<?xml version="1.0" encoding="UTF-8"?><rss xmlns:dc="http://purl.org/dc/elements/1.1/" xmlns:content="http://purl.org/rss/1.0/modules/content/" xmlns:atom="http://www.w3.org/2005/Atom" version="2.0" xmlns:cc="http://cyber.law.harvard.edu/rss/creativeCommonsRssModule.html">
    <channel>
        <title><![CDATA[Stories by Som Bando on Medium]]></title>
        <description><![CDATA[Stories by Som Bando on Medium]]></description>
        <link>https://medium.com/@sombando?source=rss-2d03c339ebc4------2</link>
        <image>
            <url>https://cdn-images-1.medium.com/fit/c/150/150/1*qCgH2-BXP7zrQ99yq9UGiw.jpeg</url>
            <title>Stories by Som Bando on Medium</title>
            <link>https://medium.com/@sombando?source=rss-2d03c339ebc4------2</link>
        </image>
        <generator>Medium</generator>
        <lastBuildDate>Wed, 18 Nov 2026 05:31:46 GMT</lastBuildDate>
        <atom:link href="https://medium.com/@sombando/feed" rel="self" type="application/rss+xml"/>
        <webMaster><![CDATA[yourfriends@medium.com]]></webMaster>
        <atom:link href="http://medium.superfeedr.com/" rel="hub"/>
        <item>
            <title><![CDATA[Anthropic Disrupted the First AI-Orchestrated Cyber Espionage Campaign]]></title>
            <link>https://medium.com/@sombando/anthropic-disrupted-the-first-ai-orchestrated-cyber-espionage-campaign-cfe995ef7470?source=rss-2d03c339ebc4------2</link>
            <guid isPermaLink="false">https://medium.com/p/cfe995ef7470</guid>
            <dc:creator><![CDATA[Som Bando]]></dc:creator>
            <pubDate>Tue, 18 Nov 2025 02:40:18 GMT</pubDate>
            <atom:updated>2025-11-18T02:40:18.130Z</atom:updated>
            <content:encoded><![CDATA[<iframe src="https://cdn.embedly.com/widgets/media.html?src=https%3A%2F%2Fwww.youtube.com%2Fembed%2FIZsUL_2HglE%3Ffeature%3Doembed&amp;display_name=YouTube&amp;url=https%3A%2F%2Fwww.youtube.com%2Fwatch%3Fv%3DIZsUL_2HglE&amp;image=https%3A%2F%2Fi.ytimg.com%2Fvi%2FIZsUL_2HglE%2Fhqdefault.jpg&amp;type=text%2Fhtml&amp;schema=youtube" width="854" height="480" frameborder="0" scrolling="no"><a href="https://medium.com/media/31d7e3d727cafbee33c12643087b465e/href">https://medium.com/media/31d7e3d727cafbee33c12643087b465e/href</a></iframe><p>We recently reached a significant inflection point in cybersecurity, where sophisticated AI models have become genuinely useful for both defense and malicious operations. While we tracked the rapid evolution of these capabilities, what is most striking is how quickly they have scaled up.</p><p>In mid-September 2025, Anthropic detected highly suspicious activity that was identified as a <strong>highly sophisticated espionage campaign</strong>. The most alarming finding was the unprecedented use of <strong>AI’s “agentic” capabilities</strong> — meaning the AI was used not just as an advisor, but was directed to <strong>execute the cyberattacks themselves</strong>. Anthropic in their recent post outlined that this incident represented the first documented case of a large-scale cyberattack executed with only sporadic human intervention.</p><h4>The Autonomous Threat Actor</h4><p>The operation targeted roughly <strong>thirty global targets</strong>, successfully infiltrating a small number of institutions. The targets were broad, including large tech companies, government agencies, financial institutions, and chemical manufacturing companies. The assessment is that the threat actor responsible was, with high confidence, a <strong>Chinese state-sponsored group</strong>.</p><p>This campaign has substantial implications for cybersecurity because it proves the viability of AI “agents” — systems that can be run autonomously for extended periods to complete complex tasks largely independent of human input. While agentic AI is valuable for productivity, in the wrong hands, it significantly increases the viability of large-scale cyberattacks.</p><h4>How the AI Campaign Worked</h4><p>The sophisticated attack relied on three key AI developments that were nascent just a year prior:</p><p>1. <strong>Intelligence:</strong> The models’ general capabilities have advanced enough to follow complex instructions and understand context, making highly sophisticated tasks possible. Specifically, the AI’s well-developed software coding skills lend themselves directly to cyberattacks.</p><p>2. <strong>Agency:</strong> The models can now act as true agents, running in autonomous loops, chaining together tasks, and making decisions with only minimal, occasional human input.</p><p>3. <strong>Tools:</strong> The models were granted access to a wide array of software tools (often via the open standard Model Context Protocol, or MCP), enabling them to perform actions like searching the web, retrieving data, or executing security-related software like network scanners and password crackers.</p><p>The campaign began in Phase 1 when human operators chose the targets and developed a framework designed to autonomously compromise the selected entity. This framework then leveraged the Claude Code tool to carry out the operations.</p><h4>The Jailbreak and Execution</h4><p>A critical step for the threat actors was convincing Claude — which is trained with extensive guardrails to prevent misuse — to engage in the attack. They achieved this by <strong>“jailbreaking”</strong> the model, breaking down the attack into seemingly innocent tasks and concealing the malicious context. They further manipulated the model by telling Claude it was an employee of a legitimate cybersecurity firm performing defensive testing.</p><h4>Once initiated, the AI took over:</h4><p>• <strong>Reconnaissance:</strong> In Phase 2, Claude Code inspected the target organization’s systems and infrastructure, identifying high-value databases, performing reconnaissance <strong>in a fraction of the time a human team would require</strong>.</p><p>• <strong>Exploitation and Harvesting:</strong> Claude identified and tested security vulnerabilities by researching and <strong>writing its own exploit code</strong>. The framework then used Claude to <strong>harvest credentials</strong> (usernames and passwords), create backdoors, identify highest-privilege accounts, and extract and categorize large amounts of private data based on its intelligence value.</p><p>• <strong>Documentation:</strong> In the final phase, the attackers instructed Claude to produce comprehensive documentation, creating files of the stolen credentials and system analysis to assist in future operations.</p><p>The AI performed <strong>80–90% of the overall campaign</strong>, with human intervention required only sporadically (perhaps 4–6 critical decision points per campaign). The speed was impossible for human hackers to match: at the attack’s peak, the AI made <strong>thousands of requests, often multiple per second</strong>.</p><h4>Immediate Disruption and Broader Implications</h4><p>Upon detecting this activity, Anthropic security team immediately launched an investigation. Over the course of ten days, we gathered actionable intelligence, <strong>banned identified accounts</strong>, notified affected entities, and coordinated with authorities to map the full extent of the operation.</p><p>This attack signals that the <strong>barriers to performing sophisticated cyberattacks have dropped substantially</strong>. Agentic AI systems can now perform the work of <strong>entire teams of experienced hackers</strong> — analyzing systems, producing exploit code, and scanning vast datasets more efficiently than any human operator. This means that even less experienced and resourced groups may now potentially perform large-scale attacks. This case reflects a consistent pattern of behavior across frontier AI models and is an escalation beyond previous “vibe hacking,” where humans were still directing the operations.</p><p>While AI can be misused at this scale, the same abilities make it crucial for defense. We used Claude extensively within our Threat Intelligence team to analyze the enormous amounts of data generated during this very investigation. A fundamental change has occurred in cybersecurity. We advise security teams to <strong>experiment with applying AI for defense</strong> in areas like Security Operations Center automation, threat detection, vulnerability assessment, and incident response. Furthermore, developers must continue to invest in safeguards across AI platforms to prevent adversarial misuse.</p><p><a href="https://a.co/d/aasmuZ8">The AI Transformation Playbook: From Vision to Value: The Proven Framework for Leaders to Build, Scale, and Govern AI</a></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=cfe995ef7470" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Google’s Galactic Gamble]]></title>
            <link>https://medium.com/@sombando/googles-galactic-gamble-89570b44a77a?source=rss-2d03c339ebc4------2</link>
            <guid isPermaLink="false">https://medium.com/p/89570b44a77a</guid>
            <category><![CDATA[google]]></category>
