all posts Artificial Intelligence | AI News

Meet Alex. He has built a working Ironman suit in his garage. Yes, a working version with a working reactor. Source | Artificial intelligence 🤖

🔺 Microsoft wants Three Mile Island to fuel its AI power needs 🔺 Microsoft just signed a deal to revive the shuttered Three Mile Island nuclear power plant. If approved by regulators, the software maker would have exclusive rights to 100 percent of the output for its Al data center needs. Constellation, the owner of the Three Mile Island plant, announced a power purchase agreement with Microsoft earlier today, which should see the site coming back online in 2028, assuming regulators approve it. 🔺 The reactor that Microsoft plans to source its energy from was retired in 2019 for economic reasons and is located next to a unit that was shut down in 1979 after the worst US nuclear accident in history. The plant that Constellation plans to reopen can generate 837 megawatts of energy, enough to power more than 800,000 homes - demonstrating the huge amount of power needed for data centers and Microsoft's Al ambitions. 🔺 Microsoft has agreed to purchase power from the plant - which will be renamed to the Crane Clean Energy Center to honor the late Chris Crane, former CEO of Exelon - for 20 years in a first-of-its-kind deal for the software giant. Microsoft's own greenhouse gas emissions are growing with its focus on Al, putting its ambitious climate goals at risk. 🔺 Bloomberg reports that this nuclear plant would help Microsoft's plans to run its data centers on clean energy by 2025 and power data center expansions in Chicago, Virginia, Pennsylvania, and Ohio. "This agreement is a major milestone in Microsoft's efforts to help decarbonize the grid in support of our commitment to become carbon negative," says Bobby Hollis, vice president of energy at Microsoft. "Microsoft continues to collaborate with energy providers to develop carbon-free energy sources to help meet the grids' capacity and reliability needs." 🔺 Microsoft has been betting on next-generation nuclear reactors to power its data center and Al plans recently, looking for someone who could roll out a plan for small modular reactors (SMR) last year. Microsoft cofounder Bill Gates is also a "big believer that nuclear energy can help us solve the climate problem." Constellation will invest $1.6 billion to revive the plant, and the company will need approval from the Nuclear Regulatory Commission to bring the site back online, alongside permits from state and local agencies. Constellation is also pursuing a license renewal to extend plant operations until at least 2054. Source | Artificial intelligence 🤖

NVIDIA AI Aerial Launches to Optimize Wireless Networks 📊 🔺 NVIDIA has launched AI Aerial, a platform for designing, simulating, training, and deploying AI radio access network (AI-RAN) technology. AI Aerial aims to optimise wireless networks and support new generative AI experiences across a range of applications, including autonomous vehicles, robots, and smart factories. 🔺 This platform is driven by a growing ecosystem of partners, including T-Mobile, Ericsson, Nokia, and Softbank, working to accelerate the commercialisation of AI-RAN. The AI-RAN Innovation Center, established through collaborations with these partners, focuses on bringing RAN and AI innovation closer together to deliver transformational network experiences. Source | Artificial intelligence 🤖

Jim Fan says that NVIDIA is building a full-stack computing platform for humanoids and other robots, which includes their inference chip, robotics foundation model, and simulation tools. He quotes Jensen: "Everything that moves will eventually be autonomous." Source | Artificial intelligence 🤖

Neura presents its humanoid robot 4NE-1 🤖 The German company's announcement comes alongside Nvidia's unveiling of new tools for humanoid robots at the Siggraph conference in Colorado. Neura is among the select companies with early access to Nvidia's development tools, which are intended to enhance the capabilities of humanoid robots. While Neura's 4NE-1 project has generated interest, it remains in the early stages as the company has primarily concentrated on traditional industrial robotics. By integrating its cognitive robotics solutions with Nvidia's advanced computing power, Neura aims to accelerate the development of humanoid robots. Source | Artificial intelligence 🤖

🚨 BlackRock and Microsoft are reportedly planning a $30B AI-focused megafund 🔹 Microsoft and BlackRock, the world's largest investment firm managing $10 trillion in assets, are collaborating to establish an investment fund aimed at raising $30 billion, with the potential to expand it to $100 billion. The fund will focus on investing in AI infrastructure, which involves developing production and supply chains, as well as building data centers and power plants, rather than directly investing in companies like OpenAI. 🔹 BlackRock appears to anticipate a future deficit in AI infrastructure and aims to position itself strategically ahead of this trend. Additionally, MGX, a significant fund from the UAE, will join as a partner in this venture, highlighting the global interest in enhancing AI capabilities and infrastructure development. Source | Artificial intelligence 🤖

You'll have a robot doing your housework within 3 years. 🤖 Several companies are racing to bring humanoid robots into homes within the next three years. Major players like Tesla, Figure, and 1X are at the forefront, with their robots offering advanced mobility, human-like interaction, and autonomous assistance. The global humanoid robot market, valued at around $2.43 billion in 2023, is expected to grow rapidly, potentially reaching $66 billion by 2032. Some expect much higher. Elon musk predicts Humanoid manufacturing to reach over a billion units per year. Are you buying a humanoid to do your dishes? Artificial intelligence 🤖

World's first fully robotic heart transplant performed 🤯 🔹A groundbreaking achievement in medical technology has been made at King Faisal Specialist Hospital and Research Centre in Riyadh, Saudi Arabia, where the world’s first fully robotic heart transplant was successfully performed. The operation, lasting about two and a half hours, was conducted on a 16-year-old patient suffering from end-stage heart failure, who specifically requested that his chest remain unopened during the procedure. 🔹The surgery was led by cardiovascular surgeon Feras Khaliel, who prepared meticulously with his team, conducting seven training sessions over three days to ensure everything went smoothly. This rigorous preparation highlights the complexity and innovation involved in performing such a pioneering operation. 🔹Majid Al Fayyad, the CEO of the medical center, praised the accomplishment as a significant advancement in healthcare, drawing parallels to the historical significance of the first heart transplants conducted in the 1960s. The fully robotic approach marks a new era in surgical procedures, emphasizing precision and minimally invasive techniques. 🔹This landmark surgery not only showcases the capabilities of robotic technology in medicine but also sets a precedent for future procedures. It represents a major step forward in cardiac care, offering hope for improved outcomes in patients requiring heart transplants. Source | Artificial intelligence 🤖

