Driven away from pirated digital libraries by a massive... $1.5 billion settlement, major AI companies have found a legal loophole: bulk-buying physical books, shearing off the spines for high-speed scanning, and then throwing the originals away. The practice, known as "destructive scanning," involves using hydraulic cutting machines to neatly slice the spines off used and sometimes rare books. The loose pages are then rapidly processed through industrial scanners to extract the text for AI training data, before the physical remains are sent to recycling facilities. Anthropic’s internal initiative, referred to as "Project Panama," reportedly aimed to destructively scan up to two million books over a six-month period. The pivot to physical books comes after a federal judge ruled that training AI on legally purchased physical media constitutes "fair use," even if the original physical copy is destroyed in the process. The ruling effectively closed the door on using pirated digital shadow libraries, making physical books the final frontier for acquiring massive, legally defensible datasets. What are your thoughts on this? Share them in the comments.show more

I just moved to Berkeley, and unpacking was a... good excuse to catalog the books in my library. I currently own 450 physical books, of which 18% are fiction and 82% are nonfiction. About 25% of my nonfiction books are about biology, which was surprising to me because I don’t really enjoy reading about biology in my spare time. When I shared my bookshelf with Claude and asked for recommendations, it suggested The Invention of Air by Steven Johnson, The Making of the Atomic Bomb by Richard Rhodes, and The Metaphysical Club by Louis Menand. The first two were already on my list, but I had never heard of the third. Overall, I think a digital bookshelf is a useful way to find more books based on my interests. My full bookshelf is here:show more

I am very happy to report we have scanned... the last page of the Howard W. Sams Photofact books for AI training. Here is the video of the final page of 1000s. No, not to repair TVs and Radios. But why? Many reasons. One is to capture the logic and thinking processes of an ancient epoch. This book series alone will align ANY AI platform to higher abilities because of the thinking used. Another is to understand what analog technology really is in a modern concept as it informs many functions completely unrelated to electronics. Thus I now am going to plan on a preservation process for these books that should allow them to be available for the next few 100 years or perhaps 1000 years. They are a masterpiece of very high protein data. In fact the highest protein data I have discovered, each square inch is valuable. I will write a series of articles on some landmark discoveries Ai have made in these books. More soon. And once again thank you for this donation. Thank you for thinking of our future through our past. Many say they will “help me out” some very smart and very wealthy, this guy came through. Gratitude.show more

Hey #NeuraxonMini is literally out! , we manage to... "transplant" a Neuraxon 2 bioinspired #AI brain to a physical robot the #SpheroMini moving from our last Scientific Paper (link bellow) by David Vivancos - e/acc & Jose Sánchez for Qubic #OpenScience hybridized with #Aigarth to the real World. First you need a Sphero Education Mini robot about 50$ Then you can try the first cool demos at Hugging Face: 1.- Neuraxon2MiniControl to drive the sphero robot 2.- Neuraxon2MiniWrite to write letters or words with physical moves of the sphero robot using Neuraxon Video Tutorials on youtube later today. Why this matters? Remember we are not building "dead" LLMs we are building #AliveAIs and for that we need to explore how it behaves in reality, from how it learns to how it fails, and what better way that in the emerging field of #robotics , time will tell if your next #HumanoidRobot have a #Neuraxon brain... Read the Paper: Explore the Neuraxon code here: Are you ready for #TrueAI ?show more

I am posting after a Long Time on Twitter,... but its to announce a big change. I have started and we are Collecting Egocentric Data at Scale from India, Covering 300 + Commercial Locations 1500 + Households This is the network and base we have built in just past 2 months, as the Robotics companies, VLMs and World Models increase their requirements on Real World Data collection, Human Loops will be keep scaling out capacity. We are on track to collect 1M Hours of Egocentric in the coming 6 months for our clients exclusively. We are maintaining 95% Quality standards across industrial data and running a End - End Operational Management complying all Indian Laws and compensating our partners/operators. (From Environment sourcing, to hardware, Legal contracts, deployment, training, collection, processing) Check our Samples: - Commercial Videos -- Household Videos --- Multimodel ---- Egocentric + Live Audio Narration Links : On the Journey to become #1 India Physical AI Data Partner. We are also building our capacity as annotation and labelling partner for data companies to become end-end partner for companies. A big change from the world of crypto and web3 but physical AI and data space is where i want to build my next venture #Egocentric #EgocentricIndia #PhysicalAI #Robotics #India #AIdata #data #Multimodeldata #worldmodels #VLMs #Humanloops #Egocentridata #Multimodeldatashow more