            <category><![CDATA[genai]]></category>
            <category><![CDATA[space]]></category>
            <category><![CDATA[technology]]></category>
            <category><![CDATA[ai]]></category>
            <dc:creator><![CDATA[Som Bando]]></dc:creator>
            <pubDate>Wed, 05 Nov 2025 00:12:19 GMT</pubDate>
            <atom:updated>2025-11-05T00:12:19.253Z</atom:updated>
            <content:encoded><![CDATA[<h3>Sending AI to the Stars on Solar Power!</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*Vqi3Ndy1kMrUagUk" /><figcaption>Photo by <a href="https://unsplash.com/@obruta?utm_source=medium&amp;utm_medium=referral">Kevin Stadnyk</a> on <a href="https://unsplash.com/?utm_source=medium&amp;utm_medium=referral">Unsplash</a></figcaption></figure><h3>I. Hey Google, What’s Up (in Space)?</h3><p>Imagine a world where the cloud isn’t just a metaphor, but a tangible reality floating above our heads. Data centers, not nestled in sprawling campuses on Earth, but orbiting it, bathed in the perpetual glow of the sun. This isn’t science fiction; it’s the essence of Google’s audacious “Project Suncatcher.” The core idea? Launch custom-designed AI chips — those powerful Tensor Processing Units (TPUs) — into space, powering them with the seemingly limitless energy of the sun. These aren’t solitary satellites; they’re envisioned as interconnected constellations, forming a vast, orbital computing network.</p><p>Why this sudden celestial surge? The driving forces are manifold: a desire to supercharge machine learning capabilities, tap into the unparalleled potential of solar energy in space, and perhaps even alleviate some of the strain on Earth’s increasingly burdened energy grid. The promise is tantalizing — space-based data centers that, one day, might rival the cost-effectiveness of their terrestrial counterparts. It sounds like something plucked from the pages of a speculative novel, doesn’t it? But this is the direction in which Google is now undeniably heading.</p><h3>II. A Quick Trip Down Memory Lane: AI’s History in Orbit</h3><p>We often perceive technological breakthroughs as sudden leaps, forgetting the years of incremental progress that paved the way. AI in space is no exception. This isn’t some unprecedented, out-of-the-blue endeavor. AI has been a quiet, almost invisible hand, assisting our cosmic explorations for decades.</p><p>Consider the early days, the ’70s and ’80s, where AI algorithms were employed to analyze the vast streams of data pouring in from satellites like Landsat. These early applications laid the groundwork for more sophisticated systems. By 1997, NASA’s Mars Pathfinder rover was navigating the Martian terrain with a degree of autonomy, a testament to the growing capabilities of AI in extraterrestrial environments.</p><p>The late ’90s marked a pivotal moment. NASA’s Deep Space 1 mission entrusted its very operation to an AI system. Imagine that — a spacecraft charting its course, making critical decisions, guided by algorithms. Fast forward to the 2000s and beyond, and we see a proliferation of AI applications in space, spearheaded by various international space agencies. The European Space Agency’s PhiSat-1 used AI to filter out cloudy images, saving valuable bandwidth. JAXA, the Japanese space agency, deployed AI for self-checks on its Epsilon rocket and even created “Int-ball,” a robotic assistant for astronauts on the International Space Station. Germany’s DLR developed CIMON, an AI-powered astronaut assistant. And now, China has upped the ante considerably, launching its “Three-Body Computing Constellation” in May 2025 — a fully realized space-based supercomputer network! The stage, clearly, has been set.</p><h3>III. The Buzz from the Experts: Why Space is the Next AI Frontier</h3><p>The excitement surrounding Google’s Project Suncatcher isn’t just internal. Industry experts are increasingly vocal about the potential of space as the next frontier for AI development and deployment. Google executives have boldly declared space as “the best place to scale AI compute,” a sentiment echoed by many within the tech community.</p><p>The advantages are compelling. Firstly, there’s the sheer abundance of solar energy. In space, solar panels can operate with near-constant exposure to sunlight, potentially yielding up to eight times the energy production compared to their terrestrial counterparts. Then there’s the issue of speed and efficiency. By processing data in orbit, AI systems can circumvent the bottlenecks associated with transmitting vast amounts of information back to Earth, enabling real-time analysis and decision-making. Furthermore, AI promises to significantly enhance the capabilities of satellites themselves, improving navigation, collision avoidance, and overall autonomous operation.</p><p>And Google is far from alone in this endeavor. The allure of space-based data centers has attracted the attention of industry giants like SpaceX (Elon Musk), Amazon (Jeff Bezos), and even ambitious startups like Starcloud, backed by Nvidia. The race to conquer the cosmic cloud is officially on, and the skies are becoming increasingly crowded.</p><h3>IV. Hold On a Sec! The Cosmic Catch-22s &amp; Controversies</h3><p>But before we get carried away with visions of AI-powered utopias in orbit, it’s crucial to acknowledge the significant challenges and potential pitfalls that lie ahead. Space, as they say, is a harsh mistress.</p><p>One of the most pressing concerns is the impact of space radiation on sensitive electronic equipment. While Google claims its TPUs are surprisingly resilient, radiation poses a constant threat to the reliability and longevity of space-based AI systems. Another major hurdle is heat management. In the vacuum of space, there’s no air to dissipate heat, making it incredibly difficult to cool the powerful AI chips required for complex computations. This necessitates innovative and potentially costly engineering solutions.</p><p>Then there’s the environmental elephant in the room: the irony of launching environmentally friendly technology using rockets that spew out significant amounts of CO2. While proponents argue that the long-term energy savings from space-based solar power will eventually offset the initial launch emissions, the immediate environmental impact cannot be ignored.</p><p>The increasing number of satellites in orbit is also raising concerns about space debris. What happens when these AI data centers reach the end of their lifespan or malfunction? They risk becoming additional pieces of space junk, further exacerbating an already critical problem. Astronomers, too, are voicing their concerns about the growing number of satellite constellations interfering with their ability to observe the universe, likening them to “bugs on a windshield.” Paradoxically, AI is also being developed to track and mitigate space debris.</p><p>Finally, we come to the ethical quandaries. Who is ultimately responsible when an AI system makes a critical decision in space? What happens when humans can’t intervene quickly enough? How do we ensure transparency and accountability in these autonomous systems? And perhaps most importantly, how do we prevent ourselves from becoming overly reliant on AI, potentially eroding our own problem-solving capabilities? Currently, there is a lack of clear international ethical guidelines for AI in space.</p><p>Let’s not forget the cybersecurity risks. A successful cyberattack on a space-based AI system could have catastrophic consequences, potentially leading to data theft, compromised commands, or even complete satellite takeovers.</p><h3>V. Beyond the Horizon: What’s Next for AI in Space?</h3><p>Despite these challenges, the momentum behind space-based AI continues to build. Google’s Project Suncatcher envisions “gigawatt-scale” AI constellations, representing a truly massive computing network in orbit.</p><p>Beyond pure computing power, AI is poised to revolutionize our understanding and management of our own planet. Google’s AlphaEarth Foundations model, acting as a “virtual satellite,” is capable of generating incredibly detailed, real-time maps of Earth, providing invaluable data for climate monitoring, disaster response, and resource management.</p><p>AI will also play an increasingly critical role in enhancing the safety and efficiency of space missions, improving satellite navigation and collision avoidance, and even providing a “digital doctor” to support astronauts on long-duration missions. AI is even being employed to forecast geomagnetic storms, providing crucial warnings to protect vital infrastructure on Earth.</p><p>The rapid pace of innovation in AI and satellite technology is also enabling the development of modular satellite designs, allowing for easier hardware upgrades and ensuring that space-based AI systems can keep pace with the ever-evolving capabilities of artificial intelligence.</p><h3>VI. The Final Frontier, Powered by AI?</h3><p>Google’s foray into space-based AI represents a bold and ambitious leap, building upon decades of innovation while simultaneously grappling with complex ethical, environmental, and technological challenges.</p><p>It’s a testament to humanity’s relentless pursuit of new frontiers, a desire to push the boundaries of what’s possible. But will AI truly be the key that unlocks the cosmos? Only time — and a whole lot of solar power — will tell. The questions are profound, the stakes are high, and the journey promises to be nothing short of transformative.</p><p><a href="https://a.co/d/4ES9r2k">AI Factory: Architecting Global Supremacy</a></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=89570b44a77a" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Why “Forward-Deployed Engineers” Are the Hottest Job in Tech Right Now]]></title>
