Guido Jouret’s Post

A game-changing development in the AI industry - researchers at MIT are bringing the concept of common sense to robots. By allowing them to self-correct and overcome disruptions independently, this leap forwards represents an exciting development. At Marketing Automaton, we believe that an integration of this type of AI innovation with digital marketing processes can push the boundaries of what's achievable in business. Do you think this could be a game-changer for your enterprise? Click on the link below to dive into the details. #AIInnovation #FutureOfMarketing

Figure 01 Robot - Science Fiction or Reality? Figure 01 robot features below: Design Features   - Figure 01 is a humanoid robot designed to be the world's first commercially viable general-purpose humanoid robot   - It features a bipedal platform called Neo, which uses electric actuators, a screen for a face, and humanoid hands   - The robot is designed to have a human-like body with advanced AI capabilities Functionality - The collaboration between Figure and OpenAI involves building specialized AI models for Figure's humanoid robots, enabling them to process and reason from language   - Figure 01 has demonstrated dynamic bipedal walking and the ability to perform tasks autonomously, such as picking up objects and placing them on conveyor belts Technological Advancements   - The robot is part of a new generation of robots that aim to revolutionize labor markets and the global economy through advanced AI integration   - It leverages cutting-edge AI technology from OpenAI, including GPT language models Future Developments   - Figure 01 is part of a broader initiative by Figure AI to develop fully autonomous commercial robots called "humanoids" for various industries and applications, including space exploration   - The company aims to replace human workers in unsafe or undesirable jobs with these advanced humanoid robots Could we soon be chatting and sharing tasks with robots as if they were another member of the family? Let me know what you think in the comments. ----- Follow me for AI insights, hit the bell for updates, and share to spread the word! ---- #openai #figureai

🤖🥪 Will AI eat my lunch? Discover Universal Field Robots Founder and MD Jeff Sterling's thoughts about the increasing shift in the number of everyday tasks that incorporate varying degrees of artificial intelligence in our latest blog post. "This subtle AI transformation need not be feared, but as they say, education is power. I would like to attempt to provide a primer to AI, help separate fact from fiction, and suggest some tips and considerations to brave the new world with AI." 💻 Learn more > https://lnkd.in/g9XDdxES #ArtificialIntelligence #AI #Technology #Robotics #Automation #Resources #Mining #Safety #Security #Efficiency #Productivity #Sustainability #UniversalFieldRobots Dush Daniel Kerry Kimberley

🤖 Robot Whisperers: When Machines Listen, Do They Truly Understand? 🤔 Ever chatted with Siri, Alexa, or any voice-activated robot? We often marvel at how they can 'listen' and 'respond'. But here's a fun thought: How do we know they're not just parroting back without grasping the essence of our words? 📜 In the world of robotics, there's a fascinating challenge: • Robots can be programmed to repeat sequences of words (called N-grams) from our instructions. It's like teaching a parrot to mimic phrases. • But ensuring that they capture the meaning behind those words? That's the real game! Imagine telling a robot, "Lift the red block." If it responds, "Raise the block that's red," it's not an exact match in words, but the essence remains the same. While tech metrics can spot word differences, gauging true understanding is a deeper challenge. 1️⃣ Parroting vs. Understanding: When we talk to robots, they often 'listen' and 'respond'. But are they just expertly mimicking us or genuinely understanding? 2️⃣ N-grams: In tech lingo, robots can be trained to recognize and repeat short sequences of words from our instructions, known as N-grams. It's akin to teaching a parrot to mimic phrases. 3️⃣ Statistical Metrics vs. Semantics: While we have statistical tools to check if robots use similar word sequences (like N-grams), the real challenge is in semantics. Semantics is about ensuring they capture the meaning behind those words. For instance, "Lift the red block" and "Raise the block that's red" might differ in words but share the same essence. #robotics #ai #research