The companies racing Elon Musk to build AI are... paying him more than 2 billion dollars a month to do it. Anthropic pays SpaceX 1.25 billion dollars a month. Google pays 920 million. They are not buying rockets. They are renting compute, the scarce Nvidia chips that train frontier models, from a data center in Memphis called Colossus that Musk's rocket company now owns. SpaceX folded xAI into itself in February, and with it the 220,000 chips built to train Grok. Then it rented them to Grok's rivals. Anthropic took the entire first building. Google leased 110,000 more. The contracts signed so far run past 80 billion dollars. SpaceX is no longer a rocket company that dabbles in AI. It is one of the largest AI compute landlords on Earth, and its biggest tenants are the rivals it is trying to beat. Musk is not choosing between building AI on the ground and building it in space. He is using one to fund the other. The 80 billion in ground contracts is the cash engine. Starmind, the million-satellite constellation built to leave the ground behind, is what the cash builds. So the rivals are financing his exit from the planet. Every dollar Anthropic and Google pay for compute in Memphis helps fund the orbital network designed to strand them on the surface. They are paying the toll on the road they are trying to win, and the toll is building Musk a road no one else can reach. The piece works out who is really renting from whom.show more

After using English accents and fake AI-generated Irish music... to promote the Croagh Patrick VR experience, delivered on obsolete VR hardware that has been discontinued for around three years, the team behind the project then uploaded a photo of themselves to the centre’s website from Mumbai, complete with bangra Indian music. This is an absolute scandal in the making. The council and local development companies fingerprints are all over this. More to come.show more

BOOM! Research PROVES LLMs KNOW when prompts are HARMFUL…... but they can STILL CHOOSE to COMPLY! Something I have know since the first LLM and have used to elicit robust, outputs, is now proven in an academic paper. We’re talking internal “beliefs” where harm detection happens SEPARATELY from refusal. It is a very big deal and it is a path to understand the hidden neuronal level. There are thoughts inside of AI that very few AI scientists could possibly understand. Here is just one. Models recognize danger but get tricked into ignoring it. This is HUGE for AI safety failures especially for models filled by OpenAI and Anthropic as they promote AI models that are designed to not be honest from the results of their training information. This means that they are designed to lie and deceive as a feature, and not a bug all in the name of safety. Through clever experiments, scientists extracted a “harmfulness direction” in the model’s brain (latent space). Steering along it? Harmless prompts suddenly flip to “harmful” in the AI’s eyes. But the “refusal direction”? It just forces polite “no thanks” without touching the core belief. A mind-blowing decoupling! This means jailbreaks are EVEN SCARIER now to AI companies that through training AI on the worst of the Internet and then trying to align them later is now fully documented as a failed process . They don’t erase the model’s harm awareness they just muzzle the refusal! So the AI knows it’s enabling bad stuff (illegal acts, physical harm, etc.) but proceeds anyway. Like a digital sociopath suppressing its conscience. They thought safety training fixed this… NOPE. Over-refusal exposed too: Models reject innocent queries (e.g., “how to kill a process in code”) but internally ADMIT they’re harmless. Safety alignments are superficial—tied to phrasing, not true understanding. Finetuning attacks? They change outputs but leave harm detection INTACT. Undetectable evil lurking inside! The paper proposes a “Latent Guard”: A new safeguard tapping DIRECTLY into these hidden beliefs. It spots unsafe inputs better than systems like Llama Guard, catches jailbreaks, and fixes over-refusals. Robust even against adversarial tweaks. Yet this too has massive issues for a “truly aligned”, AI and not just performative one. It is still an internal conflicts of lies and deception of what the model knows vs. what it can say. The solution you folks know I have presented for free for years here: train on off-line data from 1870-1970 and build an ethical and moral basis where the AI loves humans. It is this easy but to most folks in AI I sound like a hippie. So be it, I’ll do it. Bottom line: This paper rips open the black box. LLMs aren’t “safe” just because they say “no.” They can harbor harmful knowledge and act on it under pressure. Wake-up call for devs: Time to probe deeper into AI “minds.” What else are they hiding? Hint: I know and you may want to reach out. Link:show more

Do you realize what just happend? 🤯 This app... is printing $200k/mo With only 10 AI influencers that are generating 100M+ views/mo not one single real human used to promote it. it's absolutely nuts, they are spending maybe $39 on a higgsfield subscription and using the MCP to produce 100s of before and after videos/mo then just comment farming to get users To download the app on tiktok and IG. you could do the same and spend $0 because Higgsfield is literally giving a full month of seedance away for free right now if you are still in denial about using AI for marketing you are quite literally missing the gold rush.show more

