The Good Tech Companies - The Day the Cloud Cracked: AWS Outage Exposes Fragility of Centralized Internet

Episode Date: October 21, 2025

This story was originally published on HackerNoon at: https://hackernoon.com/the-day-the-cloud-cracked-aws-outage-exposes-fragility-of-centralized-internet. AWS crashed for 15 hours taking down Snapchat, Fortnite and 2,500+ companies. 11M users affected. What went wrong. Check more stories related to data-science at: https://hackernoon.com/c/data-science. You can also check exclusive content about #data, #aws, #amazon, #ai, #good-company, #argentum-ai, #cloud-infrastructure, #web3, and more. This story was written by: @ishanpandey. Learn more about this writer by checking @ishanpandey's about page, and for more stories, please visit hackernoon.com. Amazon Web Services (AWS) went dark for 15 hours on October 20th. The outage affected over 11 million users and more than 2,500 companies. The problem traced back to domain name system issues. The DePIN (Decentralized Physical Infrastructure Networks) market is surging.

This audio is presented by Hacker Noon, where anyone can learn anything about any technology. The day the cloud cracked, AWS outage exposes fragility of centralized internet.B.Y Ashon Pondi. October 20th wasn't just another day of technical difficulties, but a reminder that the infrastructure holding up modern life hangs by a thread because when Amazon Web Services, AWS, went dark for nearly 15 hours, it took with it troves of Snapchat conversations, fortnight matches, and even food orders for a whopping 15 hours, directly affecting over 11 million users and more than 2,500 companies. The problem traced back to domain name system issues, something so fundamental and buried in

the plumbing of the internet that most people never even realized what was happening till everything stopped working. For context, AWS currently controls approximately 30% of the global cloud infrastructure, nearly equal to Microsoft Azure and Google Cloud share combined, meaning that when something goes awry, its ripple effects can be seen everywhere. Market share controlled by leading cloud infrastructure providers. Source Statista, that said, what made this particular outage notable wasn't just at scale but the uncomfortable questions it raised about resilience, redundancy, and the wisdom of letting so much critical infrastructure rest on the shoulders of AS elect few

companies. In fact, this is exactly the reason why the DEPIN decentralized physical infrastructure networks, market has been surging, with more than 13 million devices already contributing to various services, enabling users to participate directly in them, earning tangible monetary value in the process. The Akash network, for instance, operates a decentralized cloud computing marketplace connecting compute buyers and sellers, while Golem has built a P2P marketplace for compute power, with its network focused on high-performance tasks like rendering, simulations, and complex batch jobs. Similarly, Flux is another offering that currently hosts over 10,000 nodes to run everything from DAPPs to distributed hosting networks,

carving up a new reality with Argentem AI leading the front. While the aforementioned projects have undoubtedly amassed massive appeal with developers in terms of rendering workloads, none of them have been built specifically for enterprise usage, an aspect that has been found to have a direct influence on the global economy. In this regard, Argentum A AI delivers enterprise-grade AI infrastructure by integrating secure enclaves, zero-knowledge proofs, ZKPs, and staking-based trust mechanisms, enabling confidential computing and verifiable execution on the hardware front. Its model centers around second and third life NVIDIA GPUs retired from hypers and data centers that still have years of productive life but would otherwise

end up in landfills or sitting idle indecommissioned wrecks. By redeploying these units, Argentum AI delivers compute at up to 70% lower cost than traditional cloud providers while simultaneously reducing e-waste. Furthermore, Argentum AI's marketplace aggregates compute capacity globally, automatically routing workloads to available GPUs anywhere in the world, eliminating regional bottlenecks and single-point failures in the process. In other words, if one region experiences issues, the network routes around it. iPhone provider goes offline, another picks up the slack, Similarly, its AI is human-centered and market-trained, learning from real marketplace behavior to optimize efficiency and fairness, creating what the company calls a living benchmark or

intelligence that can improve over time-based on actual usage patterns rather than theoretical models cooked up in Allob. Even more importantly, Argentim AI recognizes that enterprise buyers have different needs than retail developers. As a result, it has devised a framework capable of offering Fortune 500 companies and institutional buyers the infrastructure they actually require, one that is auditable, compliant, and built for AI workloads that matter. A message that needed to be heard. From the outside looking in, the AWS outage has served as a much-needed wake UP call, but it wasn't a surprise given that

centralized clouds have always carried systemic risks with them. The question was never if an outage would happen, but when, and how badly it would hurt internet users across the globe. However, Argentem AI has proved that setups of a non-local nature can work well, even at an enterprise scale, meeting requirements of any level effortlessly, even as AI workloads continue to grow and reshape practically every industry on the planet. Looking ahead, it will be interesting to see how the market evolves and adapts when situations like the AWS outage arise. Don't forget to like and share the story.

Thank you for listening to this Hackernoon story, read by Artificial Intelligence. Visit hackernoon.com to read, write, learn and publish.