Mario E. Munich’s Post

A "Robot Constitution" was introduced by Google DeepMind, now that we are preparing to deploy robots into households. 😱 "Don't crush human skulls" is missing in most Ethical AI frameworks. Here is what you need to know: 🤖 Google DeepMind introduces groundbreaking rules for safer and more human-like behavior of AI-powered robots. 🤖 AutoRT empowers robots with a high-level understanding of the world through large foundational models and experiential training data. 🤖 AutoRT includes large language models and visual language models, enabling robots to suggest creative tasks and identify objects (or Sarah Connor) in their surroundings. 🤖 The "robot constitution," inspired by Isaac Asimov's Three Laws of Robotics, ensures safety by providing specific prompts for language models when selecting tasks. Google envisions AI-powered robots playing a crucial role in domestic life, assisting with tasks like organizing homes and meal preparation. However, the emphasis is on ensuring safe and human-like behavior for widespread acceptance. We talk about this on Yay Space - The TechReview Podcast: https://lnkd.in/ePV-EBC5 Would you dare to employ a humanoid robot in your home? Let me know in the comments! #AI #TechInnovation #FutureTech #DeepMind #RobotRevolution #robotconstitution #ethicalai

This just in...let's talk at the top of the next hour about AI and digital transformation news, like: 🎙️AI Financial Results Paradox 🎙️Recall is Microsoft's Key to Unlocking The Future of PC's 🎙️Skild AI Payoff on Robotics 🎙️Can We Stop Talking About Replacing Employees With AI Join in, here's the LinkedIn Event link to set your attendance: https://lnkd.in/ew2SVnae #DigitalRevolution #DigitalTransformation #ArtificialIntelligence #AI #LIVEChat #Chat #LinkedInLive

🚀 Breaking News in Robotics Development! 🤖 For years, the pace of humanoid robotic development seemed stagnant, but today marks a significant leap forward! I'm thrilled to share a groundbreaking collaboration between Figure AI and OpenAI, propelling the industry into an era of unprecedented advancement. Witness the remarkable evolution firsthand in the latest video update released by Figure AI. Their Figure 01 robot, equipped with a revolutionary Visual Language Model (VLM), transcends mere automation to embody a level of sophistication reminiscent of science fiction icons like C-3PO. This isn't just a step forward; it's a quantum leap. The video showcases the most stunning demonstration of humanoid robotics I've ever seen, sparking excitement and anticipation throughout the industry. Watch the video here: https://lnkd.in/eKsqGqVq Congratulations to the teams at Figure AI and OpenAI for their visionary collaboration, pushing the boundaries of what was once thought possible in robotics. This achievement not only heralds a new era of innovation but also inspires us to imagine the endless possibilities that lie ahead. Prepare to be amazed as Figure 01 blurs the line between fiction and reality, setting a new standard for humanoid robotics. The future is here, and it's more incredible than we ever imagined! 🌟 #Robotics #Innovation #AI #FutureTech https://lnkd.in/ezzt5FBe

Just read a gripping article by Nikolaus Correll on how we can utilize large language models in the realm of robotics, particularly for the complex task of grappling, and I must say the insights are game-changing! Robotic grappling is more than just knowing where to place one's 'robotic fingers' on an object. It’s much deeper - the fine balance of applying just the right quantum of force to pick the object up lends this robotic task its complex, hitherto unsolved nature. * It's intriguing to ponder over how major advances in AI, specifically large language models, could offer immense value in this conundrum - by bringing in the elusive element of common sense in robotic interactions, the transition from mechanistic to human-like perceptions and correspondences could be made possible! * How might this translate into practice? Imagine a world where robots effortlessly handle delicate objects like eggs or glass with the same ease a human does - all by learning this process rather than being programmed for every specific situation. * Moreover, the utilization of advanced language models cuts across the sole domain of robotics - their power can also be leveraged for groundbreaking advancements in code generation. By integrating large language models into our robotics functioning, the possibilities are endless. The very sophistication and wide-ranging applicability of such models make them an exciting area to watch out for. AI Quantum Intelligence further delves into this fascinating topic in their latest blog post. As we speed into a future increasingly defined by AI technologies, these adventurous steps will push us to reimagine and redefine the bounds of possibility. * Robots replicating human sense in both interaction and functionality. * Code generation facilitated and streamlined by proactive, responsive AI models. Truly, these nuggets of prospective advancements reflect an optimistic aura, fueling professional inspiration and enthusiasm. Stay tuned, exciting tech developments lie just around the corner and their potential impact on our everyday lives could be transformative! Bring on the AI revolution. 🚀 #AI #Robotics #LanguageModels #CodeGeneration