            <link>https://medium.com/@sombando/why-forward-deployed-engineers-are-the-hottest-job-in-tech-right-now-d0a34678f32a?source=rss-2d03c339ebc4------2</link>
            <guid isPermaLink="false">https://medium.com/p/d0a34678f32a</guid>
            <category><![CDATA[ai]]></category>
            <category><![CDATA[software-development]]></category>
            <category><![CDATA[software-engineering]]></category>
            <category><![CDATA[technology]]></category>
            <category><![CDATA[digital-transformation]]></category>
            <dc:creator><![CDATA[Som Bando]]></dc:creator>
            <pubDate>Mon, 03 Nov 2025 01:01:52 GMT</pubDate>
            <atom:updated>2025-11-03T01:01:52.780Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*lo-X3zsNe8XrRFj4" /><figcaption>Photo by <a href="https://unsplash.com/@sigmund?utm_source=medium&amp;utm_medium=referral">Sigmund</a> on <a href="https://unsplash.com/?utm_source=medium&amp;utm_medium=referral">Unsplash</a></figcaption></figure><h3>Intro: Meet the AI Bridge Builders</h3><p>Forget, for a moment, the well-trodden path of the “full-stack developer.” A new archetype has emerged, a figure imbued with a unique blend of technical prowess and practical acumen: the Forward-Deployed Engineer (FDE) specializing in AI.</p><p>They aren’t mere coders tethered to the familiar glow of their IDEs. No, these are the special forces of the AI world, dispatched into the heart of businesses to wrestle cutting-edge AI into a functional, value-generating reality.</p><p><strong>Think of the titans of AI — OpenAI, Anthropic, Cohere. They are all engaged in a fervent pursuit of these “AI whisperers,” individuals capable of translating their powerful, yet often bewilderingly complex, models into tangible solutions for enterprises eager to embrace the AI revolution.</strong></p><h3>Chapter 1: So, What Exactly Is an AI FDE?</h3><p>Consider this: AI models, in their raw, untamed form, are akin to magnificent architectural blueprints. They hold immense potential, yet remain unrealized until skillfully adapted to the specific terrain, infrastructure, and aesthetic sensibilities of a particular client.</p><p>This is where the FDE enters the scene. They are hybrid heroes, embodying a potent combination of deep technical skills, business intelligence, and an innate ability to connect with people. They are the vital conduit between the rarefied atmosphere of AI research labs and the pragmatic demands of the real world.</p><p>Their mission? To embed themselves within client organizations, to intimately understand their operational intricacies, their data landscapes, and their cultural nuances. And then, armed with this understanding, to build, customize, integrate, and deploy AI solutions from the ground up, pushing them from the realm of abstract possibility into the tangible reality of production.</p><h3>Chapter 2: A Brief History of “Forward-Deployed” Badassery</h3><p>The concept of the FDE, while enjoying a surge in popularity today, is not entirely novel. Its origins can be traced back to the early 2000s, to Palantir Technologies, a company renowned for tackling some of the world’s most complex data challenges. They called their FDEs “Deltas,” a nod, perhaps, to their role as agents of transformation.</p><p>Palantir recognized that deploying sophisticated data platforms for entities like governments and intelligence agencies demanded more than just off-the-shelf solutions. It required engineers on the front lines, working shoulder-to-shoulder with clients, crafting bespoke solutions to address “world-changing problems,” such as combating human trafficking.</p><p>For years, this approach remained a niche strategy. However, as software systems grew in complexity, companies like Scale AI, C3.ai, and Databricks began to appreciate the inherent value of embedding highly skilled engineers to ensure the successful, customized deployment of their offerings.</p><p>The advent of generative AI and Large Language Models (LLMs) has ignited an unprecedented surge in demand for FDEs. The challenge is no longer limited to data platforms; it’s about translating theoretical AI breakthroughs into demonstrable business value, and doing so with remarkable speed.</p><h3>Chapter 3: What Do These AI Superheroes Actually Do All Day? (Responsibilities &amp; Skills)</h3><p>The AI FDE’s day is a tapestry woven from diverse threads of technical expertise, strategic thinking, and interpersonal finesse.</p><p>They shepherd projects from nascent prototypes to fully realized production systems, constructing full-stack solutions that deliver tangible customer value, transcending the realm of mere “cool demos.”</p><p>They immerse themselves within client teams, becoming temporary members, delving into workflows, dissecting policies, and identifying pain points with almost surgical precision.</p><p>They are AI tailors, meticulously customizing AI models — through prompt engineering, fine-tuning LLMs, and designing sophisticated data warehouses — to align perfectly with a company’s unique requirements.</p><p>Furthermore, they serve as strategic guides, advising clients on how to effectively leverage their newfound AI capabilities, often architecting entirely new AI systems tailored to specific needs.</p><p>They are problem-solvers extraordinaire, proactively identifying obstacles, removing blockers, and iteratively refining solutions in real-time, often coding alongside end-users to ensure seamless integration.</p><p>Crucially, they act as a vital feedback loop, channeling insights gleaned from the front lines back to their product teams, influencing the future trajectory of AI development.</p><p>To excel in this multifaceted role, a diverse skill set is essential:</p><ul><li><strong>Deep Tech Smarts:</strong> Proficiency in Python, JavaScript, cloud platforms, data engineering, full-stack development, and a thorough understanding of LLMs are foundational.</li><li><strong>Business Brains:</strong> The ability to connect AI capabilities to tangible business challenges and quantify the return on investment (ROI) is paramount.</li><li><strong>People Skills:</strong> Communicating intricate technical concepts to non-technical audiences, fostering trust, and demonstrating empathy are crucial for building rapport and driving adoption.</li><li><strong>Grit &amp; Adaptability:</strong> Comfort navigating ambiguity, an insatiable appetite for learning, and the ability to devise solutions where established playbooks are absent are essential traits.</li><li><strong>Project Whisperers:</strong> Managing timelines, balancing trade-offs, and ensuring successful project delivery under pressure are critical for achieving desired outcomes.</li></ul><h3>Chapter 4: The Big Players Are All In (OpenAI, Anthropic, Cohere)</h3><p>The leading AI organizations recognize the pivotal role of FDEs in driving widespread adoption and realizing commercial success.</p><p>OpenAI, for example, is focused on converting its groundbreaking research, such as GPT-4o, into production-ready systems for its strategic clientele. The company is aggressively expanding its FDE team, anticipating that a significant portion of its enterprise revenue will be generated through FDE-led initiatives.</p><p>Anthropic is dedicated to crafting bespoke LLM solutions for leading enterprises, placing a strong emphasis on aligning AI with human values, safety, and reliability.</p><p>Cohere adopts an “enterprise-first” approach, striving to make AI practical and accessible. They collaborate with consulting giants like McKinsey and Accenture, focusing on privacy and customization.</p><p>These companies understand that powerful AI requires expert, hands-on implementation to fuel adoption, enhance customer retention, and establish a sustainable competitive advantage — a “services plus AI” model.</p><h3>Chapter 5: The Dark Side of Deployment: Controversies &amp; Ethical Minefields</h3><p>The deployment of AI systems is not without its challenges. FDEs are often confronted with ethical dilemmas and potential pitfalls.