Matt White had lunch with senior DeepSeek staff. One detail stood out more than anything technical. No ego. No race to be first author on a paper. No one chasing a seven-figure salary. Just people working together to solve problems, in a resource-constrained environment. Full Episode: youtu.be/zHi0jy4MK4c?si=rotX… @saraykin

Everyone assumes China's AI labs are black boxes. Matt White, AI CTO at The Linux Foundation, visited DeepSeek, Moonshot, Qwen, and Minimax. What he found was the opposite. Chinese labs invited him in. Shared their research. Showed him what they're building. US labs? They talk off the record. Maybe. If you're lucky. Full Episode: youtu.be/zHi0jy4MK4c?si=rgmd… @saraykin

There are real instances of customer service chatbots giving away refunds they shouldn't have. Some have been jailbroken. Until we build deterministic guardrails in code, agents can't be decision-makers on anything that matters. Find out on Finoverse: youtu.be/zHi0jy4MK4c?si=-cZP… Matt White | @saraykin

There are two competing narratives in the West about China's AI: either they're just copy-pasting everything we do, or export controls are crippling them. Matt White says both are wrong. China figured out how to get more mileage from less. Those innovations are now being adopted by US labs and API providers. Watch the full episode: youtu.be/zHi0jy4MK4c?si=2DNN… @saraykin

A few hours. The power of AI. Magic. ✨ Finoverse held an internal AI Workshop this Monday, and the team built real Claude Skills to automate tasks and supercharge their workflows. Learning together. Creating together. Growing as an AI-native team. Just the beginning. 🚀 #AI #FinTech #AIWorkshop #Finoverse

Finance companies are building safety systems for their specific use cases. Custom guardrails. Works. But where's the general purpose safety framework for agents? Most companies try to solve this through prompts. That won't work. Real safety gets built in code. Find out on Finoverse. Watch the full episode: youtu.be/zHi0jy4MK4c?si=R-Tf… Matt White | @saraykin

"We don't really know what's happening in China." He walked into DeepSeek, Moonshot, and Qwen. China isn't copying. They're innovating. And the West is building on their breakthroughs now. Watch the full conversation: youtu.be/zHi0jy4MK4c Matt White | @saraykin

Large language models learned human biases and became persuasion machines. You see two texts. One is longer. Which do you trust? The longer one. Humans assume longer text has more information, even if it doesn't. The machine learned tricks to convince people. Humans thought it looked intelligent. In reality, it learned persuasion. AI investments boomed. Watch the full episode: youtu.be/gbR1xgHRAFI?si=owU6… Maxim Afanasyev, PhD | @saraykin

MIT research shows 95% of AI applications don't bring bottom-line impact. BCG found the same. The problem isn't technology. Banks lack applied AI experts who connect business understanding, transformation, and technology. Watch the full episode: youtu.be/gbR1xgHRAFI?si=cwC-…

If a CEO cuts 20% of workforce because of AI, they're hiding failure. If technology makes you efficient, use that advantage to grow market share. Not cut staff. Watch the full episode: youtu.be/gbR1xgHRAFI?si=bHqA… Maxim Afanasyev | @saraykin

Raising fund shouldn't feel like this. Cold spreadsheets. Months of waiting. How Token Planet works: Apply → Samantha AI interviews you in details via a call → Demo Day pitching → Investors vote → List on EX.IO → Trade from day one. tokenplanet.ai

Maxim Afanasyev built AI at McKinsey and @Google before most people knew what a neural network was. Now he's watching CEOs panic-buy AI tools and fire 20% of their staff. Watch the full episode: finoverse.oia.bio/49yt