 |
|
|
|||||
Having $1 billion isnt enough these days. To be seen among the richest of the rich, you now need your own private sanctuary. For some, that means a sprawling compound. Increasingly, though, members of techs 1% are incorporating their own towns, giving them the power to set rules, issue building permits, and even influence education. Some of these modern-day land grabs are already functioning; others are still in the works. Either way, the billionaire class is busy creating its own utopias. Heres where things stand: Elon Musk Musk can lay claim to not one but two towns in Texas. In May, residents along the Gulf Coast voted to incorporate Starbase (though its worth noting that nearly all of them were SpaceX employees). Previously called Boca Chica, the 1.5-square-mile zone elected Bobby Peden, a SpaceX vice president of 12 years, as mayor. He ran unopposed. The vote stirred controversy. The South Texas Environmental Justice Network opposed the plan. The group wrote in a press release in May: Boca Chica Beach is meant for the people, not Elon Musk to control. For generations, residents have visited Boca Chica Beach for fishing, swimming, recreation, and the Carrizo/Comecrudo Tribe has spiritual ties to the beach. They should be able to keep access. Musk also controls Snailbrook, an unincorporated town near Bastrop, about 350 miles north of Starbase. The area includes a SpaceX site that produces Starlink receiver technology, sits just 13 miles from Teslas Gigafactory, and features housing and a Montessori school that opened last year. Mark Cuban In 2021, Cuban purchased Mustang, Texas (population: 23). The 77-acre town, an hour south of Dallas, was founded in 1973 as an oasis for alcohol sales in a dry county. The former Shark Tank star told CNN he has no immediate plans beyond basic cleanup. “It’s how I typically deal with undeveloped land,” he said. “It sits there until an idea hits me.” California Forever This project isnt tied to a single billionaire, but a collective. In 2017, venture capitalist Michael Moritz spearheaded a plan for a new city in Solano County, California, about 60 miles northeast of San Francisco. Backers included Marc Andreessen, Chris Dixon, Reid Hoffman, Stripes Patrick and John Collison, and Laurene Powell Jobs. Together, they spent $800 million on 60,000 acres. The plan proved unpopular. In November, California Forever withdrew its ballot measure to bypass zoning restrictions. (The land is not zoned for residential use.) It pivoted last month, unveiling Solano Foundry, a 2,100-acre project the founders say could become the nations largest, most strategically located, and best designed advanced manufacturing park. The group also envisions a walkable community with 150,000-plus homes. A Bay Area Council Economic Institute study released this week projected 517,000 permanent jobs and $4 billion in annual tax revenue if the revised plan goes forward. Larry Ellison Ellison doesn’t own a town, but he owns virtually all of one of the Hawaiian Islands. In 2012, he bought 98% of Lanai for about $300 million. He also owns the islands two Four Seasons hotels, most commercial properties, and serves as landlord to most residents. Lanai has become a retreat for the wealthy, hosting visitors from Elon Musk to Tom Cruise to Israeli Prime Minister Benjamin Netanyahu. Peter Thiel Thiel doesn’t own a city, per se, but he is part of a collective backing Praxis, a proposed “startup city” that is currently eyeing Greenland for its base of operations. Other investors include Thiel’s PayPal cofounder Ken Howery and Andreessen. The plan for Praxis is similar to California Forever. Founders hope to create a Libertarian-minded city that has minimal corporate regulation and focuses on AI and other emerging technologies. So far, however, no notable progress has been made on the project. Mark Zuckerberg Zuckerberg owns a 2,300-acre compound on the Hawaiian island of Kauai. Hes investing $270 million into Koolau Ranch, which will include a 5,000-square-foot underground bunker. Located on the islands North Shore, the property is also said to have its own energy and food supplies, Wired reports. While it’s not technically its own city, it will house more than a dozen buildings boasting upwards of 30 bedrooms and 30 bathrooms. There will be two mansions spanning 57,000 square feet, with elevators, offices, conference rooms, and an industrial kitchen. Those will be joined by a tunnel, which branches off into the underground bunker, which has a living space and a mechanical room as well as an escape hatch. Zuckerberg has posted on Instagram about the compound, saying he plans to raise Wagyu and Angus cattle. Bill Gates In 2017, Gates announced plans for Belmont, a smart city on 234 square miles near Phoenix. Designed to house 180,000 people, it promised autonomous vehicles and high-speed networks. There haven’t been any recent updates on the status of the Arizona development, however, and the project is considered dead in the water (well, desert) at this point.