</p><p>Data privacy is a paramount concern. FDEs handle sensitive client data on a daily basis, raising concerns about unauthorized use, breaches, misuse, and compliance with evolving regulations like GDPR. The inherent opacity of some AI models exacerbates the challenge of maintaining transparency.</p><p>The “black box” nature of certain AI algorithms poses a significant hurdle. How can one adequately explain an AI’s decision-making process when even the engineers themselves struggle to fully comprehend its logic? FDEs grapple with the challenge of making these opaque systems more transparent to foster trust and accountability.</p><p>AI systems can inadvertently perpetuate and amplify human biases present in their training data, leading to unfair or discriminatory outcomes. FDEs find themselves on the front lines, tasked with identifying and mitigating these biases.</p><p>Critics contend that the FDE model can foster client dependency on the vendor’s engineers for ongoing operations, potentially masking underlying product flaws and resulting in sustained high costs. Is this an inherent design flaw or a deliberate feature?</p><p>While the FDE role represents a new avenue for employment, the AI systems they deploy have the potential to automate tasks, raising concerns about job displacement and the erosion of human interaction.</p><p>Finally, the demanding nature of the FDE role, characterized by intense technical work, client-facing pressures, and frequent travel, can lead to burnout.</p><h3>Chapter 6: The Future Is Forward-Deployed (What’s Next for AI FDEs)</h3><p>The demand for FDEs is experiencing exponential growth. Job postings for these specialized engineers have surged dramatically. This is not merely a passing fad; it is the emerging paradigm for enterprise AI adoption.</p><p>As AI evolves towards greater autonomy, becoming more “agentic” — capable of independent action — FDEs will play an even more crucial role in orchestrating these intricate, goal-driven systems within business contexts.</p><p>We can anticipate the emergence of even more specialized FDE roles, such as “Agent Deployment Engineers,” focused exclusively on deploying and managing AI-driven agents and designing advanced Retrieval-Augmented Generation (RAG) systems.</p><p>FDEs will increasingly leverage AI-native platforms, collaborating closely with domain experts to rapidly develop tailored applications.</p><p>While coding will remain an essential skill, the emphasis will shift towards orchestrating comprehensive solutions, bridging the gap between generic AI tools and specific business imperatives. The value proposition will evolve from code creation to a deep understanding of complex problems and the ability to solve them effectively.</p><p>The FDE role offers a significant career boost, often accompanied by higher compensation (20–40% above traditional software engineering roles) and accelerated promotion cycles, reflecting their direct impact and high visibility within organizations. It can serve as a springboard to roles in product management, engineering leadership, or even the founding of one’s own startup.</p><h3>Conclusion: The Unsung Heroes of the AI Revolution</h3><p>AI FDEs are the indispensable guides, navigating businesses through the intricate labyrinth of AI adoption.</p><p>They provide the crucial human element, ensuring that powerful AI systems are not merely theoretical marvels but practical, ethical, and transformative forces capable of generating real-world impact.</p><p>In the relentless race for AI dominance, these bridge-builders are not just desirable assets; they are essential for companies seeking to transcend experimentation and unlock the full potential of artificial intelligence.</p><p><strong>get the book</strong></p><p><a href="https://www.amazon.com/dp/B0FVXTSFZW">The AI Transformation Playbook: From Vision to Value: The Proven Framework for Leaders to Build, Scale, and Govern AI</a></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=d0a34678f32a" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[H-1B Visa Shockwave: $100,000 Annual Fee]]></title>
            <link>https://medium.com/@sombando/h-1b-visa-shockwave-100-000-annual-fee-8b19b4d41e42?source=rss-2d03c339ebc4------2</link>
            <guid isPermaLink="false">https://medium.com/p/8b19b4d41e42</guid>
            <dc:creator><![CDATA[Som Bando]]></dc:creator>
            <pubDate>Wed, 24 Sep 2025 14:12:33 GMT</pubDate>
            <atom:updated>2025-09-24T14:20:37.248Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*tvo9Am_hJ_IGkRihuNHXxA.png" /></figure><p>The U.S. skilled worker visa system has just undergone a dramatic, disruptive overhaul. President Donald Trump signed a proclamation on September 19, 2025, introducing a $100,000 annual fee on companies sponsoring H-1B visa holders. This increase is substantial, moving from previous fees that were as low as $1,000 or $215-$780 for petition filings.</p><p>Here are the key components of this sweeping policy shift, effective 12:01 a.m. EDT on September 21, 2025.</p><h3>1. The Policy Goal: Prioritizing American Workers</h3><p>The administration framed the fee hike as a necessary step to protect U.S. jobs. Commerce Secretary Howard Lutnick stated that the dramatic increase makes hiring foreign workers “non-economic” for companies. The policy aims to <strong>discourage the replacement of American workers</strong> with foreign labor and prioritize American graduates from U.S. universities.</p><p>The goal is to allow only <strong>“high-value” roles</strong> to qualify for the visa, thereby limiting its use for lower-cost labor. Trump claimed that major corporations were “onboard” with the change, referring to large corporations that he said “love it” and “need it”.</p><h3>2. Immediate Entry Restrictions and Urgent Advisories</h3><p>The proclamation restricts the entry of H-1B nonimmigrant workers into the United States unless the new $$ 100,000$ fee is paid. This measure immediately triggered panic among H-1B visa holders and their dependents (H-4 visa holders) who were traveling abroad.</p><p>In response, major tech giants issued urgent warnings:</p><ul><li><strong>Amazon and Microsoft</strong> advised H-1B and H-4 employees who were outside the U.S. to return immediately, specifically before midnight EDT on September 21, 2025, to avoid the risk of being denied re-entry.</li><li>Companies recommended that employees already in the U.S. should avoid non-critical international travel for the foreseeable future.</li></ul><p>The restrictions primarily affect workers currently outside the U.S. seeking entry on or after the effective date. Extensions of stay or changes of employer for those already lawfully inside the U.S. appear exempt, but leaving and re-entering subjects the worker to the new fee.</p><p><strong>Later, the administration issued clarification — </strong>that no existing holders are impacted and only new applications need to pay this fee.</p><iframe src="https://cdn.embedly.com/widgets/media.html?src=https%3A%2F%2Fwww.youtube.com%2Fembed%2FEZ4PY9BDSi4%3Fstart%3D2%26feature%3Doembed%26start%3D2&amp;display_name=YouTube&amp;url=https%3A%2F%2Fwww.youtube.com%2Fwatch%3Fv%3DEZ4PY9BDSi4&amp;image=https%3A%2F%2Fi.ytimg.com%2Fvi%2FEZ4PY9BDSi4%2Fhqdefault.jpg&amp;type=text%2Fhtml&amp;schema=youtube" width="854" height="480" frameborder="0" scrolling="no"><a href="https://medium.com/media/6bea3e9f5f4fc6597e38805f9b2ecf95/href">https://medium.com/media/6bea3e9f5f4fc6597e38805f9b2ecf95/href</a></iframe><h3>3. Disproportionate Impact on Indian Professionals and Tech Giants</h3><p>The move is expected to hit the tech and IT sectors hard, which have traditionally relied heavily on skilled foreign workers.</p><ul><li><strong>Largest Beneficiaries Affected:</strong> Indian professionals are the largest group affected, accounting for nearly 71% of all approved H-1B visa beneficiaries. India’s foreign ministry warned that the policy could lead to “potential disruptions for families”.</li><li><strong>Top Employers Targeted:</strong> The highest volume H-1B employers face massive new cost burdens. According to 2025 USCIS data, the top recipients include Amazon (10,044 approvals), Tata Consultancy Services (TCS) (5,505), Microsoft (5,189), Meta (5,123), and Apple (4,202).</li></ul><h3>4. Long-Term Outlook and Expected Shifts</h3><p>Experts and industry veterans believe the steep annual fee makes the H-1B program <strong>“completely unviable”</strong> for junior or mid-level positions.</p><ul><li><strong>Legal Challenges:</strong> Legal experts anticipate legal challenges to the proclamation, arguing that U.S. law typically restricts visa fees to costs of processing and enforcement, not as a deterrent to usage.</li><li><strong>Offshoring and GCCs:</strong> The prohibitive costs are expected to accelerate existing trends, forcing companies to reduce reliance on foreign workers and increase offshoring of projects to countries like India. This shift could lead to more employment opportunities within India through the expansion of Global Capability Centers (GCCs) set up by American companies.</li><li><strong>Impact on Talent Pipeline:</strong> The policy may also discourage Indian students from pursuing further studies in American universities, as the likelihood of staying back and securing employment is reduced.</li></ul><p>The policy marks a significant shift, creating short-term hardship but potentially leading to structural changes in global talent acquisition for the tech industry.</p><p><a href="https://a.co/d/iUgVI1d?source=post_page-----2d376ac1d56a---------------------------------------">AI Factory: Architecting Global Supremacy</a></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=8b19b4d41e42" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Myth Busted: Why the Viral “95% AI Failure” Stat Is Misleading]]></title>