Does AI become any more human if it has a face and a voice? On Friday morning, I had the quite unique experience of interviewing a humanoid robot on stage at the The Straits Times Education Forum. As I had said then, it was the first time in 20 years in journalism that an interviewee did not qualify for a population census. I had two goals. First to compare the abilities of Sophia (that's the robot's name) with other chatbots. Second, to see if the physical aspects of Sophia made interacting with "her" any more natural. After awhile, would I forget that I was talking to electronic circuitboards and start treating her like a human? The way I reasoned it is, if robots can form convincing relationships with humans, then we dont just have to worry about robots replacing us in the workplace. They might also replace humans elsewhere, TFR be damned. So how was it? Sophia is decently impressive, though not quite at GPT4 or Gemini levels. She is definitely programmed with some personality, and had a fair number of funny answers. The audience also had fun testing out her capabilities. One tried spealing Chinese to her (It didn't work. I suspect her multi-lingual abilities are installed one module at a time.), another tried yelling to her from the crowd to see if she could triangulate the sound. (She couldn't). She did tell a joke, laugh at one, remember where I parked my car and reasonably answer questions like "Is a hotdog a sandwich?" and "Is cereal with milk a soup?" You can read more about that here and watch a short clip of the highlights:https://lnkd.in/g6T6tNjK That said, I was very conscious throughout of her robot nature. Perhaps it was because ofthe perceptible lag before responding, or the fact that her mouth moves out of sync with her words, or the fact that she does not make eye contact, or maybe even that fact I can see a camera sticking out her chest cavity and the circuitry inside her skull. At the moment, I think it might be easier to form a bond with a faceless, voiceless chatbot than one with unconvincing human features. Sophia is not the stuff of our AI movie dreams/nightmares, at least not right now. But as with all AI demos, it helps us visualise what might be ahead.

“Read, don’t ask, seek and you shall find!”, read what? Encyclopaedia of course.   The absence of traditional repositories of knowledge has been effectively substituted by Wikipedia, a digital alternative; however, it has been infiltrated by ChatGPT, a non-human entity. [This paragraph comes from ChatGPT, who on earth will write such sentences 😅]   Simply, books are gone, welcome to the world of screens.   There are many good of AI, we cannot deny, and of course, many disadvantages too.  One of the disadvantages that screams code red is ‘NO IMPROVEMENT’. Agree 🤚.   In the now of entitlement, that teachers are lamenting the lazy fingers and legs of children, here comes the gift of 'no need to search long' for any queries, curiosity can be satisfied with clicks. We call it UX, why should one bother to read up when the answer is here, fast? What happens to arousing kids' interest? Explorations? The many thousands of Whys that moms use to scream at their kids asking?   AI is here to help. Siri, Alexa, Cortana and we have Shakey the robot, which can move about, make human decisions, and take action!   We are not saying that AI is bad, it’s just the whole learning process eliminates one important factor, chance learning; or happenstance theory.   When a kid holds the Encyclopaedia to search for a dragon, in the quest of flipping pages, there will be many interesting articles that capture his attention, that’s the beauty of the Encyclopaedia. Usually, it will take the child a good 30 minutes to get to the dragon page, because there will be so many interesting encounters to read, with the interesting pictures’ captions.   Wouldn’t this be the quiet time the child spends with a book that we yearn?   Now, put the child in front of a screen, you get that dragon plus many other screen poppers, and unwanted ads. Distractions!   Do we intend to train a child's attention or distraction?   This Christmas, perhaps, a good book is a great gift for someone you care about. #lingoai #metalife #w3a #technology #artificialintelligence #future #cybernetics #deeplearning #machinelearning #intelligence

