Lifeboat Foundation: Safeguarding Humanity
Toggle light / dark theme

Blog

Category: robotics/AI – Page 485

A Technique for more Effective Multipurpose Robots

With generative AI models, researchers combined robotics data from different sources to help robots learn better. MIT researchers developed a technique to combine robotics training data across domains, modalities, and tasks using generative AI models. They create a combined strategy from several different datasets that enables a robot to learn to perform new tasks in unseen environments.

Let’s say you want to train a robot so it understands how to use tools and can then quickly learn to make repairs around your house with a hammer, wrench, and screwdriver. To do that, you would need an enormous amount of data demonstrating tool use.

Existing robotic datasets vary widely in modality — some include color images while others are composed of tactile imprints, for instance. Data could also be collected in different domains, like simulation or human demos. And each dataset may capture a unique task and environment.

Tesla could release FSD v12.4.2 this weekend

The next update to Tesla’s Full Self-Driving (Supervised) could arrive this weekend, as the long-awaited v12.4.2 is scheduled to enter an internal testing phase tomorrow.

Tesla first released version 12 of FSD in March, and it was a significant release because it was the first version that relied on end-to-end neural nets, instead of over 300,000 lines of hand-written code. With the switch, CEO Elon Musk said that each revision should result in significant improvements, saying that v12.4 should see a 5 to 10 times improvement in miles per intervention.

However, v12.4 was only released to a limited number of testers earlier this month, more than four weeks after Musk initially said it would be available, and it received a luke-warm response, with a number of bugs and erratic driving behaviours reported. As a result, it has yet to go to a wide release.

I don’t think we can control AI much longer. Here’s why

Go to https://ground.news/sabine to get 40% Off the Vantage plan and see through sensationalized reporting. Stay fully informed on events around the world with Ground News.

Geoffrey Hinton recently ignited a heated debate with an interview in which he says he is very worried that we will soon lose control over superintelligent AI. Meta’s AI chief Yann LeCun disagrees. I think they’re both wrong. Let’s have a look.

🤓 Check out my new quiz app ➜ http://quizwithit.com/
💌 Support me on Donorbox ➜ https://donorbox.org/swtg.
📝 Transcripts and written news on Substack ➜ https://sciencewtg.substack.com/
👉 Transcript with links to references on Patreon ➜ / sabine.
📩 Free weekly science newsletter ➜ https://sabinehossenfelder.com/newsle…
👂 Audio only podcast ➜ https://open.spotify.com/show/0MkNfXl…
🔗 Join this channel to get access to perks ➜
/ @sabinehossenfelder.
🖼️ On instagram ➜ / sciencewtg.

#science #sciencenews #artificialintelligence #ai #technews #tech #technology

Fighting Rectal Cancer with AI: Researchers Secure $2.78M Grant

What You Should Know:

– A glimmer of hope emerged today for rectal cancer patients as a collaborative effort between Case Western Reserve University (CWRU), Cleveland Clinic, and University Hospitals (UH) received a $2.78 million grant over five years from the National Institutes of Health and National Cancer Institute. This grant will fuel research leveraging artificial intelligence (AI) to personalize treatment for rectal cancer patients.

– The new research effort signifies a significant step forward in the fight against rectal cancer. By harnessing the power of AI, researchers are on the path to developing more precise treatment strategies, ultimately improving patient outcomes and quality of life.

Could AI Ever Become Conscious? Here’s the Truth About Thinking, Feeling Machines

Star Trek: The Next Generation looks at sentience as consciousness, self-awareness, and intelligence—and that was actually pretty spot on. Sentience is the innate human ability to experience feelings and sensations without association or interpretation. “We’re talking about more than just code; we’re talking about the ability of a machine to think and to feel, along with having morality and spirituality,” Ishaani Priyadarshini, a Cybersecurity Ph.D. candidate from the University of Delaware, tells Popular Mechanics.

💡AI is very clever and able to mimic sentience, but never actually become sentient itself.

The very idea of consciousness has been heavily contested in philosophy for decades. The 17th-century philosopher René Descartes famously said, “I think therefore I am.” A simple statement on the surface, but it was the result of his search for a statement that couldn’t be doubted. Think about it: he couldn’t doubt his existence as he was the one doubting himself in the first place.

New work explores optimal circumstances for reaching a common goal with humanoid robots

Researchers at the Istituto Italiano di Tecnologia (IIT-Italian Institute of Technology) have demonstrated that under specific conditions, humans can treat robots as co-authors of the results of their actions. The condition that enables this phenomenon is that a robot behaves in a human-like, social manner. Engaging in gaze contact and participating in a common emotional experience, such as watching a movie, are the key.

The study was published in Science Robotics and paves the way for understanding and designing the optimal circumstances for humans and robots to collaborate in the same environment.

The research study has been coordinated by Agnieszka Wykowska, head of IIT’s Social Cognition in Human-Robot Interaction lab in Genova, and a on a project titled “Intentional Stance for Social Attunement,” which addresses the question of when and under what conditions people treat robots as intentional agents.

Multilevel development of cognitive abilities in an artificial neural network

Several neuronal mechanisms have been proposed to account for the formation of cognitive abilities through postnatal interactions with the physical and sociocultural environment. Here, we introduce a three-level computational model of information processing and acquisition of cognitive abilities. We propose minimal architectural requirements to build these levels, and how the parameters affect their performance and relationships. The first sensorimotor level handles local nonconscious processing, here during a visual classification task. The second level or cognitive level globally integrates the information from multiple local processors via long-ranged connections and synthesizes it in a global, but still nonconscious, manner. The third and cognitively highest level handles the information globally and consciously. It is based on the global neuronal workspace (GNW) theory and is referred to as the conscious level. We use the trace and delay conditioning tasks to, respectively, challenge the second and third levels. Results first highlight the necessity of epigenesis through the selection and stabilization of synapses at both local and global scales to allow the network to solve the first two tasks. At the global scale, dopamine appears necessary to properly provide credit assignment despite the temporal delay between perception and reward. At the third level, the presence of interneurons becomes necessary to maintain a self-sustained representation within the GNW in the absence of sensory input. Finally, while balanced spontaneous intrinsic activity facilitates epigenesis at both local and global scales, the balanced excitatory/inhibitory ratio increases performance. We discuss the plausibility of the model in both neurodevelopmental and artificial intelligence terms.

Keywords: artificial consciousness; cognitive architecture; global neuronal workspace; synaptic epigenesis.

PubMed Disclaimer

/* */