Snap's Spectacles feature a suite of cameras and sensors for multimodal AI and contextual understanding. They're powered by AI and can create shared immersive experiences with people nearby. BUT the battery only lasts for 45 minutes, and they cost $99/m. Source | Artificial intelligence 🤖

AI predicts crimes BEFORE they happen! 🤯 🔹 Dejaview examines human behavior, location data, and historical incident records, boasting an 82.8% accuracy rate in crime prediction. 🔹 It is capable of identifying repeat offenders to assess their intentions and help prevent crimes before they occur. The system is currently implemented in airports, factories, and public areas in South Korea. 🔹 The developers aim to launch it to the general public by 2025. Source | Artificial intelligence 🤖

The recent hearings titled "AI Oversight: An Insider's View," conducted by the Subcommittee on Privacy, Technology, and Law of the U.S. Senate Judiciary Committee, featured significant testimonies about the future of artificial intelligence. Helen Toner, a former OpenAI board member known for her critical stance on the organization, warned senators that human-level AI is approaching rapidly, and society is inadequately prepared for its implications. Toner emphasized that as discussions around AI evolve, the Overton window—the range of ideas considered acceptable in public discourse—is expanding significantly. This shift highlights the urgent need for policymakers and society to address the challenges and risks associated with advanced AI technologies before they become a reality. Source | Artificial intelligence 🤖

🔔 AI has returned chipmaking to the heart of computer technology 🔺 The rapid development of modern AI models has led to an increasing demand for computing resources, with processing requests doubling every six months, a significant shift from the previous trend that aligned with Moore's law. This surge in demand has drawn renewed attention to microprocessor companies, which have historically operated in the shadow of software giants. 🔺 Despite advancements in technology, the fundamental concepts behind chip creation have remained largely unchanged for decades, with only minor improvements made over time. This stagnation suggests that the current methods of chip development may no longer suffice to meet the escalating performance requirements driven by AI. 🔺 To achieve exponential performance gains, chip manufacturers must explore new ideas and approaches. Some of these ideas involve tighter integration between hardware and software, which are considered incremental improvements, while others call for more radical shifts in design philosophy. 🔺 These radical ideas may include rethinking the materials used in silicon chips or even moving away from traditional digital processing techniques altogether. As the demand for AI capabilities continues to grow, the industry must innovate to keep pace with the evolving technological landscape. Source | Artificial intelligence 🤖

Tesla's in-house production of 4680 cells has hit the 100 million mark, with some help from Optimus! 🤯 Source | Artificial intelligence 🤖

1. post for deep machine 👉 Ready for a real adventure? Introducing Hamster. Creation 🐹 — your chance to build and scale your own empire! New characters, exciting challenges, and endless growth opportunities await you. 💼 Develop your entrepreneurial skills, manage your business, and watch your profits grow with every step. 👾 Start the game first and take the initiative while others are just getting started! https://shorturl.at/SaAKY

💠 Microsoft's Copilot 2.0 is out. 🔺 Microsoft has released Copilot 2.0, an enhanced AI tool that now integrates seamlessly with Excel for data analysis using Python. This new version significantly boosts productivity by enabling users to perform complex data tasks more efficiently. 🔵 One of the standout features of Copilot 2.0 is its ability to generate, insert, and execute code based on user queries almost instantaneously. This functionality streamlines the coding process, making data manipulation more accessible for users without extensive programming knowledge. 🔵 Additionally, the AI excels at creating visually appealing graphs and charts that rival those produced by professional designers. It can handle any formulas, make predictions, and manage financial transactions, providing comprehensive bookkeeping capabilities within Excel. 🔵 Best of all, Copilot 2.0 is available for free across multiple Microsoft applications, including Excel, PowerPoint, Outlook, Word, and OneDrive, making advanced data analysis tools accessible to a wider audience. Source | Artificial intelligence 🤖

At Disneyland's Avengers Campus, a "Stuntronics" animatronic robot developed by Disney engineers performs Spider-Man's aerial stunts up to 65 feet high. It looks so lifelike that visitors fully believe they are seeing an actual stunt performer. Source | Artificial intelligence 🤖

🔺 OpenAI's o1: The new era of reasoning in artificial intelligence 🔹 OpenAI has introduced its latest AI model, o1, which is distinguished by its advanced reasoning capabilities. This innovation enables the model to effectively handle complex tasks in fields such as science, programming, and mathematics, achieving performance levels akin to those of PhD students. 🔹 In a notable performance at the International Mathematical Olympiad, o1 achieved an impressive 83% accuracy, significantly surpassing the results of its predecessor. Sam Altman, co-founder of OpenAI, acknowledges that while the model has its limitations, it represents a significant advancement in tackling difficult challenges across various domains, including medical research and advanced physics. 🔹 The o1 model benefits from efficient training methodologies and is designed for continuous improvement in learning. This adaptability positions it to transform how professionals engage with complex problems in their everyday work. 🔹 Overall, OpenAI's o1 marks a significant leap forward in artificial intelligence, particularly in its reasoning capabilities, indicating a promising future for its application in various specialized fields. Source | Artificial intelligence 🤖