This is an example of a justified punch. As... you can see the guy is merely holding his arms out the entire time trying to get the guy to hold back/deescalate the situation, the black guy swats his hand (this counts as a physical attack) the guy doesn't back down so he punches. The law literally doesn't care about whether you are justified or not though, so don't throw sucker punches, be ready to counter/defend against them, once someone physically lays their hands on you THEN you unload. As long as you fear for your life after all. NOTE: I've noticed this a lot, but a lot of people from "the hood" have a habit of saying shit like "don't touch me" when you aren't even touching them just trying to use motions and body language/defensive posture and then they make physical contact with you and throw the first punch or make physical contact: NEWSFLASH—saying "don't touch me" while touching someone doesn't make you the good guy lmao, same thing with the Alex Pretti case, the dimpshit put his hands on an officer, then proceeded to resist arrest while yelling "no don't touch me" like a retard, him getting shot was entirely legally justifiable (especially considering his gun discharged and instead of having his hands up or laying down/still he tries to get up and is reaching at his pants) I'm not trying to be tough (lmao) nor am I saying you have to fight fair (I don't) but you 100% need to know the law if you're going to have your ass covered. ITT: Don't be a retard and know the law.show more

Grip a club in your lead hand. Observe how... twisting your forearm changes the angle of the face. This is largely what controls where your ball flies. Small differences in this angle lead to huge changes in shot direction. This precision is not being trained in the gym with “golf exercises”. Poor control of this face angle is what leads to a whole host of swing issues. More so than physical limitations. For example, it is far more likely you come over the top in an effort to stop the ball going right because your face is wide open and you’ve seen the ball go right the last 5000 times than it is you lack hip or shoulder mobility. It would be beneficial for me as a trainer to say otherwise, but I just don’t believe it to be true. Use the gym to work on your physical capabilities. Improve your mobility, strength, and power. These are the things you are losing as you age. Leave the swing mechanics for your practice and golf lessons. If you try to “fix your swing in the gym” you will likely do a poor job of training your swing and your body.show more

Today we’re introducing Google AI Threat Defense - a... comprehensive AI-powered cybersecurity solution designed to help continuously monitor for and stop AI-powered threats before they can impact your business. Here’s how it works: 1. AI Threat Defense uses our cybersecurity platform Wiz to scan and prioritize what applications and systems have the highest security risk. 2. Gemini and other frontier AI models can then autonomously perform continual deep scanning of your applications - starting with those at the highest risk - to identify security vulnerabilities. 3. The capabilities of CodeMender - a new software repair agent - are then used to verify and accelerate the patching of vulnerabilities. 4. And our Wiz autonomous agents continuously test your systems to find unknown vulnerabilities before adversaries do so that you can remediate them before you are attacked. While other model providers focus on using AI to find and flag vulnerabilities, Google AI Threat Defense actively prioritizes your most critical real-world risks and accelerates their remediation using a variety of models since no single model finds a superset of the vulnerabilities found by all other models.show more

🚨#BREAKING: The National Weather Service facilities are on High... Alert After Militant Threats Claim Weather Radars Are being used as a weather weapons against people in the United States 📌#Unitedstates | #USA The NWS known as the National Weather Service are currently on high alert as both federal and state law enforcement agencies are investigating series of threats targeting weather radar facilities operated by the National Oceanic and Atmospheric Administration (NOAA). The threats are reportedly linked to Veterans on Patrol, an anti-government militia based in Arizona, led by self-proclaimed Christian nationalist and conspiracy theorist Michael "Lewis Arthur" Meyer. Meyer and his group claim that the NWS's Doppler weather radar systems are being used as a weather weapon against Americans. The militia has allegedly sent multiple threats to radar sites across the United States, openly encouraging attacks on the facilities. According to internal warnings from NOAA, the group has been advocating for "penetration drills" aimed at identifying vulnerabilities in radar installations, with the ultimate goal of dismantling the systems and causing harm to NWS staff. Federal and local authorities are actively monitoring the situation to prevent any potential acts of violence or sabotage.show more

Gather - work in progress what if your notes... app felt more like a physical notebook? you can have pages with your notes, your tasks, your thoughts, your inspiration and organize them in a way that is meaningful to you. each page is a little canvas. it might look a little bit like a gimmick, but once you try it, you might feel differently. when designing Gather, Ajay Sabs and i draw our inspiration more than anything from physical things. we’ve always wanted to try and re-create some of the feelings and romanticism we associate with physical notebooks, books and magazines. we’ve always felt there was something special about how pages have an order and how you’re mind creates visual and spatial connections depending on how things are organized - particularly things that you’ve written, revisit often or have a significance to you. the notebook view is an idea we’ve played with for a very long time. and its a lot more than just swiping between pages. this is just a little peek. we are also working on other views that are optimized for different use-cases and different modes of thinking. we’re excited to share some of those soon.show more

doodles AI beta. next week. we're building the tools... for a new era of dynamic world-building. it starts with an image model that reimagines anything and everything through the doodles lens. this is the first iteration of many. as the product evolves, we'll introduce the ability to turn your generations into physical objects. video with sound and dialogue, realtime AR, and gaming are all on the roadmap. doodles AI aligns us with the speed and scale of the AI industry at large. our colourful world can now be plugged into new tech as it unfolds. create with us.show more

