are presented with a series of data challenges and must use their analytical skills to solve puzzles, manage resources, and optimize outcomes. The game’s design reflects real-world data management scenarios, making it both a fun and educational experience. Key Features of Blum Drop 1. Engaging Gameplay: Blum Drop offers a range of levels, each with unique challenges and objectives. The game’s mechanics are designed to be both entertaining and thought-provoking, ensuring players remain engaged and motivated. 2. Educational Value:While primarily a game, Blum Drop also incorporates elements of data analytics and decision-making. Players will find themselves applying concepts related to data management, making it a great tool for learning and improving analytical skills. 3. Leaderboard and Rewards: To add an extra layer of excitement, Blum Drop features a competitive leaderboard. Players can track their progress, compare their scores with others, and earn rewards based on their performance. It’s a fantastic way to foster a sense of community and friendly competition. 4. User-Friendly Interface: Blum Drop is designed with ease of use in mind. The intuitive interface ensures that players of all skill levels can jump right in and start playing without a steep learning curve. Why Play Blum Drop? Blum Drop isn’t just about having fun; it’s also about enhancing your understanding of data-related concepts in an engaging way. By playing, you’ll: - Sharpen Your Analytical Skills: The game’s challenges are designed to improve your ability to analyze data and make strategic decisions. - Learn Through Play: Blum Drop offers a hands-on approach to learning about data management, making complex concepts more accessible and enjoyable. - Join a Community: Connect with other players, share your achievements, and be part of a growing community that values both fun and learning. Get Started with Blum Drop Today! Ready to dive into Blum Drop? Head over to https://lnkd.in/gtWJG8Vz and start playing today.

2

🏄‍♀️ Can Krea AI's New Video Generation Feature Dethrone Open's Sora Model? When it comes to AI video generation, Open’s unreleased Sora model has so far this year stolen the spotlight away from compelling rivals Runway and Pika. 👩‍💻 But now comes a new challenger: krea.ai, a lesser-known San Francisco generative AI media startup co-founded by Víctor Perez and diego rodríguez that launched its public beta in late 2023. 🏃‍♀️ Unlike Pika and Runway, which work mainly by adding motion to regions of user-uploaded still images, or converting text prompt descriptions into whole new videos, or even transforming pre-shot video into new styles, Krea presents the user with the option to use both “key frame” still images and text prompts in the same project. 🤹‍♀️ Initially, a low-resolution preview is generated, giving users a quick glimpse of their creation. This is followed by a second phase that enhances the video to high resolution. The startup is now seeking to commercialize its latest AI breakthroughs. Check Out the full article on VentureBeat : https://lnkd.in/gX3nKiPu #AI #KreaAI #VideoGeneration #Sora

3

#AI Super Resolution. For those who saw my #CTVWS and CTAM Europe talks, I spoke about using AI in video super resolution. This can bring new life to library standard definition content by making it high definition for #FAST for instance. Bitmovin is one of the companies working in this space. Andy Francis explains we can now do this using a part of #machinelearning called Deep Neural Networks (DNNs). Bitmovin's ATHENA research lab partner proposes DeepStream--scalable content-aware per-title encoding, which provides higher video quality on GPU-based end-user devices. When combined with DeepCABAC, context-aware binary arithmetic encoding for DNN compression, shows ~35% bitrate savings while maintaining perceptual quality (PSNR and VMAF). (I dug into the paper for a bit more understanding.) They're also partnering with Pixop who do AI upscaling.

16

1 Kommentar

Move over #GPTs... its time for #LFMs to make the next big wave in #GenAI For the uninitiated - most of the current GenAI platforms and tools (ChatGPT, Claude, etc) are based on an architecture that's transformer based and comes huge memory usage and GPU requirements. The LFM approach is a new architecture which (to put it simply) significantly reduces the memory requirements and performs as good as the current models (if not better). To put this in perspective - Liquid’s(the company pioneering LFMs) LFM-3B requires only 16 GB of memory compared to the more than 48 GB required by Meta’s #Llama-3.2-3B model. This is one step closer to making LLMs more accessible, lighter and easy on the pocket, hence increasing adoption. Check out this article ... makes for a good weekend read !