OpenAI showed off the capabilities of its new O1 model — and it’s amazing! 🔹 The neural network created a working game with complex conditions from a single request. 🔹 He solved a quantum physics equation (!), showing the entire solution process. 🔹 Overcame a complicated logic puzzle. 🔹 She learned to translate characters with a literary approach. The language barrier has fallen. 🔹 The developers believe that the model will be able to make scientific discoveries. For example, find a cure for cancer. Source | Artificial intelligence 🤖

🍁Autumn is the time for cryptocurrency growth! Let Crypto Baron show you which coins can bring maximum profit. Simple tips for both beginners and experts. Get free recommendations for smart investing! 🚀 Don’t miss out! Subscribe now: https://t.me/+W0QGUSP0SXgxYzEy

Daily interact with humanoid robots. "🤖 The UNITREE H1: a humanoid robot that's not just a conductor at a music academy, but a symbol of the endless possibilities of robotics! Source | Artificial intelligence 🤖

🎧 Real-time Translation Headphones Launched in the USA! Timekettle unveiled the W4 pro AI headphones, providing real-time translation in 40 languages. These headphones are designed to enhance communication in meetings, video calls, and foreign content consumption. They can even generate post-meeting summaries, ensuring seamless understanding. Source | Artificial intelligence 🤖

According to communications sent to regulators, Tesla is hinting at making its self-driving vehicles with seats that can rotate to face backward. Source | Artificial intelligence 🤖

World's Biggest iPhone — Two-Meter Pro Max! 🤯 YouTuber Mrwhosetheboss has created a true giant — a fully functional replica of the iPhone 15 Pro Max standing over two meters tall! This isn’t just a model; it's a fully operational smartphone capable of sending messages, running apps, and even features a massive flashlight and charging capability. It's now officially in the Guinness World Records! Source | Artificial intelligence 🤖

Artificial Intelligence will eventually be able to understand more than just text,image, and audio, but revolutionarily understand biology itself. Artificial intelligence 🤖

AI is amazing! 🤯✨ The progress we're seeing is unbelievable. Artificial intelligence 🤖

What happens when a humanoid robot is linked to Chat GPT? With OpenAI, Figure 01 can now have full conversations with people Source | Artificial intelligence 🤖

New video of Reachy 2 by France-based Pollen Robotics shows it performing teleoperated high-precision tasks. The beta version of the bot with open-source ecosystem launches next month. Prices: $75K for the full robot with a mobile base; $38K for a stationary one-armed model. Source | Artificial intelligence 🤖

Offices can be cleaned 30% cheaper with robots. Loki Robotics has unveiled a robot commercial cleaning service that claims to be much cheaper than any competitor in the U.S., by using a mixed team of robots and humans. Would you hire a robot to do your cleaning? 🤔 Source | Artificial intelligence 🤖

Eric Jang explains that the "ChatGPT moment in robotics" is difficult because testing robots in millions of scenarios is hard. Unlike LLMs, robots face a constantly changing world, making reproducible tests challenging. This makes it harder to establish the scaling laws. Source | Artificial intelligence 🤖

Meet Toyota’s CUE6, an AI-powered basketball #robot with impressive shooting accuracy. This 6th-gen robot, initially a passion project, is a glimpse into a future where robots could excel in various sports like tennis and golf. 🏀 Source | Artificial intelligence 🤖

A useful post summarizes the news about OpenAI's latest advanced model, o1: 🔹 The improvement in quality stems from the model's ability to reason before providing an answer. While the reasoning process itself won't be shown, there will be a brief summary with a high-level overview. 🔹 Previous models could reason as well, but with less effectiveness. OpenAI has focused on enhancing the model's ability to arrive at the correct answer more frequently through iterative self-correction and reasoning. 🔹 o1 is not intended to replace gpt-4o for all tasks. It excels in math, physics, and programming, follows instructions more accurately, but may struggle with language proficiency and has a narrower knowledge base. The model should be viewed as a reasoner (akin to "thinker" in Russian). According to OpenAI, the mini version is comparable to gpt-4o-mini, with no major surprises. 🔹 The model is currently available to all paid ChatGPT Plus subscribers, but with strict limits: 30 messages per week for the large model and 50 for the mini version. So, plan your requests carefully! 🔹 If you have frequently used the API and spent over $1,000 in the past, you can access the model via API with a limit of 20 requests per minute. 🔹 However, costs are high: the junior version of o1-mini is slightly more expensive than the August gpt-4o, but you’re paying for reasoning (which you won’t see) that will be substantial. Thus, the actual markup could range from 3 to 10 times, depending on the model's "thinking" time. 🔹 The model handles Olympiad-level mathematics and programming problems with the skill of international gold medalists, and for complex physics tasks resistant to Google searches, it performs at a PhD student level (~75-80% correct). 🔹 Currently, the model cannot use images, search the internet, or run code, but these features will be added soon. 🔹 The context for models is still limited to 128k tokens, similar to older versions. However, an increase is anticipated in the future, as OpenAI claims the model currently "thinks" for a couple of minutes at a time, with aspirations for longer durations. 🔹 As with any initial release, there may be some simple bugs where the model fails to respond to obvious prompts or leads to jailbreaks. This is normal, and such issues should decrease in 2-3 months once the model transitions from preview status. 🔹 OpenAI already possesses a non-preview version of the model, which is currently being tested and is reportedly better than the current release—see the attached image for details. 🔹 The new model operates without needing prompts; you won’t have to ask it to respond in a thoughtful, step-by-step manner, as this will be handled automatically in the background. Welcome to Strawberry Era! 🔥 Source | Artificial intelligence 🤖

Hard to believe she’s not real. This AI-generated video, made with Runway’s Gen-3 from Midjourney images, is truly impressive. 📷 The future of AI-generated films is here! Artificial intelligence 🤖