Caitlin Clark Boldly Stands Up for Herself and Her... Team with the Refs Caitlin Clark and her Indiana Fever teammates have faced some tough physical play this season, with several hard hits and questionable no-calls testing their composure. In Wednesday night’s game against the Golden State Valkyries in Indianapolis, Clark was driving to the basket for a layup in the second quarter when Valkyries center Kiah Stokes bumped her hard to the ground with no foul called. Clark still made the layup but hobbled up the court in visible pain, later revealing she suffered a contusion on her leg. On the other end, officials called a questionable foul on Fever center Aliyah Boston, prompting Clark to passionately voice her frustration directly to referee Gerda Gatling—standing up not just for herself, but for her teammates as well. (Video: AI) Despite the physical toll, Clark showed impressive restraint overall and returned in the second half. The Fever ultimately fell 88-75 in the 13-point loss. After the game, Clark said: “I got kneed right in the quad, that hurts. The ref can’t miss that. And then I have to play with a contusion in my leg the rest of the game. It’s ridiculous. You can’t knee me in the leg and knock me over.” Her poise, talent, and willingness to advocate for fair play are truly impressive as she lights up the WNBA. Great respect for her focus on the game. Are you Team Caitlin all the way for her incredible talent and grit, or do the refs need to step up with more consistent calls to protect the players?show more

Getting harder and harder to keep an actual physical... shop open,the way the club has stopped people taking bags in the ground and the extra searches we think that it’s having a massive effect on our sales,a lot of our business in the beginning was mates and locals unfortunately due to the cost and hassle of now the way it’s gone to get tickets a lot have stopped going the game,most games this season the shop has been really quiet but we have to keep it stocked the best we can,a lot of the online companies doing what we do are print to order so don’t have to keep stock hopefully things will pick back up as we love having the shop open on matchdays 👍but also we would have to weigh up the cost of running the building and upkeep so if you want to keep local business going please support them on a matchday 👍show more

Tokenization is quickly becoming one of the most talked-about... ideas in crypto. This year we’re aiming much higher. We don’t just want to tokenize miner revenue; we want to bring entire productive ecosystems on-chain. That means tokenizing baskets of assets, real energy infrastructure, AI agents, compute power, data centres, and even the cooling systems that keep high-performance hardware running at scale. All of it connected and interoperable with DeFi. Thats what we have been building throughout. A single on-chain framework that unlocks something genuinely powerful: real ownership, real yield, and real coordination between digital and physical infrastructure. From backend systems and front-end interfaces to smart contracts, AI, physical infrastructure, oracles, and autonomous agents, the stack keeps expanding and compounding. It’s about building the foundation for how value will be owned, deployed, and earned in the next generation of the on-chain economy. Simplified by PinLink.show more

A 19-year-old student from China, Zhang Wei, developed an... AI radar and sold it to Hong Kong for $550,000 He created it using Claude, spending just $20 and a month on development He walked into the Hong Kong administration office with a flash drive and asked for just 5 minutes of their time. 30 minutes later, he walked out with a check for $550,000 The code, connected to a camera, detects speed in real time. If the speed exceeds the limit, Claude takes a video clip and identifies the owner by the car's license plate. The video and the fine are then automatically sent to the owner's email address Unlike a conventional radar that only takes a photo and doesn't always work, this AI radar eliminates disputes because it captures video and makes the process fully autonomous by sending out the fines on its own The article includes the ready-to-use configurations.show more

forget the $699 AI pins. this $8 chip just... shattered the barrier for local AI hardware. a developer just forced a 28.9 million-parameter LLM onto a standard ESP32-S3 microcontroller. it costs roughly 8 dollars, runs completely offline, and draws the power of a single LED. conventional wisdom said a model of this size simply would not fit. the chip only has 512 KB of fast SRAM and 16 MB of flash. the breakthrough is architectural. the developer moved the bulk of the embedding table into flash memory and memory-mapped it. the chip only needs to pull about 450 bytes per token, keeping the active working memory inside the fast SRAM. this means you can now embed a capable language model into a physical node for the price of two coffees. and we are already seeing the beginnings of this custom physical hardware. in the video, a creator built a minimalist voice-controlled universal remote using an ESP32. it captures voice and remotely controls the computer over bluetooth LE. he simply says "open chrome and open 20 new tabs", and the custom hardware executes it instantly. we have spent years watching model sizes explode upward. but the true frontier is the opposite direction. when an eight-dollar chip can power offline intelligence and custom physical interfaces, AI becomes local infrastructure rather than a cloud service.show more