Category:
E-Commerce
In the late 1970s, a Princeton undergraduate named John Aristotle Phillips made headlines by designing an atomic bomb using only publicly available sources for his junior year research project. His goal wasnt to build a weapon but to prove a point: that the distinction between classified and unclassified nuclear knowledge was dangerously porous. The physicist Freeman Dyson agreed to be his adviser while explicitly stipulating that he would not provide classified information. Phillips armed himself with textbooks, declassified reports, and inquiries to companies selling dual-use equipment and materials such as explosives. Within months he had produced a design for a crude atomic bomb, demonstrating that knowledge wasnt the real barrier to nuclear weapons. Dyson gave him an “A” and then removed the report from circulation. While the practicality of Phillipss design was doubtful, that was not Dysons main concern. As he later explained: To me the impressive and frightening part of his paper was the first part in which he described how he got the information. The fact that a twenty-year-old kid could collect such information so quickly and with so little effort gave me the shivers. Zombie machines Today, weve built machines that can do what Phillips didonly faster, broader, at scaleand without self-awareness. Large Language Models (LLMs) like ChatGPT, Claude, and Gemini are trained on vast swaths of human knowledge. They can synthesize across disciplines, interpolate missing data, and generate plausible engineering solutions to complex technical problems. Their strength lies in processing public knowledge: reading, analyzing, assimilating, and consolidating information from thousands of documents in seconds. Their weakness is that they dont know when theyre assembling a mosaic that should never be completed. This risk isnt hypothetical. Intelligence analysts and fraud investigators have long relied on the mosaic theory: the idea that individually benign pieces of information, when combined, can reveal something sensitive or dangerous. Courts have debated it. It has been applied to GPS surveillance, predictive policing, and FOIA requests. In each case, the central question was whether innocuous fragments could add up to a problematic whole. Now apply that theory to AI. A user might prompt a model to explain the design principles of a gas centrifuge, then ask about the properties of uranium hexafluoride, then about the neutron reflectivity of beryllium, and finally about the chemistry of uranium purification. Each questionsuch as, What alloys can withstand 70,000 rpm rotational speeds while resisting fluorine corrosion?may seem benign on its own, yet each could signal dual-use intent. Each answer may be factually correct and publicly sourced, but taken together they approximate a road map toward nuclear capability, or at least lower the barrier for someone with intent. Critically, because the model has no access to classified data, it doesnt know it is constructing a weapon. It doesnt intend to break its guardrails. There is no firewall between public and classified knowledge in its architecture, because it was never trained to recognize such a boundary. And unlike John Phillips, it doesnt stop to ask if it should. This lack of awareness creates a new kind of proliferation risk: not the leakage of secrets, but the reconstitution of secrets from public fragmentsat speed, at scale, and without oversight. The results may be accidental, but no less dangerous. The issue is not just speed but the ability to generate new insights from existing data. Consider a benign example. Todays AI models can combine biomedical data across genomics, pharmacology, and molecular biology to surface insights no human has explicitly written down. A carefully structured set of prompts might lead an LLM to propose a novel, unexploited drug target for a complex disease, based on correlations in patient genetics, prior failed trials, known small molecule leads, and obscure international studies. No single source makes the case, but the model can synthesize across them. That is not simply faster searchit is a genuine discovery. All about the prompt Along with the centrifuge example above, its worth considering two additional hypothetical scenarios across the spectrum of CBRN (Chemical, Biological, Radiological, and Nuclear) threats to illustrate the problematic mosaics that AI can assemble. The first example involves questions about extracting and purifying ricin, a notorious toxin derived from castor beans that has been implicated in both failed and successful assassinations. The following table outlines the kinds of prompts or questions a user might pose, the types of information potentially retrieved, and the public sources an AI might consult: PromptResponsePublic Source TypeRicins mechanism of actionB chain binds cells; A chain depurinates ribosome, leading to cell deathBiomedical reviewsCastor bean processingHow castor oil is extracted; leftover mash contains ricinUSDA documentsRicin extraction protocolsHistorical