🐵 Artificial intelligence 🤖

OpenAI released a new o1 model today 🔥 It is 5-6 (!) times better than the past strongest neuron GPT-4o. This is the very secret project that has been worked on for so long development. The new model shows itself 5 times better in math problems and 6 times better in writing code! This strong increase in quality is due to the fact that the model “thinks” before giving you the answer. Access starts today. Source | Artificial intelligence 🤖

Mark Zuckerberg showed off the capabilities of Ray-Ban's Meta sunglasses via a series of videos taken in Montana. The glasses use audio to provide a verbal description of Big Sky Mountain and the history of the Roosevelt Arch while explaining (like a caveman) how snow is formed. Source | Artificial intelligence 🤖

Automated AI Forecasting Recently, I’ve been testing FiveThirtyNine, which uses AI to generate predictions for all kinds of questions—politics, tech, world events, you name it. You just enter a question, and the AI gives you a probability based on its analysis. You can mess around with your own questions at forecast.safe.ai. Artificial intelligence 🤖

🇨🇳 vs 🇺🇸 in AI: A Tech Race Update! Kai-Fu Lee, ex-head of Google China, recently commented on the state of AI development between the US and China. Here's the scoop: The Gap: According to Lee, China's AI models are now only 6 to 9 months behind their U.S. counterparts. This is a significant catch-up from being years behind not too long ago. Why It Matters: AI isn't just tech; it's a cornerstone for future innovation, military applications, and economic dominance. The Context: Despite this narrowing gap, there are nuances: Investment: Reports suggest China invests heavily in AI for military and civilian use, potentially outpacing U.S. public sector investment in terms of percentage of budget allocation. Talent: The U.S. remains a magnet for global AI talent, with a high retention rate of AI researchers educated there. However, China's domestic AI education and industry are booming, keeping most of their talent home. Tech & Chips: While China makes strides in AI application, there's mention of a lag in chip manufacturing technology, crucial for AI advancement, where U.S. companies still lead. The Discussion: This isn't just about who's ahead today. It's about innovation ecosystems, government support, and the integration of AI into broader societal functions. Future Predictions: If this trend continues, we might see China not just catching up but setting new benchmarks in certain AI applications, especially given their rapid deployment and integration in various sectors. What's Next?: Will the U.S. respond with increased investments or policy changes to foster AI innovation? Or will China's strategic focus on becoming an AI leader by 2030 reshape global tech landscapes? 🤔 Source | Artificial intelligence 🤖

The world's first AI-powered dumbbells. These smart weights offer automatic rep counting, real-time form correction, and velocity-based training to optimize every workout. Source | Artificial intelligence 🤖

Huawei launches the world's first triple-folding smartphone after the iPhone 16 Huawei has burst into the smartphone market with its Mate XT Ultimate Design, the world's first triple-folding smartphone, unveiled just hours after the long-awaited launch of the iPhone 16. This innovative device offers three modes of use: as a conventional phone, in foldable Fold mode and expanded into a tablet, with an impressive screen that varies between 6.3 and 10.2 inches. Equipped with HarmonyOS 4.2 and powered by artificial intelligence, it has up to 16 GB of RAM and storage options that reach one terabyte. The camera stands out for its XMAGE technology, including a 50-megapixel main sensor and a 5.5x optical zoom. Its price ranges from $2,809 to $3,371, depending on the internal storage of each version. Source | Artificial intelligence 🤖

🇺🇸 OpenAI officially announced at KDDI SUMMIT 2024 that ChatGPT has surpassed 200 million users and the new model is 100 times more powerful than GPT-4 Source | Artificial intelligence 🤖

Elon on All-in summit: ⦿ "Next-generation Optimus hand, which we have in prototype form, has actuators that have moved to the forearm, just like humans, and they operate the fingers through cables, just like human hands." ⦿ Chamath: "If you just provide Grok to the robot, then the robot has personality." Elon: "Oh yeah!" ⦿ After three production versions of Optimus and scaling to 1m+ units, the cost (labor + material) won't be more than $10k. ⦿ "The useful humanoid robot opportunity is the single biggest opportunity ever." Source | Artificial intelligence 🤖

OpenAI Co-founder Andrej Karpathy predicts transformers may soon surpass the human brain in capabilities "Transformers have the potential to surpass the human brain in efficiency and memorization capabilities, despite current limitations in data availability and training. As AI technology advances, transformers may become more efficient learners than the human brain, which operates under various limitations" Source | Artificial intelligence 🤖

Step into the future of transportation with the “Land Aircraft Carrier” flying car. A new way to travel has arrived! Source | Artificial intelligence 🤖

The UK’s 🇬🇧 first "teacherless class", using AI instead of human teachers, is about to start 📚 David Game College, a private school in London, opens its new course for 20 GCSE students in September. Source | Artificial intelligence 🤖

AI will make it possible for one person to build a billion dollar company Artificial intelligence 🤖

Waeve introduced Isaac, a personal robot that's built for the home and is shipping next year It will "autonomously tidy up endless messes, fold your laundry, and care for your home while you’re away." Controlled via voice, text, or from their phone app Source | Artificial intelligence 🤖

Altera's Project Sid created the first simulation of 1,000 fully autonomous agents collaborating in a virtual world (Minecraft) Without human intervention, the AI agents built economies, cultures, governments, and religions Source | Artificial intelligence 🤖

Enjoy our content? Advertise on this channel and reach a highly engaged audience! 👉🏻 It's easy with Telega.io. As the leading platform for native ads and integrations on Telegram, it provides user-friendly and efficient tools for quick and automated ad launches. ⚡️ Place your ad here in three simple steps: 1 Sign up 2 Top up the balance in a convenient way 3 Create your advertising post If your ad aligns with our content, we’ll gladly publish it. Start your promotion journey now!