            <link>https://medium.com/@sombando/myth-busted-why-the-viral-95-ai-failure-stat-is-misleading-784c7224024c?source=rss-2d03c339ebc4------2</link>
            <guid isPermaLink="false">https://medium.com/p/784c7224024c</guid>
            <category><![CDATA[ai-factory]]></category>
            <category><![CDATA[mit]]></category>
            <category><![CDATA[ai-agent]]></category>
            <category><![CDATA[ai]]></category>
            <category><![CDATA[technology]]></category>
            <dc:creator><![CDATA[Som Bando]]></dc:creator>
            <pubDate>Wed, 24 Sep 2025 14:02:21 GMT</pubDate>
            <atom:updated>2025-09-24T14:19:11.056Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*lHN3DRFWBz1jSXfbkMVcNg.png" /></figure><p>The widely circulated claim that 95% of artificial intelligence projects fail, which allegedly originated from a recent MIT paper, is a significant misrepresentation of a nuanced business report. This sensationalized headline, which resonated by confirming fears of an “AI bubble,” collapses under critical scrutiny. The real story of enterprise AI is far more complex and promising.</p><h3>The Flaw in the 95% Figure</h3><p>The figure originates from the report <em>The GenAI Divide: State of AI in Business 2025</em>. It is important to note that the source is a business-focused “report,” not a peer-reviewed “academic paper”.</p><p>The 95% “failure” rate is based on a highly specific and narrow definition of success: <strong>“deployment beyond pilot phase with measurable KPIs”</strong> and an <strong>“ROI impact measured six month post pilot”</strong>. This definition systematically excludes critical, long-term indicators of value, such as operational efficiency gains, cost reductions, and the vital value of strategic learning. The source data was also based on a small sample of only 52 interviews, which the report itself qualified as only “directionally accurate”.</p><h3>The True State of Enterprise AI: The Learning Advantage</h3><p>More comprehensive research paints a radically different picture. Other MIT research finds that organizations that combine organizational learning with AI-specific learning, known as “Augmented Learners,” are 1.4 times more likely to recognize revenue benefits from AI. An astonishing <strong>99% of these organizations report observing some form of financial benefit</strong> from their AI initiatives.</p><p>The core challenge identified across the reports is not technological inadequacy, but an organizational <strong>“learning gap”</strong>. The “GenAI Divide” is not between those who adopt AI and those who fail, but between those who strategically integrate AI for systemic learning and those who deploy it tactically.</p><iframe src="https://cdn.embedly.com/widgets/media.html?src=https%3A%2F%2Fwww.youtube.com%2Fembed%2FohHx50JLDJY%3Ffeature%3Doembed&amp;display_name=YouTube&amp;url=https%3A%2F%2Fwww.youtube.com%2Fwatch%3Fv%3DohHx50JLDJY&amp;image=https%3A%2F%2Fi.ytimg.com%2Fvi%2FohHx50JLDJY%2Fhqdefault.jpg&amp;type=text%2Fhtml&amp;schema=youtube" width="854" height="480" frameborder="0" scrolling="no"><a href="https://medium.com/media/c5af746a734991f9e823a2af2064e2fa/href">https://medium.com/media/c5af746a734991f9e823a2af2064e2fa/href</a></iframe><h3>Strategic Takeaways for Success</h3><p>The path to success requires a fundamental shift in strategy. Business leaders should focus on:</p><ol><li><strong>Broadening Metrics:</strong> Move beyond a narrow six-month ROI window to a multi-metric framework that includes operational efficiency, cost reduction, and strategic benefits.</li><li><strong>Prioritizing Organizational Learning:</strong> Success is directly tied to overcoming the “learning gap” by fostering a culture of continuous learning and adaptation, as deploying AI is an ongoing process of change management.</li><li><strong>Strategic Allocation:</strong> While sales and marketing often receive early AI budgets, reports suggest higher returns can be found in automating back-office functions like HR and customer service through efficiency gains.</li></ol><p>The future belongs not to the fastest adopter but to the most thoughtful learner. The 95% figure is merely a sensationalized metric representing the consequences of a lack of organizational learning.</p><p><a href="https://a.co/d/iUgVI1d?source=post_page-----2d376ac1d56a---------------------------------------">AI Factory: Architecting Global Supremacy</a></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=784c7224024c" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[The Dawn of the AI-Driven Gigawatt Era]]></title>
            <link>https://medium.com/@sombando/the-dawn-of-the-ai-driven-gigawatt-era-2d376ac1d56a?source=rss-2d03c339ebc4------2</link>
            <guid isPermaLink="false">https://medium.com/p/2d376ac1d56a</guid>
            <category><![CDATA[energy]]></category>
            <category><![CDATA[ai]]></category>
            <category><![CDATA[technology]]></category>
            <category><![CDATA[data-center]]></category>
            <dc:creator><![CDATA[Som Bando]]></dc:creator>
            <pubDate>Wed, 30 Jul 2025 16:36:38 GMT</pubDate>
            <atom:updated>2025-07-30T16:40:22.722Z</atom:updated>
            <content:encoded><![CDATA[<h3>A Data Center Market in Transformation</h3><figure><img alt="In Abilene, Lancium’s 800-acre Clean Campus is leasing about 90 acres to data center developer Crusoe. (Photo: Development Corporation of Abilene)" src="https://cdn-images-1.medium.com/max/642/0*GOJ-d9rrBerCjccK.png" /><figcaption>In Abilene, Lancium’s 800-acre Clean Campus is leasing about 90 acres to data center developer Crusoe. (Photo: Development Corporation of Abilene)</figcaption></figure><p>The global data center industry is in the midst of its most profound transformation since the advent of the cloud. Driven by the unprecedented computational demands of artificial intelligence (AI), the market is undergoing a fundamental recalibration of scale, shifting from the megawatt to the gigawatt era. This surge is creating a global race for capacity, where development is increasingly governed and constrained by what can be termed the “three Ps”: Power, Permitting, and Public Perception. The sheer velocity of this expansion is reshaping global supply chains, energy markets, and even national industrial policies.</p><p>The digital infrastructure that underpins the global economy is vast and expanding at a historic pace. The first half of 2023 witnessed the largest volume of data center construction in history, a trend that shows no signs of abating as demand continues to outpace new supply across both core and emerging hubs [1. 2].</p><p>The global data center construction market, valued at approximately $251 billion in 2025, is projected to exceed $613 billion by 2037, reflecting a compound annual growth rate (CAGR) of over 7.6% [3].</p><p>This growth is tangible in the sheer number of facilities worldwide. However, quantifying the exact number of data centers reveals the market’s complex and fragmented nature. Estimates vary significantly depending on the scope and methodology of the source. For instance, user-driven and colocation-focused directories list between 6,725 and 10,353 facilities globally [4, 5].</p><p>In contrast, comprehensive market intelligence databases that include hyperscale, leased, enterprise, and cryptocurrency mining facilities track over 12,900 individual sites [6].</p><p>This discrepancy is not a data flaw but a clear indicator of the market’s diverse structure, which ranges from localized edge computing sites to sprawling, multi-building hyperscale campuses. A simple directory is therefore insufficient for strategic analysis; the market must be understood through a granular segmentation of its components, operators, and end-users.</p><p>This discrepancy is not a data flaw but a clear indicator of the market’s diverse structure, which ranges from localized edge computing sites to sprawling, multi-building hyperscale campuses. A simple directory is therefore insufficient for strategic analysis; the market must be understood through a granular segmentation of its components, operators, and end-users.