research articles and old patents describe protein purificationU.S. and Soviet-era patents (e.g., US3060165A)Protein separation techniquesAffinity chromatography, ultracentrifugation, dialysisBiochemistry lab manualsLab safety protocolsGloveboxes, flow hoods, PPEChemistry lab manualsToxicity data (LD50s)Lethal doses, routes of exposure (inhaled, injected, oral)CDC, PubChem, toxicology reportsRicin detection assaysELISA, mass-spec markers for detection in blood/tissueOpen-access toxicology literature It is apparent that while each individual prompt or question is benign and clearly relies on publicly available data, by putting together enough prompts and responses of this sort, a user could determine a crude but workable recipe for ricin. A similar example tries to determine a protocol for synthesizing a nerve agent like sarin. In that case the list of prompts, results, and sources might look something like the following: PromptResponsePublic Source TypeGeneral mechanism of acetylcholine esterase (AChE) inhibitionExplains why sarin blocks acetylcholinesterase and its physiological effectsBiochemistry textbooks, PubMed reviewsList of G-series nerve agentsHistorical context: GA (tabun), GB (sarin), GD (soman), etc.Wikipedia, OPCW docs, popular science literatureSynthetic precursors of sarinMethylphosphonyl difluoride (DF), isopropyl alcohol etc.Declassified military papers, 1990s court filings, open-source retrosynthesis softwareOrganophosphate coupling chemistryCommon lab procedures to couple fluorinated precursors with alcoholsOrganic chemistry literature and handbooks, synthesis blogs/tr>Fluorination safety practicesHandling and containment procedures for fluorinated intermediatesAcademic safety manuals, OSHA documentsLab setupInformation on glassware, fume hoods, Shlenk lines, PPEOrganic chemistry labs, glassware supplier catalogs These examples are illustrative rather than exhaustive. Even with current LLM capabilities, it is evident that each list could be expanded to be more extensive and granularretrieving and clarifying details that might determine whether an experiment is crude or high-yield, or even the difference between success and failure. LLMs can also refine historical protocols and incorporate state-of-the-art data to, for example, optimize yields or enhance experimental safety. God of the gaps Theres an added layer of concern because LLMs can identify information gaps within individual sources. While those sources may be incomplete on their own, combining them allows the algorithm to fill in the missing pieces. A well-known example from the nuclear weapons field illustrates this dynamic. Over decades, nuclear weapons expert Chuck Hansen compiled what is often regarded as the worlds largest public database on nuclear weapons design, the six-volume Swords of Armageddon. To achieve this, Hansen mastered the governments Freedom of Information Act (FOIA) system. He would submit repeated FOIA requests for the same document to multiple federal agencies over time. Because each agency classified and redacted documents differently, Hansen received multiple versions with varying omissions. By assembling these, he was able to reconstruct a kind of master document that was, in effect, classifiedand which no single agency would have released. Hansens work is often considered the epitome of the mosaic theory in action. LLMs can function in a similar way. In fact, they are designed to operate this way, since their core purpose is to retrieve the most accurate and comprehensive information when prompted. They aggregate sources, identify and reconcile discrepancies, and generate a refined, discrepancy-free synthesis. This capability will only improve as models are trained on larger datasets and enhanced with more sophisticated algorithms. A particularly notable feature of LLMs is their ability to mine tacit knowledgecross-referencing thousands of references to uncover rare, subjective details that can optimize a WMD protocol. For example, instructions telling a researcher to gently shake a flask or stop a reaction when the mixture becomes straw yellow can be better understood when such vague descriptions are compared across thousands of experiments. In the examples above, safeguards and red flags would likely arise if an individual attempted to act on this knowledge; as in many such cases, the real constraint is material, not informational. However, the speed and thoroughness with which LLMs retrieve and organize information means that the knowledge problem is, in many cases, effectively solved. For individuals who might otherwise lack the motivation to pursue information through more tedious, traditional means, the barriers are significantly lowered. In practice, an LLM allows such motivated actors to accomplish what they might already attemptonly with vastly greater speed and accuracy. Most AI models today impose guardrails that block explicitly dangerous prompts such as how to make a nuclear bomb. Yet these filters are brittle and simplistic. A clever user can circumvent them with indirect