🚨 SB 1047, a regulatory act for AI in California, has passed the State Senate and is awaiting the governor's signature. 🔹 While the governor can still veto the act, various stakeholders are voicing their opinions for and against it. Notably, Anthropic supports the act, believing it will do more good than harm, despite having concerns about specific points that they recommend changing. In contrast, OpenAI expresses strong opposition to the legislation. 🔹 Jan Leike, a former co-leader of SuperAlignment and now with Anthropic, argues that concerns about the act stifling innovation, startups, and open-source projects are unfounded. He contends that teams investing over $10 million in model retraining should be able to afford to create a model security plan and conduct necessary assessments. Leike emphasizes that if a model leads to significant harm or casualties, it indicates a serious failure in its development. 🔹 Addressing criticisms regarding the need for legal teams, Leike questions why organizations spending vast sums on AI development would balk at hiring legal counsel to ensure compliance. He argues that the act does not represent regulatory capture, implying that responsible practices should be standard in the industry. 🔹 Leike also refutes the analogy that blaming AI developers for misuse is akin to holding hammer manufacturers responsible for crimes. He points out that society often holds producers liable for harm caused by their products when they fail to exercise reasonable care, citing examples from pharmaceuticals to automobiles. 🔹 Finally, Leike clarifies misconceptions about the act's implications, particularly regarding safety regulations. He draws a parallel with historical aviation safety measures, asserting that the act does not ban AI development but requires companies to demonstrate safety protocols, similar to how airships would need to prove the safety of hydrogen use to avoid liability. This framework aims to enhance accountability without stifling innovation. Source | Artificial intelligence 🤖

PNDbotics joins the growing list of Chinese humanoid robotics startups. They aim to advance humanoid robotics by developing modular actuator hardware, original robot designs, and cutting-edge control algorithms. Here's their first robot, Adam. Source | Artificial intelligence 🤖

The future is here! Artificial intelligence 🤖

OpenAI CTO Mira Murati says the definition of intelligence will continue to evolve and we will have incredibly advanced AI systems within the next 5 years Source | Artificial intelligence 🤖

Robots have become completely intimidating in China. They are afraid of everything and everyone, jumping back in horror. I hope we don't get the same backlash in a couple of years. But seriously, this is the X30 robot dog from the Chinese company Deep Robotics. And they taught it to avoid contact with people and other objects for safety's sake, of course. The price of such a four-legged friend starts at $65,000. And that's not a small amount. Boston Dynamics' robot dog cost $75k when they were still selling them (now, it seems, sales are made only by individual agreements). Source | Artificial intelligence 🤖

Meet EELS, a snake-like robot being developed to autonomously map and explore previously inaccessible destinations on Earth, the Moon, and other worlds in our solar system. Inspired by a desire to look for signs of life below the icy crust of Saturn’s moon Enceladus, the robot’s testing and development phase has shown that EELS is versatile and can be adapted to a wide variety of terrain including sand and ice, cliff walls, craters too steep for rovers, underground lava tubes, and labyrinth-like spaces within glaciers. Source | Artificial intelligence 🤖

1X CEO talks about the AI data-engine with unlimited potential for what it can learn. "The future of science is billions of droids running experiments in labs all over the world and, through that, progressing our understanding of the universe." Source | Artificial intelligence 🤖

🎥 OpenAI and Anthropic will share their models with the US government 👍 OpenAl and Anthropic have agreed to let the US government access major new Al models before release to help improve their safety. The companies signed memorandums of understanding with the US Al Safety Institute to provide access to the models both before and after their public release, the agency announced on Thursday. 👍 The government says this step will help them work together to evaluate safety risks and mitigate potential issues. The US agency said it would provide feedback on safety improvements, in collaboration with its counterpart agency in the UK. Sharing access to Al models is a significant step at a time when federal and state legislatures are considering what kinds of guardrails to place on the technology without stifling innovation. 👍 On Wednesday, California lawmakers passed the Safe and Secure Innovation for Frontier Artificial Intelligence Models Act (SB 1047), requiring Al companies in California to take specific safety measures before training advanced foundation models. It's garnered pushback from Al companies including OpenAl and Anthropic that warn it could harm smaller open-source developers, though it's since undergone some changes and is still awaiting a signature from Governor Gavin Newsom. 👍 In the meantime, the White House has worked to secure voluntary commitments from major companies on Al safety measures. Several leading firms have entered non-binding commitments to invest in cybersecurity and discrimination research and work on watermarking Al-generated content. US Al Safety Institute director Elizabeth Kelly said in a statement that the new agreements were "just the start, but they are an important milestone as we work to help responsibly steward the future of AI." Source | Artificial intelligence 🤖

What is Musk planning for AI? Likely goal: superhuman AI mathematician in 2025-2026 Elon Musk appears to be focused on developing a superhuman AI mathematician, aiming for its completion by 2025 or 2026. This initiative involves significant investment in Colossus, touted as the world's most powerful AI cluster, indicating a shift in priorities away from his Mars colonization plans. Unlike others in the AI field, Musk is not pursuing artificial general intelligence (AGI), which he views as a more remote and less practical goal. Musk's approach contrasts sharply with that of figures like Sam Altman, who is willing to spend billions on the pursuit of AGI without concern for costs. Musk is known for his pragmatic focus on achievable goals, and his ambition to create a superhuman AI mathematician aligns with this mindset. This goal is seen as tangible and attainable, distinguishing it from the more abstract concept A notable comment from Christian Szegedy, a key member of the team at Musk's xAI, who has revised his timeline for achieving a superhuman AI mathematician from 2029 to potentially 2025. This reflects a growing confidence in rapid advancements in AI capabilities. Szegedy's assertion suggests that the necessary technology and hardware are within reach, making the goal more plausible. Achieving a superhuman AI mathematician could have significant implications, possibly exceeding the excitement surrounding space exploration plans. The focus on powerful hardware and practical applications of AI mathematics underscores Musk's intent to create a groundbreaking technology that could transform various fields, shifting the narrative from speculative ambitions to concrete achievements. Source | Artificial intelligence 🤖

