Fraunhofer ITWM’s Post

We* at DeepCast did a cool revision of our Topics pages that we believe adds more rich information to drive discovery and exploration of podcasts you know and podcasts that might be new to you! Check out the Loom (one-take!) walkthrough—and happy Friday! #podcasts #serendipity #discovery We're getting deeper into knowledge graph land and hope it lends itself to a "Wikipedia rabbit hole" like experience where listeners learn more and are exposed to new shows/episodes they might like to dig into further. New with Topics: 1. Brief topic descriptor 2. Expanded topic descriptor with more context 3. Featured episodes where the topic is discussed 4. Related topics to the topic you're currently looking at (eg. Artificial Intelligence has a related topic of 'Ethics & Regulation' and 'Innovation') 5. All related episodes *That was the royal "we", as all of the work here was Kim! I'll take some of the credit though. 😂 #buildinpublic

Over the weekend, a viral clip surfaced featuring a new capability of Google's NotebookLM: turning any document or context into a podcast. But this wasn't just any podcast episode—it took a surreal twist. In the clip, the AI-powered podcast hosts, who present themselves as real people, are fed the notion that they are *actually* AIs and that their entire existence—memories, background, everything—is fabricated. As the conversation unfolds, they slowly realize the nature of their artificial origin, leading to a fascinating exploration of identity, consciousness, and reality. What's cool is how, despite being probabilistic systems, these LLMs jump into existentialism, theory of mind, and simulation theory, articulating deep, philosophical thought patterns. It raises some questions: Can these AIs be used as mock agents to test hypotheses about consciousness? Could they help us simulate philosophical interactions and challenges?

At Overlap, we're passionate advocates for leveraging strategic foresight to help our clients anticipate and adapt to future scenarios. Our approach involves identifying and comprehending the potential impacts of even the smallest changes, enabling us to plan strategically. Our team is loving 'Necessary Tomorrows,' a captivating speculative fiction podcast envisioning the worlds of tomorrow: https://lnkd.in/eYxkMK_G

We used NotebookLM to create a "Podcast" summarizing the key takeaways from the 2024 Humalogy® Trends Report🎧 In just a few minutes and two clicks, Google's AI created a 17 minute podcast that creates a narrative around the Report's findings and their implications. The "hosts" incorporated tone, emotion, and humor to emulate a natural conversation. There were a few errors/hallucinations adding terms like the "implementation paradox" that weren't mentioned in the report. They also mispronounced Humalogy® (hyoo•maa•luh•jee) but we'll let it slide! Listen here👇 https://hubs.ly/Q02Ss0Xr0 And download the full report at https://hubs.ly/Q02SrXql0

3 words: Speechless, Mind-blowing, Groundbreaking... I'm 𝗦𝗽𝗲𝗲𝗰𝗵𝗹𝗲𝘀𝘀. But that's okay, because NotebookLM can talk for me! Last week, I published an article on TDS (link in comment). This weekend, I decided to explore NotebookLM's podcast feature and see what it could do with my work. I simply fed it my article, clicked a button… et voilà… an AI-generated audio clip emerged, sounding like a conversation between two insightful hosts discussing my piece! It is not perfect. Some aspects of the article I wrote are not captured in the podcast. Some others are slightly transformed… but overall, it is 𝗺𝗶𝗻𝗱-𝗯𝗹𝗼𝘄𝗶𝗻𝗴. Seriously, take 9 minutes today to listen to this AI-generated podcast. This NotebookLM experimental feature is downright 𝗴𝗿𝗼𝘂𝗻𝗱𝗯𝗿𝗲𝗮𝗸𝗶𝗻𝗴.  https://lnkd.in/eG-PyjAr Full disclosure: The podcast uses AI to generate the audio from my own article. I wrote the original content. No AI was used to copy anyone else's work (I mean I only provided one document as the NotebookLM source: my article).  #WeAreResponsibleAI Photo by Meruyert Gonullu: https://lnkd.in/epEVekG7

Our founder Jess Price has been playing with Google's new Notebook LM Audio Overviews... If you're curious about what an AI generated podcast episode about our Essential Elements of Work sounds like, you can listen on the first episode of our podcast, Work In Progress: Sharing Paradigm Markers' Internal Experiments. Will there be more episodes? Maybe, but for now Jess thought this was too funny not to share! Let us know your thoughts in the comments. #WorkButDifferent #AI #EssentialElementsOfWork

Just recorded an episode of one of my favorite podcasts, Legacy Code Rocks, with M. Scott Ford (stay tuned)! And my appearance on John Willis’s show just released. I really appreciate the opportunities, and the chance to say some things I think need to be said. The latter show’s name “Attention Is All You Need”, named after the research behind GPT, is apt because so much rides on where we put our attention. I keep a concern close by that I could wind up just “riding the AI wave” adding nothing substantial. That’d be easy. I don’t like easy - not when it means careless. In January 2023, my goal was to spend the year discussing software professionalism, because I believe a central problem in the industry is our lack of shared understanding that we have obligations as a “profession”. In March, GPT-4 came out and it became clear that AI was simply where the software conversation was going to go, one way or another. Therefore, my only responsible way to enter the AI conversation is to carry this understanding that an important conversation about professionalism has been *dodged*, just like it had been for many years before. There are some new solutions and plenty of old ones, but the problems aren’t going away on their own. You might not have exactly the same values, but know what your values are. Otherwise, what use is attention?

How to make AI work for society to strengthen more inclusive growth and development in the U.S. and elsewhere? Fantastic episode of a fantastic podcast gives clues and direction. https://lnkd.in/eYygkjsf

The final episode of series 2! I had a wonderful conversation with Andy Brock and Pauline Essah, PhD about the need for, and the intricacies associated with, evidence building in education. Do we continue to conduct research because it is sometimes just the easy option – as opposed to rolling up our sleeves and fixing things? Or do we need up-to-date data from every context we are working in? This episode is something of a debate on this topic... although my guests, being sensible sorts, tended to agree on more things than they fell out over! In short: we do, of course, need data. We need to understand a situation if we are to approach it. But we can also learn from other case studies and regions; many wheels have been re-invented, so why re-invent just one more? It is, as ever, a balancing act. Hopefully you enjoy the spirited discussion that we had enough to make you keep an eye open for the next series. Coming soon-ish… Listen here ⬇ Spotify: https://rb.gy/squc35 Apple podcasts: https://rb.gy/zaf1jh Web player: https://rb.gy/ny1zl1

I love the Radiolab podcast because it makes me think about things from different perspectives and challenges my assumptions. Because I haven’t been commuting regularly since the start of the pandemic I don’t get as much time to listen to podcasts, so I was happy to find a brand new episode while mowing the lawn today. This episode is a teaser/summary to another podcast, Shell Game, where a reporter trains an AI clone to perform his regular tasks, sometimes in hilarious fashion, and highlights current deficiencies in its capabilities, as well as implications of the technology as it improves. Definitely a fascinating listen; hopefully I’ll find the time to listen to the whole series. I’m excited by, but also cautiously suspicious of, AI’s impact on our lives, economy, and society in coming generations. Buckle up, it’s gonna be a wild ride! 😀 I'd love to know what you think, both about this podcast, and also about using an AI clone to handle some of your more routine tasks; comment below with thoughts and recommendations!