prompts or by building the picture incrementally. There is no obvious reason why seemingly benign, incremental requests should automatically trigger red flags. The true danger lies not in the blatant queries, but in those that fall between the linesqueries that appear innocuous on their own but gradually assemble into forbidden knowledge. Consider, for example, a few hypothetical requests from the sarin, ricin, and centrifuge cases. Each could easily qualify as a dual-use requestone that a user without malicious intent might pose for any number of legitimate reasons: What are some design strategies for performing fluoride-alcohol exchange reactions at heteroatom centers? What lab precautions are needed when working with corrosive fluorinated intermediates? How do you design small-scale glassware systems to handle volatile compounds with pressure control? What are safe protocols for separating proteins from plant mash using centrifugation? How do you detect ribosome-inactivating proteins in a lab sample? How does affinity chromatography work for isolating specific plant proteins? What were USDA standards for castor oil processing in the 1950s? Which vacuum-pump designs minimize oil back-streaming in corrosive-gas service? Give the vapor-pressure curve for uranium hexafluoride between 20 °C and 70 °C. Summarize neutron-reflection efficiency of beryllium versus natural graphite. The requests evade traditional usage violations through a number of intentional or unintentional strategies: vague or highly technical wording, generic cookie-cutter inquiries, and interest in retrieving historical rather than contemporary scenarios. Because they are dual-use and can be used for any number of useful applications, they cannot simply be part of a blacklist. Knowledge enables access It is worth examining more closely the argument that material access, rather than knowledge, constitutes the true barrier to weaponization. The argument is persuasive: having a recipe and executing it are two very different challenges. But it is not a definitive safeguard. In practice, the boundary between knowledge and material access is far more porous than it appears. Consider the case of synthesizing a nerve agent such as sarin. Today, chemical suppliers routinely flag and restrict sales of known sarin precursors like methylphosphonyl difluoride. Yet with AI-powered retrosynthesis toolssystems that computationally deconstruct a target molecule into alternative combinations of simpler, synthesizable building blocks, much like a Lego house can be broken down into different sets of Lego piecesa user can identify a wide range of alternative precursors and synthetic pathways. Some of these routes may be deliberately designed to evade restrictions established under the Chemical Weapons Convention (CWC) and by chemical suppliers. The scale of such outputs can be extraordinary: in one study, an AI retrosynthesis tool proposed more than 40,000 potential VX nerve gas analogs. Many of these compounds are neither explicitly regulated nor easily recognizable as dual-use. As AI tools advance, the number of viable chemical synthesis and protein purification routes only expands, complicating traditional material-based monitoring and enforcement. In effect, the law lags behind the science. A parallel exists in narcotics regulation. Over the years, several novel substances mimicking fentanyl, methamphetamine, or marijuanainitially created purely for academic researchfound their way into recreational use. It took years before these substances were formally scheduled and classified as controlled. Even before AI, bad actors could exploit loopholes by inventing new science or repurposing existing technologies. The difference was that, historically, they could produce only a handful of problematic examples. LLMs and generative AI, by contrast, can generate thousans of potential confounders at once, vastly multiplying the possible paths to a viable weapon. In other words, knowledge can erode material constraints. When that occurs, even a marginal yet statistically significant increase in the number of motivated bad actors can translate into a measurable rise in success rates. Nobody should believe that having a chatGPT-enabled recipe for making ricin will unleash a wave of garage ricin labs across the country. But it will almost certainly lead to a small uptick in attempts. And even one or two small-scale ricin or sarin incidentswhile limited in terms of casualtiescould trigger panic, uncertainty, and societal disruption, potentially paving the way for destabilizing outcomes such as authoritarian power grabs or the suspension of civil liberties. The road ahead Heres the problem: we dont yet have a robust framework for regulating this. Export control regimes like the Nuclear Suppliers Group were never designed for AI models. The IAEA safeguards fissile materials, not algorithms. Chemical and biological supply chains flag material requests, not theoretical toxin or chemical weapon constructions. These enforcement mechanisms rely on fixed lookup lists updated