ChatGPT currently costs $20 a month. How much do you think it might go up in the near future with the release of new models? $50? $75? How about $200 or $2,000? You might be surprised, but such numbers (yes, two thousand) were mentioned in internal discussions at OpenAI. This issue is especially acute now (well, if you believe that the next generation of models will pleasantly surprise) - after all, the Strawberry (improving the reasoning skills of models) and Orion (probably GPT-5) projects will require more resources to work. According to a fairly popular hypothesis, the models will need time to "think" before giving an answer, and all this time a neural network will be spinning in the background. Of course, raising the price (especially to over $100) would also mean that OpenAI believes that its existing ChatGPT customers will find these new models much more valuable for their day-to-day work. OpenAI is currently making about $2 billion a year on $20 subscriptions. I wonder where they'll be in a year.👀 Source | Artificial intelligence 🤖

Ilya Sutskever's Safe Superintelligence startup, which is only a couple of months old, has attracted a gigantic seed round of $1 billion. Investors there include a16z, Sequoia, DST Global and others. Reuters sources say that the valuation is $5 billion. Let me remind you that Sutskever left OpenAI, took a month off and started this project with his buddies, which is developing safe superintelligence. There are currently 10 people on staff, and offices in Palo Alto and Tel Aviv. It turns out that $100 million per employee - it should be enough for the first time. Source | Artificial intelligence 🤖

Anthropic CEO Dario Amodei says AI could increase the rate of discovery in biology by 100x, compressing the total amount of progress that occurred in the 20th and 21st centuries to just a few years Source | Artificial intelligence 🤖

AI spots cancer and viral infections at nanoscale precision - Spanish scientists have developed an artificial intelligence which can differentiate cancer cells from normal cells, as well as detect the very early stages of viral infection inside cells. The findings, published in the journal Nature Machine Intelligence, pave the way for improved diagnostic techniques and new monitoring strategies for disease. The tool, AINU (AI of the NUcleus), scans high-resolution images of cells. AINU detects and analyses tiny structures inside cells at the molecular level. The researchers trained the model by feeding it with nanoscale-resolution images of the nucleus of many different types of cells in different states. The model learned to recognize specific patterns in cells by analysing how nuclear components are distributed and arranged in three-dimensional space. Source | Artificial intelligence 🤖

Artificial Intelligence, MIT Vision Laboratory, 1971, Artificial intelligence 🤖

Come & meet Hero.io on the roadshow: 🗓️ September 11-12: Dubai AI & Web3 Festival Learn more: https://dubaiaiweb3festival.com/ 🗓️ September 14: Redacted Airways Party at Burj Khalifa Learn more: https://redactedairways.com/ 🗓️ September 17-18: Token 2049 – We're hosting 3 amazing side events! 1️⃣ AI for the People: Empowering Web3 Economy Register here: https://lu.ma/a7q99fap 2️⃣ Investing in the Age of AI Boom: Opportunities for Web3 Register here: https://lu.ma/hltpwo6h 3️⃣ Cambridge University AI Research for Web3 Register here: https://lu.ma/avv2qd5a

🔺 ChatGPT’s weekly users have doubled in less than a year 🔹 OpenAI says that more than 200 million people use ChatGPT each week, as first reported by Axios. OpenAl spokesperson Taya Christianson confirmed the number to The Verge, which is now double the 100 million weekly active users OpenAl reported last November. 🔹 Additionally, Christianson says that 92 percent of Fortune 500 companies are using OpenAl's products, while API usage has doubled following the release of the company's cheaper and smarter model GPT-40 Mini. 🔹 Since ChatGPT's launch in late 2022, Google, Microsoft, and Meta have also launched Al chat interfaces of their own. Today, Meta CEO Mark Zuckerberg revealed that the company's Al assistant has reached over 400 million monthly active users and 185 million weekly active users - despite not rolling out in the UK, Brazil, or EU yet. 🔹 Earlier today, the US Al Safety Institute also announced that OpenAl and Anthropic have agreed to let the government evaluate major Al models before being launched to the public. Reports are also circulating that Apple and Nvidia could be among OpenAl's next round of investors. Source | Artificial intelligence 🤖

Meet the AI robot cleaning up cigarette butts 💡 Legged robots are being developed to help clean up the 4.5 trillion cigarettes that are thrown into our environment every year. IIT' Dynamic Legged Systems lab developed VERO, which uses AI-based detection to help clean beaches, small alleys and parks, particularly for removing cigarette butts, in terrains that are hard to reach for wheeled and tracked robots. The onboard vacuum cleaner is connected with hoses to each leg. Vacuum cleaner nozzles are attached to each foot, allowing the robot to autonomously collect the cigarettes efficiently while walking, by simply stepping close to the detected object and activating the vacuum. Source | Artificial intelligence 🤖

IShowSpeed made a fuss at the robo-dog, for which he received a deserved blast from the flamethrower 🤬 Artificial intelligence 🤖

OpenAI's first in-house chip will be built by TSMC using its A16 Angstrom process technology and will primarily be used for video generation in Sora. Industry insiders say OpenAI was in active talks with TSMC to develop a dedicated foundry for its own chip, but those plans were cancelled. It gets more interesting: The internal chip is said to boost Sora's video generation capabilities, which could help boost sales of Apple devices as the feature will be integrated into the company's suite of generative AI features. It's possible that OpenAI and Apple are working closely together on this project, as the latter has years of experience developing chipsets that are ahead of the competition, but nothing has been confirmed yet. Source | Artificial intelligence 🤖

Meet Galbot G1, the 1st-generation robot by Chinese startup Galbot, designed for generalizable, long-duration tasks. Founded in May 2023 by a Stanford PhD, the company currently has 80 engineers and researchers working on algorithms, software, and hardware development. Source | Artificial intelligence 🤖

