MENGENAI EPISOD INI
In this episode, Keira sits down with Gaurav Sharma, CEO of io.net, a decentralized GPU network that allows gamers, individuals, and data centres around the world to contribute computing power to a global marketplace.Gaurav has spent more than 16 years building internet infrastructure. His career has included engineering roles at Calsoft, eBay, Amazon, Agoda, and Binance. While helping scale Agoda to more than 100 million users, he experienced the GPU shortage firsthand—even a major company with the money to buy Nvidia GPUs had to wait months to receive them.That problem is now at the centre of io.net’s mission.Gaurav explains why the concentration of AI compute among Amazon, Google, and Microsoft is limiting global innovation, and why access to GPUs can determine which startups survive. io.net is building what he describes as the “Airbnb for GPUs,” connecting businesses with idle computing power from data centres, gamers, and individual suppliers across more than 130 countries.They also discuss:- Why io.net can offer compute at significantly lower prices than traditional cloud providers- How decentralized GPU networks could make AI more accessible globally- Why blockchain is important for transparency and trust—not just tokens- How io.net puts marketplace activity and company finances on-chain- Why Gaurav does not believe AI is currently in a bubble- How AI agents and vibe coding could enable the rise of single-person companies- Why founders need to become comfortable failing and moving quickly- What Gaurav would tell his 17-year-old selfGaurav also shares the story of a university student who travelled from Gwalior to Delhi simply to thank him for making GPUs affordable enough to complete his graduation project.KEY TAKEAWAYS- AI compute is concentrated among a small number of companies, making it difficult for startups and developers to access the GPUs they need.- io.net connects idle GPUs from data centres, gamers, and individuals with businesses that need computing power.- Decentralized compute could dramatically reduce the cost of building and scaling AI products.- Blockchain allows io.net to make pricing, payments, revenue, and marketplace activity transparent and verifiable.- The demand for GPUs continues to grow as more companies adopt AI.- AI tools, decentralized infrastructure, and vibe coding are lowering the barrier to entry for founders around the world.KEY TIMESTAMPS00:00 – Gaurav Sharma’s journey through 16 years of internet infrastructure01:46 – Why three companies control most AI compute03:00 – How GPU access shaped competition inside Agoda05:00 – Why a funded startup waited four months for GPUs09:47 – How io.net’s decentralized GPU marketplace works13:15 – What micro data centres look like15:02 – How GPU providers join the io.net network16:34 – How io.net protects the network from bad actors19:24 – Who needs GPUs and why demand is growing21:33 – What a GPU is, explained simply23:27 – Is AI currently in a bubble?29:36 – Why decentralized compute can be significantly cheaper33:21 – Competing with Amazon, Google, and Microsoft35:38 – Why blockchain matters beyond token incentives42:42 – io.net’s growth and Gaurav’s move from CTO to CEO47:07 – The future of the agentic economy49:55 – Will AI make humans dispensable?55:34 – The student who travelled to thank Gaurav59:27 – Lessons from 16 years in technology01:04:14 – Advice to his 17-year-old selfGUESTGaurav Sharma – CEO, io.netWebsite: io.netX: @Gaurav_ionetio.net on X: @ionetLinkedIn: Gaurav SharmaHOSTKeira Nesdale – MH Ventures Portfolio ManagerX: @RealMissAILinkedIn: Keira NesdaleDealFlow PodcastX: @dealflowpodcastMH VenturesWebsite: mhventures.ioLinkedIn: MH Ventures#ionet #DePIN #GPU #AI #DecentralizedCompute #Web3 #Blockchain #AIInfrastructure #MachineLearning #Crypto #Startups #DealFlowPodcast