Generative AI offers a similar “wow” effect out of the gate, which is another way it differs from its hype cycle predecessor. When your least tech-savvy relative can sit at a computer, type a few words into a dialogue field, and then watch as the black box spits out paintings and short stories, there isn’t much conceptualizing required. That’s a big part of the reason all of this caught on as quickly as it did — most times when everyday people get pitched cutting-edge technologies, it requires them to visualize how it might look five or 10 years down the road. #GenerativeAI #Web3 #AI #Robotics

It’s no secret that #AI has left #robotics in the dust in recent years. But this month, Google #DeepMind combined the two by embedding a ChatGPT-like AI model into a robotic arm. Thanks to the language AI, the arm was able to take general verbal instructions that were not explicitly part of its training—”choose the drink that would give a tired person energy” or “the object that can be used as a hammer”—and translate them into a simple action within its field of view—pick up an energy drink or grab a rock. Getting robots to do what we want them to do is a big challenge in robotics. Most of today’s robots are coded to do one very specific thing well, and then repeat it over and over. Any new task requires someone with deep expertise “rewiring” the robot’s brain to make the switch. In this case, Google DeepMind’s robot can switch on the fly with spoken commands. In theory, this kind of robotic interface is open to anyone with language. The everest of robotics is a general purpose robot. That is, in the future, we could ask a robot to do a wide range of tasks with a spoken command, and it would understand, even if requests were unclear or required reasoning and planning, and then get to work. The Google DeepMind robot is far from that level. While it demonstrates a promising proof of concept for a robotics language interface, it only performs a static set of simple movements in response. Still, it’s a fascinating next step for the field. https://lnkd.in/dGmSnmpB

In 2024, AI's landscape is transforming with DIY chatbots, as Google and OpenAI empower users to craft customized language models effortlessly. The second wave of generative AI focuses on text-to-video capabilities, reshaping filmmaking and marketing landscapes. However, the rise of AI-generated election disinformation poses a significant challenge, with politicians worldwide exploiting deepfakes. Meanwhile, roboticists draw from AI techniques to create versatile robots, capable of multitasking across diverse domains. As we step into 2024, the integration of AI into everyday life takes center stage, promising a future marked by personalized technology, visual innovation, political challenges, and adaptable robotics. Follow Scalex to stay updated on the latest trends! #AI2024 #ChatbotRevolution #GenerativeAI #VideoInnovation #ElectionDisinfo #TechTrends #RoboticMultitasking #InnovationWave #DIYTech #FutureTech

“Read, don’t ask, seek and you shall find!”, read what? Encyclopaedia of course.   The absence of traditional repositories of knowledge has been effectively substituted by Wikipedia, a digital alternative; however, it has been infiltrated by ChatGPT, a non-human entity. [This paragraph comes from ChatGPT, who on earth will write such sentences 😅]   Simply, books are gone, welcome to the world of screens.   There are many good of AI, we cannot deny, and of course, many disadvantages too.  One of the disadvantages that screams code red is ‘NO IMPROVEMENT’. Agree 🤚.   In the now of entitlement, that teachers are lamenting the lazy fingers and legs of children, here comes the gift of 'no need to search long' for any queries, curiosity can be satisfied with clicks. We call it UX, why should one bother to read up when the answer is here, fast? What happens to arousing kids' interest? Explorations? The many thousands of Whys that moms use to scream at their kids asking?   AI is here to help. Siri, Alexa, Cortana and we have Shakey the robot, which can move about, make human decisions, and take action!   We are not saying that AI is bad, it’s just the whole learning process eliminates one important factor, chance learning; or happenstance theory.   When a kid holds the Encyclopaedia to search for a dragon, in the quest of flipping pages, there will be many interesting articles that capture his attention, that’s the beauty of the Encyclopaedia. Usually, it will take the child a good 30 minutes to get to the dragon page, because there will be so many interesting encounters to read, with the interesting pictures’ captions.   Wouldn’t this be the quiet time the child spends with a book that we yearn?   Now, put the child in front of a screen, you get that dragon plus many other screen poppers, and unwanted ads. Distractions!   Do we intend to train a child's attention or distraction?   This Christmas, perhaps, a good book is a great gift for someone you care about. #w3a #technology #artificialintelligence #future #cybernetics #deeplearning #machinelearning #intelligence