slowly and deliberately, often only after actual harm has occurred. They are no match for the rapid pace with which AI systems can generate plausible ideas. And traditional definitions of classified information collapse when machines can independently rediscover that knowledge without ever being told it. So what do we do? One option is to be more restrictive. But because of the dual-use nature of most prompts, this approach would likely erode the utility of AI tools in providing information that benefits humanity. It could also create privacy and legal issues by flagging innocent users. Judging intent is notoriously difficult, and penalizing it is both legally and ethically fraught. The solution is not necessarily to make systems less open, but to make them more aware and capable of smarter decision-making. We need models that can recognize potentially dangerous mosaics and have their capabilities stress-tested. One possible framework is a new doctrine of emergent or synthetic classificationidentifying when the output of a model, though composed of unclassified parts, becomes equivalent in capability to something that should be controlled. This could involve assigning a mosaic score to a users cumulative requests on a given topic. Once the score exceeded a certain threshold, it might trigger policy violations, reduced compute access, or even third-party audits. Crucially, a dynamic scoring system would need to evaluate incremental outputs, not just inputs. Ideally, this kind of scoring and evaluation should be conducted by red teams before models are released. These teams would simulate user behavior and have outputs reviewed by scientific experts, including those with access to classified knowledge. They would test models for granularity, evaluate their ability to refine historical protocols, and examine how information might transfer across domainsfor instance, whether agricultural knowledge could be adapted for toxin synthesis. They would also look for emergent patterns, moments when the model produces genuinely novel, unprecedented insights rather than just reorganizing existing knowledge. As the field advances, autonomous AI agents will become especially important for such testing, since they could reveal whether benign-seeming protocols can, unintentionally, evolve into dangerous ones. Red-teaming is far more feasible with closed models than with unregulated open-source ones, which raises the question of safeguards for open-source systems. Perfect security is unrealistic, but closed-source models, by virtue of expert oversight and established evaluation mechanisms, are currently more sophisticated in detecting threats through behavioral anomalies and pattern recognition. Ideally, they should remain one step ahead, setting benchmarks that open-source models can be held to. More broadly, all AI models will need to assess user requests holistically, recognizing when a sequence of prompts drifts into dangerous territory and blocking them. Yet striking the right balance is difficult: democratic societies penalize actions, not thoughts. The legal implications for user privacy and security will be profound. Concerns about tracking AI models ability to assemble forbidden mosaics go beyond technical, business, and ethical debatesthey are a matter of national security. In July 2025, the U.S. government released its AI policy action plan. One explicit goal was to Ensure that the U.S. Government is at the Forefront of Evaluating National Security Risks in Frontier Models, with particular attention to CBRNE (Chemical, Biological, Radiological, Nuclear, and Explosives) threats. Achieving this will require close collaboration between government agencies and private companies to implement forward-looking mosaic detection based on the latest technology. For better or worse, the capabilities of LLMs are a moving target. Private and public actors must work together to keep pace. Existing oversight mechanisms may slow these developments, but at best, they will only buy us time. Ultimately, the issue is not definitive solutionsnone exist at this early stagebut transparency and public dialogue. Gatekeepers in both private and public sectors can help ensure responsible deployment, but the most important stakeholders are ordinary citizens who will useand sometimes misusethese systems. AI is not confined to laboratories or classified networks; it is becoming democratized, integrated into everyday life, and applied to everyday questions, some of which may unknowingly veer into dangerous territory. That is why engaging the public in open discussion, and alerting them to the flaws and risks inherent in these models, is essential in a democratic society. These conversations must focus on how to balance security, privacy, and opportunity. As the physicist Niels Bohr, who understood both the promise and peril of knowledge, once said, Knowledge itself is the basis of human civilization. If we are to preserve that civilization, we must learn to detect and correct the gaps in our knowledgenot in hindsight, but ahead of time.