Beauty and robots? 💄 The California-based company LUUM created a robot that can apply eyelash extensions to humans. In order to make the process of custom-eyed lash extensions faster, a professional lash artist works alongside the machine. Preparation begins with cleaning the lashes, then applying barcoded patches to the under-eyes and eyelids that can be read by the machine. Another robot isolates each lash, dips the extensions in glue, and places them on the isolated one. When a client touches a robotic hand, the hand will simply fall off. In each eye, it can apply 25 lashes, with the rest being applied by the lash artist. Source | Artificial intelligence 🤖

Booster Robotics' adorable little humanoids drew large crowds on the final day of the 2024 World Robot Conference in Beijing. Source | Artificial intelligence 🤖

Researchers at the University of Tokyo developed a human-like forearm with a radioulnar joint using a compact bone-muscle module. It replicates human forearm proportions, weight, and muscle arrangement, enabling skillful motions like turning screws and swinging a racket. Source | Artificial intelligence 🤖

Chinese copycat of Boston Dynamics' Atlas just released a video of their robot doing push-ups. 😂 Source | Artificial intelligence 🤖

Magic's new AI model has a 100M token context window, insane For non-AI nerds, that means the model can ingest and understand up to 750 novels worth of text, essential for useful autonomous AI agents The previous leader was Google DeepMind with a 10M context window in research Source | Artificial intelligence 🤖

🚨 California State Assembly passes sweeping AI safety bill 🔹 The California State Assembly has passed the Safe and Secure Innovation for Frontier Artificial Intelligence Models Act (SB 1047), Reuters reports. The bill is one of the first significant regulations of artificial intelligence in the US. The bill, which has been a flashpoint for debate in Silicon Valley and beyond, would obligate Al companies operating in California to implement a number of precautions before they train a sophisticated foundation model. 🔹 Those include making it possible to quickly and fully shut the model down, ensuring the model is protected against "unsafe post-training modifications," and maintaining a testing procedure to evaluate whether a model or its derivatives is especially at risk of "causing or enabling a critical harm." Senator Scott Wiener, the bill's main author, said SB 1047 is a highly reasonable bill that asks large Al labs to do what they've already committed to doing: test their large models for catastrophic safety risk. 🔹 "We've worked hard all year, with open source advocates, Anthropic, and others, to refine and improve the bill. SB 1047 is well calibrated to what we know about forseeable Al risks, and it deserves to be enacted." Critics of SB 1047 - including OpenAl and Anthropic, politicians Zoe Lofgren and Nancy Pelosi, and California's Chamber of Commerce — have argued that it's overly focused on catastrophic harms and could unduly harm small, open-source AI developers. 🔹 The bill was amended in response, replacing potential criminal penalties with civil ones, narrowing enforcement powers granted to California's attorney general, and adjusting requirements to join a "Board of Frontier Models" created by the bill. After the State Senate votes on the amended bill - a vote that's expected to pass - the Al safety bill will head to Governor Gavin Newsom, who will have until the end of September to decide its fate, according to The New York Times. 🔹 Anthropic declined to comment beyond pointing to a letter sent by Anthropic CEO Dario Amodei to Governor Newsom last week. OpenAl declined to comment and pointed to a letter by OpenAl chief strategy officer Jason Kwon sent to Senator Wiener, also last week. Source | Artificial intelligence 🤖

There is no person in the suit — you are looking at a real robot — NEO for household tasks. It will help you unload your groceries, wash the dishes, cook dinner and wash the dishes again. It will do literally everything you want. Characteristics: — Height - 1.65 meters; — Weight - 30 kg; — Walking speed - 4 km/h; — Running speed - 12 km/h; — Load capacity - 20 kg. The start dates of mass production and prices are still unknown. For some reason, scenes from the movie "I Robot" immediately come to mind Source | Artificial intelligence 🤖

We start the weekend with a funny but useful post - it can be thrown into chats to people who are sure that WELL THIS SHOWS THAT LLM IS A HYPE AND THERE IS NOTHING SMART THERE 😂 7 signs your daughter may be an LLM: 1️⃣Does she have trouble multiplying numbers beyond 2-3 digits if she's not allowed to write out the steps? 2️⃣If you ask her a question whose answer she doesn't know, does she sometimes make something up? 3️⃣Is she incapable of matching the heights of human intellect, not able yet to independently advance the frontiers of science and technology without outside assistance? 4️⃣If asked to draw a photorealistic image of a person, do the resulting anatomical proportions or fine details sometimes look off on a close inspection? 5️⃣Does her code sometimes contain bugs? 6️⃣Does she start to forget exact details of what she's already read after the first 10 million tokens? 7️⃣Does she sometimes claim to be conscious? Source | Artificial intelligence 🤖

Donald Trump says AI will be "the oil of the future" and to meet the electricity requirements it needs will require investing in nuclear power and building lots of small plants at the sites of data centers Source | Artificial intelligence 🤖

Elon Musk's AI system, Grok, generated a deepfake video showing Musk, Trump, and others committing armed robbery. The hyper-realistic clip, created for fun, has sparked debates about the ethics and future impact of AI-generated content. Source | Artificial intelligence 🤖

Can a swarm of robots build a car from scratch? 🤔 Ambots3D has developed a proof of concept demonstrating - Mobile 3D printers & arms collaborating to produce a car without a human in sight. Will factories of the future be built along these lines? Source | Artificial intelligence 🤖

Robo-slime! 🤯 A robot made from magnetic slime could grab objects inside your body. The magnetic slime robot can navigate narrow passages, grasp objects, and fix broken electrical circuits because of its custard-like consistency. This robot was invented by researchers at The Chinese University of Hong Kong. 🇨🇳 The device could be deployed inside the body to retrieve swallowed objects, for example. The toxicity of its magnetic particles remains a safety concern, despite its potential medical use. Robots don't need to be mechanical devices as you can see! 👀 Source | Artificial intelligence 🤖