11

1 Kommentar

Reflecting on a great weekend at TEDAI in Vienna. Too much for one LinkedIn post so cherry-picking some highlights: ⌨ Chin-Teng Lin showing wearable tech that can read the words and images you’re THINKING! Will brain computer interfaces make keyboards a thing of past? 🌐 Selena Deckelmann of Wikipedia advocating companies building AI learn from the transparency adopted by Wikipedians. 📜 Digital archaeologists showing how they used pattern analysis to read and decode what was written in unrolled fossilised scrolls. AI isn’t just changing the future - it’s changing the past! ✍ Victor Riparbelli suggesting our grandchildren could be the last generation to read or write as attention spans drop and AI automates video and audio generation. Maybe the written word isn’t morally superior to audio and video? But what about allowing our imagination to fill in the gaps? Discuss! 🎮 Microsoft’s Katja Hofmann showing an AI model trained on 8 years of user gaming data that can help gaming companies test new creative ideas in games, showing AI CAN augment creativity, 🤔 Lots of arguments about regulation hurting small companies, the merits of open-source models v walled gardens, and the negative real-world consequences of unchecked AI like: > Equity (85% of LLMs are based on Latin languages, AI recruitment tools that select candidates based on traits that correlate with gender, ethnicity etc) > Environment (depending on the platform used, 0.5 Litres of water are used to power 10-50 LLM responses!) > And a powerful talk by Northern Irish MLA Cara Hunter on how she was deep faked into a porn video in an attempt to subvert her re-election. 🍃 Raia Hadsell of Google DeepMind encouraging us to think beyond the hype of chatbots to how AI will solve the unknown, from weather forecasting to earth models. 🤦‍♂️ Psychologist Riccardo Loconte on the process of creating an AI lie detector with 80% accuracy, and the implications for trust. 🧠 Oliver Brock on how we tend to use the word ‘intelligence’ indiscriminately, and how the next AI paradigm shift will be active networks, not trained networks. 🤖 An argument by Andrea Lavazza that we should be considering robotic digital successors to continue human civilisation in case of an extinction event. 🌏 Linked to that, Jürgen Schmidhuber on how 2029 will be the year when general purpose robots finally reach the mainstream, and that the world ends as we know it in 2042 as humanity passes the mantle to AI (which expands throughout the universe without the limitations of human biology). Maybe tongue in cheek, maybe not!? 🎤 Oh, and a truly hilarious session by British comic and freestyle rapper Chris Turner in a rap battle against ChatGPT (spoiler: Chris won hands down). I remain a firm techno-optimist from this event- it seems right now for every harm AI poses there are at least 5 applications of AI that improve the world we live in. This AI revolution is truly unstoppable. #AI #TED #TechForGood TEDAI Vienna

55

13 Kommentare

So many people come and reflect their thoughts about the Global XR Hackathon, September 6-8th. Everybody is so positive and people keep asking me for the next step. At the same time, I'm keeping the dialogue with Linus Kimselius, who hacked the ChatGPT-driven bartender at the Hackathon. He was not competing because he was a mentor. We discuss how many cool integrations can be done and new features like multiplayer, other settings, Spotify integration, and so on. So my question is what kind of activity we should aim for next? Meet and play, show potential - what's possible today, or just an after-work? Please help me out, what are you interested in next? Here's a video of the experience. https://lnkd.in/dbBGrKCu