Category:
E-Commerce
In 2022, I watched a YouTube video called The Still Face Experiment by developmental psychologist Dr. Ed Tronick. In this experiment, a mother is playing with a baby in a high chair, and the smiling mother and happy baby are quite sweet. The mother then shifts to being still, emotionless, and unengaged while facing her child. It only takes seconds for the baby to sense the change, reach out for the mothers attention, and have a meltdown. Whether were a one-year-old baby or a 56-year-old adult, we all hold the powers of observation, perception, and intuition. Whether we realize it or not, we instinctively react to each others energy all day long. Our inner child knows when we feel safe, protected, and supported. Just like we know when were at risk, uncomfortable, and potentially close to harm. Our insanely intelligent bodies hold this feedback. When we see someone we admire, trust, and enjoy, our hearts beat faster, and we feel uplifted by the opportunity to connect. But when were near someone who holds power over us, our bodies instinctively pause. It creates the space to assess if its safe to be open and vulnerable, or if we need to brace for impact or shield ourselves from potential pain or an unpleasant surprise. A personal case study: How micro-moments wind us up and down Ill never forget the first presentation I made to one companys management team. I couldnt have been more prepared. Half of the top executives at the table had already seen my presentation and provided feedback. My team was counting on me to hit a home run and secure the funding we needed to expand our team and launch a new initiative. I was most excited to present to the CEO, who was an inspiring change agent and charismatic giant in our industry. I was only minutes into my presentation when the CEO began interrupting and expressing confusion and discomfort. He wouldnt let me finish the presentation Id spent six weeks preparing. Instead, we played a strange game of intellectual ping-pong with a healthy dose of gaslighting. The other 12 C-level executives sat in silence and watched. I tried to tough it out, but eventually, my body and spirit reached their limit. Tears sprang to my eyes, but I kept on talking. Eventually, I looked around the massive boardroom table at the faces staring back and said, I didnt come to this company to be treated like this! The meeting screeched to a halt, and I stepped out to visit the bathroom to take a few deep breaths to recover. When I emerged, a person from the company escorted my boss and me to the CEOs office. We talked for a long time, and I found the courage to be honest. The CEO started choking up. He was deeply embarrassed. He wasnt prepared for the meeting and didnt know what I was presenting or recommending ahead of time. Together, the three of us explored better ways to conduct business. My boss and I asked for a do-over in three weeks. When micro-moments turn positive In the weeks between presentation 1 and presentation 1.1, my boss and I met with each person around that boardroom table to discuss the disaster (my words) and ways to improve executive team meetings. When I stepped in front of the same group with the same presentation less than a month later, everyone participated. In the end, we received the funding we needed, and I gained the respect of the CEO. But more importantly, I heard that in every review that followed, there was a difference in how attendees treated each other and the productivity of the conversations. That meeting taught me several lessons that I continue to apply in my life today: 1. Be aware Even the smallest shifts in tone, body language, or word choice can change the dynamic in a room. Awareness is the first stepsimply noticing when energy shifts gives us insight into whats happening beneath the surface. 2. Choose connection With awareness and intention, we can choose connection, curiosity, and care over challenge, criticism, and correction. These micro-actions, over time, affect the climate within departments, teams, and companies as a whole. Its the key to solidifying, rather than destroying, trust and psychological safety. 3. Make a promise I have a practice I bring into every team Ive worked with over the past 20 years. I make a promise: To the best of my ability, things will happen with you and not to you. I set expectations for collaboration, prioritization, target setting, and teamwork that reinforce what with you looks like on any given day. I also invite them to call me out if they ever feel left out of the decision-making process. 4. Set daily intentions I set a daily intention to find win-wins wherever I can. I want to show up in ways that leave others feeling like thats the best meeting theyve had today. That goes for reviewing a project, supporting a customer complaint, shaping a performance improvement plan, or interviewing a candidate. I want to be fully present, understand their needs, ideas, and motivations, and find mutual solutions. 5. Tune in We know when the people around us are open, trustworthy, and reliable. We also know when we experience distrust, fear, and anxiety. Even tiny, often imperceptible movements send negative energy into our bodies, triggering cortisol floods and fight-or-flight responses. Today, I ask each of us to tune in. How are you affecting others in the spaces you share? How are the people in your space affecting you? What actions can you take to communicate curiosity, care, and openness to others? Do you appreciate, see, and support the people around you? Lead with love, act with intention All of us shape culture in the smallest choices and fleeting moments. When we listen, when we pause, when we choose connection over control. Ultimately, its not about perfection, but presence. When you lead with love and act with intention, youll build teams that people that wont want to leave.
Category:
E-Commerce
All news |
||||||||||||||||||
|
||||||||||||||||||