AI Expo Europe brings together leading minds in artificial intelligence on October 6-7. Industry pioneers from IBM, NVIDIA, Microsoft, Google, and emerging startups will showcase real-world AI applications and discuss their impact across sectors. Because we believe in accessible knowledge (and we're not too shabby at machine learning ourselves), we've predicted you'd love a discount. Use AINEWS50 for 50% OFF any ticket type on www.aiexpoeurope.com

Artist Matthew Ritchie says ChatGPT was predicted by Jonathan Swift in Gulliver's Travels, with his idea of The Engine, a machine that produced books because people became too lazy to write themselves Source | Artificial intelligence 🤖

OpenAl's lawyers have made a smart move in their dispute with fiction authors who have accused the developer of violating their copyrights and using book texts to train an Al model. In response to the amended lawsuit, OpenAI noted that the learning process is not aimed at reproducing, but at creating new content. Any use of works to train models is necessary so that the model can analyze sentence syntax, spelling, and word frequency. In short, all of this is fair use, and what is not fair use is simply not fully understood. And why did they “come in beautifully”? Because they included in the text the result of ChatGPT generation - a description using legal analogies of the building at “450 Golden Gate Avenue, San Francisco” - the very building where the court that is hearing their case is located. But the cherry on top is that ChatGPT generated a response that used the legal principle of “stare decisis” (stand by what has been decided) as an analogy - the principle of respecting court decisions that have become precedents. Subtle. Source | Artificial intelligence 🤖

Introducing ACE - A Cross-Platform Visual-Exoskeletons System! 🤩 ACE is a low-cost, cross-platform teleoperation system using visual-exoskeletons for precise, real-time 3D hand pose tracking. It enables versatile control across diverse robots, supporting effective imitation learning. Source | Artificial intelligence 🤖

This robot from Disney Research can imitate human facial movements, specifically blinking and subtle head movements. A sensor in its chest area alerts the robot when to turn and face a person in front of it, and its eye movements shift from direct eye contact to the rapid eye movements known as saccades. It also moves slightly up and down to mimic breathing. The robot was developed by engineers at Disney’s Research division, Walt Disney Imagineering, and robotics researchers from the University of Illinois, Urbana-Champaign and the California Institute of Technology. Source | Artificial intelligence 🤖

AI turns wi-fi routers into "cameras" that see people through walls 😳 Source | Artificial intelligence 🤖

Spiderman's stuntman is a robot! 🕷️ Walt Disney Imagineering Imagineers created an advanced robotics figure capable of tucking, somersaulting, slowing down, and climbing 25 meters in the air in real time. 🤹🏼‍♀️ This is Spider-Man in Avengers Campus, doing gravity-defying stunts that have never been seen in a Disney park before. That's incredible to see how robotics is changing different verticals! 🎞️ In the future, we can expect even more sophisticated and realistic characters and experiences as robotics continues to evolve. They just can't create a robot that won't steal the show, right? 😐 Source | Artificial intelligence 🤖

⚡️ Following Vitalik Buterin, SB-1047 was supported by Elon Musk Artificial intelligence 🤖

Goodbytz introduces the Robotic Kitchen Assistant, set to transform the food service industry! 🚀 This AI-powered marvel boosts efficiency, cuts labor costs, and expands menu possibilities with advanced automation. 🍽️ 🤖 Source | Artificial intelligence 🤖

Elon Musk has posted a video on his X from the giant new Cortex AI training supercluster being built at Tesla's Austin headquarters to solve "real-world AI problems." Usually they don’t allow you to take videos or photos in such clusters, and you can only walk around there in shoe covers, robes, and tinfoil hats – so let’s enjoy it while we can. Source | Artificial intelligence 🤖

From Beijing World Robotics Conference 😳 Source | Artificial intelligence 🤖

Boston Dynamics' Atlas does push ups and a burpee. 👀 Source | Artificial intelligence 🤖

Facts 🤷‍♂️ Artificial intelligence 🤖

New Chinese humanoid robot startups will stop emerging about the same time penguins start flying. Meet TLIBOT T1Pro, designed by SICHUAN TLIBOT Co. — a humanoid robot with a 1.25 kWh battery, 5'3" height, 95 lbs weight, currently being developed for industrial applications. Source | Artificial intelligence 🤖

Terence Tao says AI will enable mathematics to be done at an unprecedented scale: instead of solving one theorem at a time and solving each problem separately, AI could work on the space of problems and classes of 1000s of problems at a time Source | Artificial intelligence 🤖

A game-changing step into the future of cooking at home, Moley brings you the first fully robotic kitchen. The Moley Robotics Kitchen allows you to save time, free up your day from routine cooking, plan and adapt your menu according to different diets and lifestyles, enjoy international cuisine anytime, control calories and get cooking tips and recipes from chefs around the world. Source | Artificial intelligence 🤖

✈️

Florida-based Boardwalk Robotics just unveiled their first commercial humanoid robot, Alex, developed entirely in-house. 🤖 While legs are in the plans, Boardwalk is prioritizing advanced manipulation, considering many applications don't require legs for the robot to be useful. Source | Artificial intelligence 🤖

Before everyone gets a cyborg body, there's a huge opportunity to revolutionize prosthetics for amputees. Atom Touch is the first prosthetic arm with individual finger control — uses electromyography for precise operation. Clinical trial & FDA approval expected in 12-18 months. Source | Artificial intelligence 🤖

A crazy guy from TikTok built Robo-Jesus It's an art object where everyone can touch the "divine machine" — in reality, a set of boards with ChatGPT. The last photo shows the final stage of the project. 🗿 Source | Artificial intelligence 🤖