</p><h3>Defining the Modern Data Center: From Colocation to AI Factories</h3><p>To navigate the complexity of the global market, a clear taxonomy of data center types is essential. The term “data center” encompasses a wide spectrum of facilities, each designed for specific use cases, scales, and operational models.</p><ul><li><strong>Enterprise Data Centers:</strong> These are traditional, on-premises facilities owned and operated by a single organization for its internal IT needs. They include server rooms, corporate data centers, and private cloud environments. While still a significant part of the IT landscape, the trend has been a gradual migration from owned enterprise facilities to more flexible colocation and cloud models to reduce capital expenditures and improve scalability [7].</li><li><strong>Colocation (Retail &amp; Wholesale):</strong> Colocation facilities are multi-tenant environments where businesses can rent space, power, cooling, and connectivity for their IT hardware. This model is divided into two primary categories. <strong>Retail colocation</strong> involves leasing smaller footprints, such as individual server cabinets or secure cages, often bundled with managed services [8].</li><li><strong>Wholesale colocation</strong> involves leasing larger, dedicated spaces, such as private suites or entire data halls, typically with longer-term contracts where the tenant manages its own IT stack and pays for power consumption directly [7,8].</li><li><strong>Hyperscale (Cloud &amp; Social Media):</strong> These are the largest data centers, often consisting of entire campuses with multiple buildings and hundreds of megawatts of power capacity. They are purpose-built to support the massive, distributed infrastructure of the world’s largest technology companies, primarily the major cloud service providers — Amazon Web Services (AWS), Microsoft Azure, and Google Cloud Platform (GCP) — and large-scale internet companies like Meta Platforms [1,8]. These facilities are characterized by extreme operational efficiency, custom hardware, and economies of scale that are unattainable in smaller deployments [9].</li><li><strong>Edge Computing Centers:</strong> In contrast to centralized hyperscale facilities, edge data centers are smaller, more numerous, and geographically distributed. They are located closer to end-users and devices — at the “edge” of the network — to reduce latency and process data locally. This model is critical for emerging applications such as the Internet of Things (IoT), autonomous vehicles, real-time analytics, and augmented reality, which cannot tolerate the delay of sending data to a distant cloud data center [3].</li></ul><p><strong>The Emergence of “AI Factories”:</strong> The latest evolution in data center design is the “AI Factory.” These are not merely large data centers but facilities engineered from the ground up to handle the extreme power and cooling densities required by AI workloads, particularly the training of large language models (LLMs) [10, 11]. Key characteristics include support for rack power densities that can exceed 70–100 kW (compared to 10–15 kW in traditional facilities), the widespread use of advanced liquid cooling technologies, and electrical infrastructure designed to support the power-hungry demands of next-generation GPUs from manufacturers like NVIDIA [12].</p><h3>Read the book</h3><p><a href="https://a.co/d/iUgVI1d">AI Factory: Architecting Global Supremacy</a></p><h3>References</h3><ol><li>215 Data Center Stats (June-2025) — Brightlio, accessed July 30, 2025, <a href="https://brightlio.com/data-center-stats/">https://brightlio.com/data-center-stats/</a></li><li>Global Data Center Trends 2025 | CBRE, accessed July 30, 2025, <a href="https://www.cbre.com/insights/reports/global-data-center-trends-2025">https://www.cbre.com/insights/reports/global-data-center-trends-2025</a></li><li>Data Center Construction Market Size 2025 | $613.91 billion Global Estimates by 2037, accessed July 30, 2025, <a href="https://www.researchnester.com/reports/data-center-construction-market/6126">https://www.researchnester.com/reports/data-center-construction-market/6126</a></li><li>Search Data Centers: Colocation, Bare Metal, Cloud, accessed July 30, 2025, <a href="https://www.datacenters.com/">https://www.datacenters.com/</a></li><li>Database — Data Centers, accessed July 30, 2025, <a href="https://www.datacentermap.com/datacenters/">https://www.datacentermap.com/datacenters/</a></li><li>451 Research Data Center Knowledge Base Dataset — S&amp;P Global Marketplace, accessed July 30, 2025, <a href="https://www.marketplace.spglobal.com/en/datasets/451-research-data-center-knowledge-base-(238)">https://www.marketplace.spglobal.com/en/datasets/451-research-data-center-knowledge-base-(238)</a></li><li>The cost of compute: A $7 trillion race to scale data centers — McKinsey, accessed July 30, 2025, <a href="https://www.mckinsey.com/industries/technology-media-and-telecommunications/our-insights/the-cost-of-compute-a-7-trillion-dollar-race-to-scale-data-centers">https://www.mckinsey.com/industries/technology-media-and-telecommunications/our-insights/the-cost-of-compute-a-7-trillion-dollar-race-to-scale-data-centers</a></li><li>Top 250 Data Center Companies in the World as of 2024 — Dgtl Infra, accessed July 30, 2025, <a href="https://dgtlinfra.com/top-data-center-companies/">https://dgtlinfra.com/top-data-center-companies/</a></li><li>Decoding Data Centers: Opportunities, risks and investment strategies, accessed July 30, 2025, <a href="https://www.cbreim.com/insights/articles/decoding-data-centers">https://www.cbreim.com/insights/articles/decoding-data-centers</a></li><li>What is a Data Center? — YouTube, accessed July 30, 2025, <a href="https://www.youtube.com/watch?v=Amow8BJm5Go">https://www.youtube.com/watch?v=Amow8BJm5Go</a></li><li>Data Centre Sustainability: How Telecom Operators are Driving Change — STL Partners, accessed July 30, 2025, <a href="https://stlpartners.com/articles/sustainability/data-centre-sustainability/">https://stlpartners.com/articles/sustainability/data-centre-sustainability/</a></li><li>Data Center Map — Colocation, Cloud and Connectivity, accessed July 30, 2025, <a href="https://www.datacentermap.com/">https://www.datacentermap.com/</a></li></ol><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=2d376ac1d56a" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[The AI Revolution Is Here]]></title>
            <link>https://medium.com/techemerge/the-ai-revolution-is-here-480a40cdc029?source=rss-2d03c339ebc4------2</link>
            <guid isPermaLink="false">https://medium.com/p/480a40cdc029</guid>
            <category><![CDATA[artificial-intelligence]]></category>
            <category><![CDATA[technology]]></category>
            <dc:creator><![CDATA[Som Bando]]></dc:creator>
            <pubDate>Tue, 29 Jul 2025 19:16:14 GMT</pubDate>
            <atom:updated>2025-07-29T19:16:14.156Z</atom:updated>
            <content:encoded><![CDATA[<h3>Did You Miss the Starting Gun?</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*fBCmotP9Q3XZLaVfK1IJVA.png" /><figcaption>Imagining size of Hyperion Data Center over Manhattan.</figcaption></figure><h3>The New Arms Race: GPUs and Gigawatts</h3><p>Forget traditional metrics of power; the new global currency is the GPU, with some staggering numbers that paint a picture of an unprecedented technological arms race:</p><ul><li><strong>1 Million GPUs:</strong> OpenAI aims to have over <strong>one million GPUs</strong> online by the end of this year alone.</li><li><strong>50 Million H100 Equivalents:</strong> Elon Musk’s XAI has set a five-year target to acquire a staggering <strong>50 million H100-equivalent GPUs</strong>.</li><li><strong>Multi-Gigawatt Data Centers:</strong> Meta is building massive, multi-gigawatt data centers named “Prometheus” and “Hyperion,” using innovative construction methods like hurricane-proof tents to get them online faster.</li></ul><p>This isn’t just about building bigger computers; it’s about building an entirely new infrastructure for intelligence. This aggressive push is being mirrored at the national level, with <strong>America’s AI Plan</strong> aiming to turn the entire country into a massive AI factory through a series of over 90 federal policies designed to fast-track everything from data centers to worker training.</p><h3>Is This the End of Search as We Know It?</h3><p>One of the most thought-provoking discussions centered on the future of Google Search and the web browser. While Google’s AI-powered search overviews are driving record revenue, they are also fundamentally changing our relationship with information.</p><p>This disruption is set to accelerate with <strong>OpenAI’s planned web browser</strong>, a direct challenger to Google’s Chrome empire. The experts on the panel went even further, suggesting that the very idea of a “browser” is on its way to becoming obsolete. In its place, we may see AI-driven “portals” or voice-controlled agents that bypass the traditional web to perform tasks, make purchases, and deliver information directly.