39

5 Kommentare

How can player preferences be used to design better ad creatives for games? 😯 🦉 Game Surgery #1 provides a 2-minute deep dive into this case! 👾 Let's examine a rising star: Tile Family - a relaxing mahjong interior cultivation family drama from Code Dish GmbH. ✏️ We look forward to your comments and questions. Opinions from anyone behind Tile Family are especially welcome 👋 Ben(Xurong) Lin, Emircan Erdogan! Enjoy! 🎯 Powered by GEMBASE.io #MobileGaming #UserAcquisition #GameCreatives #AdCreatives #Marketing #Gaming #MarketResearch #TileFamily

13

1 Kommentar

Check out the latest XReality Pro Podcast #12 episode featuring Wagner James Au, discussing his book "Making A Materverse That Matters". The topics we will discuss: 1. What's Metaverse/metaverse? 2. What are the unique challenges of design in M/m? 3. The history of M/m, what failed? What succeeded? 4. What are the M/m Economy and Monetization (Creator Economy)? 5. How is Gen Z shaping the M/m? And the transition from social media to M/m? 6. Future Direction - What's the next big trend in M/m? If you are interested in the discussion, subscribe XReality Podcast to learn more! Partner with GatherVerse. 🎥 YouTube: https://lnkd.in/g9AgBsau 🔊 Spotify: https://lnkd.in/g2DNDJTg 🍎 Apple Podcast: https://lnkd.in/gtejtpz7 #metaverse #Metaverse #SecondLife #Meta #HorizonWorlds #VRChat #Roblox #CreatorEconomy #Monitization #XR #SocialMedia #Facebook #Entreprenuership #UX #GenZ

30

4 Kommentare

In this interview with Jelmer Althuis, we dive into the importance of Spatial sound👉https://lnkd.in/eNWVkpXS During this episode of XR AI Spotlight we: - Define the real fundamentals of what makes sound sounds real - We look at the production workflow including the recording process, the best libraries and the SDK you can use right now - Look at the impact sound can have on users and inhabitants of virtual worlds

4

👨🏻‍🚀🎨🎥 Luma AI's First / Last Frame feature is ingenious in combination with Midjourney styles: 1) in MidJourney I generated the prompt “portrait of an astronaut, looking directly into the camera, black short stubble hair, round glasses” with different sref styles 2) in Luma AI, I used one image as the first frame and another image as the second frame 3) then I used the second image as the first frame and another image as the second frame 4) and again this second image... and so on and so forth until I used the overall first picture as the last frame 5) I placed all the videos one after the other in FotoMagico and added music generated in Suno. 🎉 Done :-) You’ll find more Styles and Style blends in our Style Code eBook. Check 'AI Imagelab' on Gumroad or follow the link in the comments.

24

10 Kommentare

How i plan a storyboard | videocontent | coding | neuronal network? The basic of all my work and planings are mindmaps. In my picture you see a mindmap about our new project "International Online Academy". The whole Online Academy have 124 videos that are related to each other. It is perfect for creating connections, a workflow for later work and you have a perfect, simple, logically structured overview. And its a perfect way to bring your vision, plan, idea or what ever you plan in a simple overview, where you can change the related items, structure perfectly. Note: Yes, It exists AI mindmap tools. But make sure that you understand AI or mashine learning. The ai knows nothing about your "strategy or Relationships of your elements". You can teach the Ai, but it doesn't make sense for "simple" projects! With deeper experience, you can import mindmaps in a ai. #startup #learning #motorsport #technology

3

1 Kommentar

What is an AR (Augmented Reality) Framework and How Does It Work? Welcome to WikiGlitz! Augmented Reality (AR) is transforming how we interact with the world around us by overlaying digital information onto our physical environment.  In this comprehensive guide, we’ll explore the concept of AR frameworks, how they work, and their applications in various industries.  Join us as we delve into the fascinating world of AR technology. For more expert advice, visit: https://lnkd.in/eKdyfq2U and Follow WikiGlitz #usa #uk #canada #singapore #londan #australia #techcommunity #augmentedreality #edutech #learnskill #vrandar #techblog #expertguidence #virtualreality 

1