Let’s be real Arnold Schwarzenegger as an AI chatbot would be way cooler than Elon Musk’s recently announced sarcasm-infused “Grok” for X (Twitter). Schwarzenegger said the development of artificial intelligence has made his sci-fi classic “The Terminator,” a movie in which an AI-based defense network known as Skynet becomes self-aware and conquers humanity, more truth than fiction. “It’s not any more fantasy or kind of futuristic,” he said. “It is here today.” But is it? Clearly, robots and artificial intelligence have started to blend and work together. But before we deem Schwarzenegger’s apocalyptic prediction true, let’s stay for a minute in our positive world. Let’s talk about how robots and AI work together to improve society. We’ll worry about our robot overlords when they come for us. Full blog 🤖 https://lnkd.in/eQcCMFAY #AI #ArtificialIntelligence #Grok #Elonmusk

AI robots are revolutionizing industries from manufacturing to healthcare, enhancing efficiency, safety, and innovation. Discover how robots like ASIMO, Pepper, and Digit are transforming our world in this insightful article. For example, Manufacturing companies use AI robots for tasks like assembly, welding, painting, and quality inspection. These robots increase productivity, reduce errors, and enhance safety by performing repetitive and dangerous tasks. Advanced AI algorithms allow these robots to learn and adapt to new tasks, making them versatile and valuable assets in manufacturing environments. It's very important to mention, that every humanoid developed must be part of an end-to-end program that fundamentally includes how it will assist current human workers and what alternatives will be provided for those displaced by these robots. It's essential to ensure that the introduction of humanoids benefits all stakeholders and addresses potential job displacement issues effectively. Below is the list of Humanoid AI-robots in the article: - ASIMO by Honda - Research in human-robot interaction - Pepper by SoftBank Robotics - Retail, hospitality, healthcare - Atlas by Boston Dynamics - Search and rescue missions - Sophia by Hanson Robotics - Media, entertainment, companionship - Nao by SoftBank Robotics - Education, research - Romeo by SoftBank Robotics - Domestic care for the elderly and disabled - Digit by Agility Robotics - Logistics and delivery - HRP-5P by AIST - Construction and industrial tasks - Walker by UBTECH Robotics - Home and office environments - Talos by PAL Robotics - Research, industrial applications #AI #Robotics #Innovation #Industry4.0 #HumanAndRobot #Humanoids

Google is working on a new artificial intelligence model for garbage-collecting robots, and with it, it explains that different artificial intelligence models can be developed for various purposes. That is, artificial intelligence models are no longer just imitating human behavior, but are also becoming usable in different fields of expertise. #robots  #google #ai  #ml  #artificialintelligence

This is a must read for current and future roboticists who are struggling to keep up with current trends of A.I. and LLMs, ViMs. It is a fantastic summary of the interactions and developments of A.I. and traditional robotics. I also want to share my views on the still nuanced concept of using general/foundation models for performing specific robotics tasks in an unstructured environment. With the advent of chatGPT and other LLMs, we have seen the explosive potential for LLMs to respond to general inquiries with natural language. This capability has largely expanded into multimodal inputs with GPT4 and DALL•E. A lot of people have naturally asked the question of whether we can modify the output of these models to control joints, manipulators and mobile bases. There is no doubt that these models have an impressive understanding of language and images, but these data are inherently lower dimensional than the data generated by robots that operated in 3D space and a time horizon. I think that at least at the very first stage, we will see robots that employ these models to translate camera and text/audio data into high level commands and relinquished the low level controls to more traditional model-based approaches. For example going from language -> high level tasks -> position and trajectory of joints and graspers -> motor commands. I think the focus should be put on the task at hand rather than actuating the robot in a very complex state space. Since the latter would require the models to have a fundamental grasp on physics, space and dynamics of interaction with the environment which even the most current state-of-the-art models have largely failed to demonstrate. I, myself have started exploring the potential of using A.I. and deep learning to augment the capabilities of a robot dog, the Unitree Go 2, my plan is to first start using LLMs and ViMs to output text and infer actions from that output. If all goes well, I will try to experiment with using these models to infer task specific actions. https://lnkd.in/gahm8twE