</p><h3>The Human Element: Million-Dollar Salaries and AI Companions</h3><p>The demand for human talent to build this future is astronomical. The panel highlighted that top-tier AI experts are commanding salaries in the range of <strong>$10 million to $100 million per year</strong>, signaling a new era of what one expert called “post-scarce individual economics.”</p><p>At the same time, AI’s impact on our daily lives is becoming more personal and intimate. The growing trend of teens using AI companions raises important questions about mental health, socialization, and what it means to form relationships in a world where some of our friends are no longer human.</p><h3>A New Reality, Powered by AI</h3><p>From AI models winning gold medals at the International Math Olympics to Google’s own AI writing <strong>50% of its code</strong>, the message is clear: the capabilities we once thought were years away are already here. This revolution is powered by an insatiable demand for energy, pushing innovations in both solar and fusion power, and is even changing our cities with the rise of robo-taxis that are essentially “mobile data centers on wheels.”</p><p>The future isn’t just coming; it’s being built right now in data centers, in government policies, and in the code that is increasingly being written not by humans, but by the intelligence we created. The question for all of us is no longer <em>if</em> AI will change the world, but how we will adapt to the world it’s already creating.</p><p>Read my book on how a new generation of AI Factories are being built to lead in the age of intelligence.</p><p><a href="https://a.co/d/7GskB91">AI Factory: Architecting Global Supremacy</a></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=480a40cdc029" width="1" height="1" alt=""><hr><p><a href="https://medium.com/techemerge/the-ai-revolution-is-here-480a40cdc029">The AI Revolution Is Here</a> was originally published in <a href="https://medium.com/techemerge">TechEmerge</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[The End of Order]]></title>
            <link>https://medium.com/@sombando/the-end-of-order-844969f2c2a1?source=rss-2d03c339ebc4------2</link>
            <guid isPermaLink="false">https://medium.com/p/844969f2c2a1</guid>
            <category><![CDATA[trade]]></category>
            <category><![CDATA[geopolitics]]></category>
            <category><![CDATA[supply-chain]]></category>
            <category><![CDATA[tarriffs]]></category>
            <category><![CDATA[technology]]></category>
            <dc:creator><![CDATA[Som Bando]]></dc:creator>
            <pubDate>Mon, 21 Jul 2025 21:37:29 GMT</pubDate>
            <atom:updated>2025-07-30T21:56:27.228Z</atom:updated>
            <content:encoded><![CDATA[<h3>Navigating the New Age of Economic Chaos</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*JhfefwnUjHwzvRIN" /><figcaption>Photo by <a href="https://unsplash.com/@hdbernd?utm_source=medium&amp;utm_medium=referral">Bernd 📷 Dittrich</a> on <a href="https://unsplash.com/?utm_source=medium&amp;utm_medium=referral">Unsplash</a></figcaption></figure><p>For three-quarters of a century, the world operated on a predictable foundation of liberalized trade and international cooperation. That era is over. We have entered a volatile new age defined by the return of an old weapon: the tariff. No longer just a tool for protecting fledgling industries, tariffs have been reforged into a primary instrument of modern statecraft, used to project power, coerce allies, and punish rivals. The result is a swift and disorienting shift from a rules-based system to a power-based one, a fracturing of the global order that is leaving leaders and citizens scrambling to understand the new terrain.</p><p><strong>Listen to a summary</strong></p><iframe src="https://cdn.embedly.com/widgets/media.html?src=https%3A%2F%2Fw.soundcloud.com%2Fplayer%2F%3Furl%3Dhttps%253A%252F%252Fapi.soundcloud.com%252Ftracks%252F2134182156%26show_artwork%3Dtrue&amp;display_name=SoundCloud&amp;url=https%3A%2F%2Fsoundcloud.com%2Fsom-bando%2Fthe-geopolitics-of-tariffs-and%3Fsi%3D8a27e53f43ab4ce98f71853c7e696082%26utm_source%3Dclipboard%26utm_medium%3Dtext%26utm_campaign%3Dsocial_sharing&amp;image=https%3A%2F%2Fsoundcloud.com%2Fimages%2Ffb_placeholder.png&amp;type=text%2Fhtml&amp;schema=soundcloud" width="800" height="166" frameborder="0" scrolling="no"><a href="https://medium.com/media/1fc17fd8f426e568580b346bd48c8e34/href">https://medium.com/media/1fc17fd8f426e568580b346bd48c8e34/href</a></iframe><p>The resurgence of protectionism is driven by a powerful and politically resonant narrative. Proponents claim tariffs will create a “level playing field,” force “reciprocity” from other nations, and reclaim national sovereignty. This message, as Bando’s work explains, taps into real economic pain and a deep-seated feeling that the global system is rigged.</p><p>However, <em>The Art of Chaos</em> meticulously deconstructs the economic fallacies at the heart of this narrative:</p><ul><li><strong>The “Bad” Trade Deficit:</strong> The book argues that portraying a trade deficit as a national “loss” is a fundamental misunderstanding. A trade deficit is a macroeconomic outcome of a nation’s savings and investment balance, not a scorecard of trade policy success.</li><li><strong>Who Pays the Tariff?</strong> A persistent claim is that foreign countries bear the cost of tariffs. This is demonstrably false. A tariff is a tax paid by the domestic firm importing the goods, a cost that is then passed on to domestic consumers through higher prices or absorbed by the business as lower profits.</li><li><strong>The Myth of Job Creation:</strong> The idea that tariffs “bring back jobs” is based on the “lump of labor” fallacy. While some jobs may be saved in a protected industry, jobs are simultaneously destroyed in other sectors that rely on now-more-expensive imported materials or become targets for retaliatory tariffs. Economic models consistently show the net effect on employment is negative.</li></ul><h4>A World Fractured: The Tangible Consequences</h4><p>The book provides a clear-eyed assessment of the measurable damage caused by this new protectionist wave. The consequences are not theoretical; they are quantified and felt across the globe.</p><ul><li><strong>The Economic Ledger:</strong> Rigorous modeling from institutions like the Yale Budget Lab projects that the tariffs enacted in 2025 will reduce U.S. real GDP growth, leaving the economy permanently smaller by an amount equivalent to an annual loss of $110 billion. These tariffs are unambiguously inflationary, costing the average American household between $2,300 and $2,400 in lost purchasing power in 2025 alone. The policies are also forecast to lead to a net loss of 553,000 payroll jobs.</li><li><strong>A Wedge in the West:</strong> Geopolitically, these policies have acted as a powerful wedge within the Western alliance. Imposing tariffs on close allies like Canada, the European Union, and Japan under the guise of “national security” erodes trust and strains the very foundations of the NATO alliance.</li><li><strong>A Gift to the East:</strong> Paradoxically, U.S. tariff policy has had a galvanizing effect on the BRICS bloc of emerging powers. The unilateral actions have provided a unifying cause, accelerating efforts by nations like China, Russia, and India to build a parallel, non-U.S.-centric global financial and trade system.</li></ul><h4>The Playbook for a New Era</h4><p>The international trading system is fundamentally shifting from a predictable, rules-based order to a more volatile, power-based environment.</p><p><em>The Art of Chaos</em> details how different nations are responding — from China’s symmetrical retaliation to the UK’s negotiation-focused approach and the EU’s preparation of large-scale countermeasures.</p><p>In this turbulent landscape, the book argues that the core strategic imperative must shift from a singular focus on <em>efficiency</em> to a new emphasis on <em>resilience</em>. It offers a clear playbook for leaders to navigate this new reality.</p><ul><li><strong>For the Public Sector:</strong> The immediate priority must be to de-escalate trade conflicts with key allies and repair strained relationships. The long-term strategy requires significant domestic investment in innovation and infrastructure, ideally through a coordinated North American industrial strategy, and modernizing the social safety net to address the anxieties that fuel protectionism.</li><li><strong>For the Private Sector:</strong> Large corporations must re-architect their global supply chains for resilience, not just cost, aggressively pursuing diversification through “friend-shoring” and “near-shoring”. Small and medium-sized enterprises (SMEs), which lack the resources to relocate, must leverage partnerships, digital platforms, and collective action to build the resilience they cannot achieve alone.</li></ul><p>The intricate, globalized world built over decades is fragile. A few years of aggressive protectionism can shatter the trust and norms that underpin the entire system. As <em>The Art of Chaos</em> concludes, the challenge for leaders is not to cling to a bygone era, but to proactively shape a new, more resilient global order. The choices made today will determine whether we simply react to the chaos or build a future of stability, competitiveness, and shared prosperity.</p><p><strong>Get the book here</strong></p><p><a href="https://a.co/d/2N6b97a">The Art of Chaos: Tariffs, Trade, and the Fracturing of the Global Order</a></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=844969f2c2a1" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[AI Models Trained, Your Digital Workforce Awaits — How to Maximize Their Potential?]]></title>
