Archish Arun

At NVIDIA, the pace is set at the TOP ⚡ Jensen Huang doesn’t enforce speed with policies. He enforces it by example. He carries more pressure than anyone else. That becomes the standard. ▫️ Intensity works as enforcement ▫️ Effort is visible, not negotiated ▫️ Authority comes from absorbing the strain first The message is clear: this is the cost of being here. I broke down why Jensen Huang leads this way, and the price he chose to pay to build NVIDIA to last. Full breakdown 👇 https://lnkd.in/eUTqBYA2

26

2 Comments

Incredible blog post about the inner workings of GPUs. Extremely informative and accessible. Highly recommended for machine learning people in the age of GPUs.

2

Unpopular opinion: The RL hype is a red flag for technical depth. When I hear founders or VCs casually dropping "reinforcement learning" in pitches, it's usually code for "we don't actually understand our problem space." Here's the reality check: RL only works when you have: - Perfect simulation environments - Extremely fast iteration cycles - Clear reward functions - Tolerance for massive compute costs That's why it succeeds in games (perfect simulation) and search optimization (Massive compute capacity at Deepmind AlphaGo/Anthropic/OpenAI) but fails spectacularly in the messy, slow-feedback real world. The graveyard of "RL-powered" startups tells the story. They burned through $$$$ trying to apply algorithms to problems that needed simple heuristics or supervised learning. Most of the time, a well-tuned evolutionary algorithm will outperform your fancy RL setup while using <1/10th the compute and delivering results you can actually explain to stakeholders. The next time someone pitches you an RL solution, ask them: "Have you tried solving this with a basic optimization approach first?" Watch how quickly the conversation changes. I am surprised when people are surprised that Direct Preference Optimization (DPO) can exceed PPO-based RLHF [1] or Reflective Prompt Evolution Can Outperform Reinforcement Learning [2]. [1] https://lnkd.in/gCY3bcTQ [2] https://lnkd.in/gQRwyQyM

45

2 Comments

Podonos just raised $2.4 million in pre-seed funding, and the signal is loud and clear: the voice AI industry finally has someone measuring the sound, not just making the noise. Based in Los Gatos, Podonos is building the infrastructure layer that decides whether voice AI feels real, human, and ready for primetime. The industry is racing toward a projected $47.5 billion by 2034, but growth without standards is chaos. Podonos is making sure the hype actually holds up when you press play. At the center is ✦Soohyun Bae, PhD in Computer Engineering from Georgia Institute of Technology and Y Combinator W22 alumnus. His track record stretches from engineering roles at Google Maps to leading AR mapping at Niantic, plus co-founding Bobidi and TickTock AI. He has seen what happens when technology scales without proper #guardrails. With Podonos, he's betting the future of AI isn't just about how smart the models get, but how believable and trustworthy they sound. Podonos evaluates models across naturalness, similarity, emotion, recognition accuracy, pronunciation, tone, and resilience. Add in #personaconsistency, and suddenly you're not just testing machines, you're testing their ability to pass as human. The twist? They deliver results in under twelve hours, when the legacy approach drags on for two months. In AI time, that's the difference between relevance and irrelevance. Investors caught the beat. Serac Ventures, led by Kevin Moore, took the lead, backed by NAVER D2SF, the venture arm of South Korea's NAVER, and KAIST Venture Investment Holdings. Together, they've now backed Podonos twice, with this $2.4 million round stacking on top of a $750,000 raise in March 2025. Total funding sits at $3.15 million, and every dollar is pointed toward scaling #engineering, refining #analysistools, and launching into the U.S., Europe, and Southeast Asia. This isn't theory. Podonos already serves Resemble AI, Play AI, and Sanas AI, hitting six-digit ARR by November 2024 with just four employees. Their secret weapon is reach: 150,000 evaluators spread across nine languages and thirteen locales, forming a human-in-the-loop platform that blends speed, rigor, and global coverage. That reach lets Podonos deliver evaluations at a scale and pace that no competitor has matched. The strategy is ambitious: expand into high-demand sectors like #healthcare, #finance, #gaming, and #advertising, then move beyond voice into #multimodal model evaluation, spanning video AI and large language models. The mission isn't just to check if AI works, it's to decide if AI works for people. #Startups #StartupFunding #EarlyStage #VentureCapital #PreSeed #AI #AIInfrastructure #Voice #VoiceAI #VoiceTech #Infrastructure #Technology #Innovation #TechEcosystem #StartupEcosystem #Hiring #TechHiring If software engineering peace of mind is what you crave, Vention is your zen.

6

1 Comment

Do you know what is being built by the "king of inference," NVIDIA? I spent part of my holiday feeding my inner tech nerd and catching up on NVIDIA GTC 2026. It did not disappoint. GTC ran March 16–19 in San Jose, California. A couple of things stood out for me from Jensen Huang's keynote which lasted for almost three hours. The scale of what NVIDIA is building is hard to ignore. They now expect at least $1 trillion in AI infrastructure revenue between 2025 and 2027, double what they had projected before. The Blackwell Ultra chips, the Vera Rubin architecture coming next, NVIDIA Dynamo for running AI at scale. This is not a company iterating slowly. Things are moving fast. OpenClaw was another highlight. In simple terms it is a platform that allows AI agents to actually do things on your behalf. For businesses this means automating workflows, handling customer interactions, managing operations and freeing up time for the work that actually needs a human. We are moving from asking AI questions to AI actually handling tasks for you. The physical AI section was my favourite part. Mercedes, Toyota, BYD, GM, Hyundai all building autonomous vehicles on NVIDIA platforms. Uber rolling out robotaxis across 28 cities. Robotics partnerships with ABB, FANUC, KUKA, Figure and others putting AI into factories and real world environments. And then Jensen brought Disney's Olaf robot on stage. Actually walking. On stage. I did not see that coming and I loved every second of it. On sovereign AI, Japan, France, Saudi Arabia, India, the UK were all named as countries partnering with NVIDIA to build their own national AI infrastructure. Africa is actually in the mix through Cassava Technologies. They are building NVIDIA-powered AI Factories starting in South Africa, with Kenya, Egypt, Morocco and Nigeria next in line. Beyond the infrastructure, they are working with African governments on national AI initiatives, training local startups, building AI models in African languages and developing use cases around agriculture, healthcare and mobile payments. Honestly, I was glad to see this and it is one I am keen to monitor. I am excited about where all of this is heading, genuinely. But I am also aware that the pace of innovation brings with it real questions around safety, governance and who gets to participate. Should we be excited about these developments or more worried? I would love to hear your thoughts. #NvidiaGTC2026 #AI #PhysicalAI #SovereignAI #CassavaTechnologies #AfricaTech #Innovation #FutureOfAI

4

2 Comments

An early Christmas present came to Groq and Jonathan Ross in a non-exclusive licensing agreement with Nvidia worth $20 Billion! Groq innovated the LPU or language processing unit specialize in inference computation that LLMs rely on… clearly something Nvidia was interested in! On TWiST, Alex and I break down the deal, what it means for Nvidia and for bestie Chamath Palihapitiya. I go deep on the VC fund math for grand slams, like Groq, for Series A venture funds in comparison with seed stage funds which can invest in hundreds of early stage startups.

49

18 Comments