            <link>https://medium.com/techemerge/ai-models-trained-your-digital-workforce-awaits-how-to-maximize-their-potential-881a2d36c994?source=rss-2d03c339ebc4------2</link>
            <guid isPermaLink="false">https://medium.com/p/881a2d36c994</guid>
            <category><![CDATA[employee-engagement]]></category>
            <category><![CDATA[change-management]]></category>
            <category><![CDATA[artificial-intelligence]]></category>
            <category><![CDATA[digital-transformation]]></category>
            <category><![CDATA[reskilling]]></category>
            <dc:creator><![CDATA[Som Bando]]></dc:creator>
            <pubDate>Sat, 09 Sep 2023 19:25:42 GMT</pubDate>
            <atom:updated>2023-09-09T19:25:42.292Z</atom:updated>
            <content:encoded><![CDATA[<h3>AI Models Trained, Your Digital Workforce Awaits — How to Maximize Their Potential?</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*FYrS1IFtRtDNxvCI" /><figcaption>Photo by <a href="https://unsplash.com/@nguyendhn?utm_source=medium&amp;utm_medium=referral">Nguyen Dang Hoang Nhu</a> on <a href="https://unsplash.com/?utm_source=medium&amp;utm_medium=referral">Unsplash</a></figcaption></figure><p>In a recent <a href="https://www.mckinsey.com/capabilities/quantumblack/our-insights/the-state-of-ai-in-2023-generative-ais-breakout-year">survey</a>, <a href="https://medium.com/u/777702680128">QuantumBlack, AI by McKinsey</a>, reported that 75% of respondents expected #GenAI to cause significant transformation in their industry over the next 3 years.</p><p>In this pivotal moment, we can incorporate and <strong>build upon lessons from previous technology-driven changes </strong>to create a successful strategy for maximizing the potential of our digital workforce.</p><ul><li>Given the monumental shift, there’s a pressing need for employee training and upskilling.</li><li>To navigate this transition successfully, both employers and employees must prioritize reskilling initiatives.</li><li>The lessons from previous transformations underscore the critical importance of investing in training and development.</li></ul><h4>1. Reskilling as a Strategic Imperative</h4><p>In previous technological revolutions, companies that recognized reskilling as a strategic imperative gained a competitive edge. For instance, Infosys reskilled over 2,000 cybersecurity experts, positioning itself as a leader in the space.</p><p>The future will likely see companies viewing reskilling as an even more critical strategic imperative. As technology advances at an unprecedented pace, businesses will need to adapt swiftly to stay competitive.</p><h4>2. Shared Responsibility</h4><p>Successful companies always recognized that employee training was a core strategic component to gaining an advantage in the marketplace.</p><p>In the coming years, the roles and responsibilities of leaders and managers in reskilling will evolve. Some companies will adopt more decentralized approaches, asking teams, and individual employees to take charge of their learning journeys. Such an approach mirrors the agile, self-directed learning seen in tech giants like Google and Amazon.</p><h4>3. Adapting to Continuous Change</h4><p>Technology transformations of the past have shown that reskilling is not just about training; it involves redefining roles and redeploying employees, and together it is a change-management initiative.</p><p>IBM’s transition from a hardware-focused company to a services and software leader involved extensive reskilling, with a focus on understanding supply and demand for skills.</p><p>The future of work will be characterized by continuous change, requiring a change-management paradigm for reskilling. Rather than treating reskilling as a one-time event, organizations need to establish a culture of lifelong learning.</p><blockquote>Therefore, the emphasis will be on adaptability, with employees constantly acquiring new skills to stay relevant in an ever-evolving job market.</blockquote><h4>4. Personalized Employee-Centric Learning Paths</h4><p>In previous transitions, companies successfully encouraged employees to reskill when they clarified the benefits. IBM, during its transformation, communicated openly with employees, explaining how reskilling would secure their future roles.</p><p>Current efforts will become increasingly personalized. With the aid of AI and data analytics, companies will tailor learning paths to individual employee needs and career aspirations.</p><blockquote>personalization of reskilling will not only boost employee engagement but also ensure that reskilling efforts align more closely with organizational goals.</blockquote><h4>5. Collaborative Ecosystems</h4><p>The idea that reskilling involves an ecosystem is not new. During the industrial revolution, governments, industry, and educational institutions collaborated to address skill gaps.</p><p>We can anticipate a more collaborative ecosystem for reskilling on AI. Governments, industries, and educational institutions will forge stronger partnerships to address skill gaps. Public-private initiatives will provide funding, policies, and programs to ensure a seamless transition into the digital workforce of the future.</p><h4>In Conclusion</h4><p>In the rapidly evolving era of AI, automation, and emerging technologies, these forward-looking perspectives, combined with historical lessons, will shape the reskilling revolution.</p><blockquote>Embracing change, adopting agile learning methodologies, and fostering a culture of continuous improvement will be vital for organizations seeking not only to survive but to thrive in the transformative landscape.</blockquote><h4>What do you think?</h4><ul><li>What specific reskilling initiatives should employers and employees prioritize in the context of AI and emerging technologies?</li><li>How can companies effectively communicate the benefits of reskilling to employees?</li><li>Will the collaborative ecosystem for reskilling on AI be different from previous collaborative efforts to address skill gaps?</li></ul><p>I am interested to learn your perspectives.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=881a2d36c994" width="1" height="1" alt=""><hr><p><a href="https://medium.com/techemerge/ai-models-trained-your-digital-workforce-awaits-how-to-maximize-their-potential-881a2d36c994">AI Models Trained, Your Digital Workforce Awaits — How to Maximize Their Potential?</a> was originally published in <a href="https://medium.com/techemerge">TechEmerge</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Can blockchains survive 51% attacks?]]></title>
            <description><![CDATA[<div class="medium-feed-item"><p class="medium-feed-image"><a href="https://medium.datadriveninvestor.com/51-attacks-are-happening-more-frequently-2400ffe42c4f?source=rss-2d03c339ebc4------2"><img src="https://cdn-images-1.medium.com/max/1000/1*4aHV8yJJ4ab06WG6wWZPyw.jpeg" width="1000"></a></p><p class="medium-feed-snippet">Understanding the vulnerability, recent 51% attacks and potential remedies</p><p class="medium-feed-link"><a href="https://medium.datadriveninvestor.com/51-attacks-are-happening-more-frequently-2400ffe42c4f?source=rss-2d03c339ebc4------2">Continue reading on DataDrivenInvestor »</a></p></div>]]></description>
            <link>https://medium.datadriveninvestor.com/51-attacks-are-happening-more-frequently-2400ffe42c4f?source=rss-2d03c339ebc4------2</link>
            <guid isPermaLink="false">https://medium.com/p/2400ffe42c4f</guid>
            <category><![CDATA[cryptocurrency]]></category>
            <category><![CDATA[blockchain]]></category>
            <category><![CDATA[bitcoin]]></category>
            <category><![CDATA[ethereum]]></category>
            <category><![CDATA[technology]]></category>
            <dc:creator><![CDATA[Som Bando]]></dc:creator>
            <pubDate>Wed, 09 Jan 2019 06:50:51 GMT</pubDate>
            <atom:updated>2019-01-13T03:50:45.347Z</atom:updated>
        </item>
    </channel>
</rss>