• AI KATANA
  • Posts
  • China Hosts First Humanoid Robot Boxing Match

China Hosts First Humanoid Robot Boxing Match

Also: Markets Brace for Nvidia’s Q1 Results Amid AI Frenzy

In partnership with

Hire an AI BDR to Automate Your LinkedIn Outreach

Sales reps are wasting time on manual LinkedIn outreach. Our AI BDR Ava fully automates personalized LinkedIn outreach using your team’s profiles—getting you leads on autopilot.

She operates within the Artisan platform, which consolidates every tool you need for outbound:

  • 300M+ High-Quality B2B Prospects

  • Automated Lead Enrichment With 10+ Data Sources Included

  • Full Email Deliverability Management

  • Personalization Waterfall using LinkedIn, Twitter, Web Scraping & More

Hello!

Today’s roundup dives deep into the accelerating convergence of AI with industry, entertainment, and sustainability. China has launched its first humanoid robot boxing match—more than a flashy spectacle, it serves as a proving ground for embodied AI, potentially opening the door to a new genre of robotic esports. Meanwhile, all eyes are on Nvidia’s upcoming earnings, with market watchers debating whether its AI-fueled momentum can withstand margin pressures. On the healthcare front, Google’s AMIE AI model is now outperforming doctors in some diagnostic scenarios, blending text and image data in groundbreaking ways. But as AI systems become more powerful, so do their environmental costs—analysts are ringing alarm bells about the massive energy consumption of AI data-centres. Nvidia also makes a strategic play in China, designing a pared-down chip to sidestep U.S. export controls while still appealing to local demand. In AI tools, Google’s Gemini Live is adding powerful real-time camera-sharing features to Android, intensifying its rivalry with OpenAI. Finally, Google has launched a major investment initiative to back AI startups, reinforcing its stake in the next wave of innovation.

Sliced just for you:

  • 🤖 China Hosts First Humanoid Robot Boxing Match

  • 💹 Markets Brace for Nvidia’s Q1 Results Amid AI Frenzy

  • 🏥 Google’s AMIE: Advancing Medical Imaging with AI

  • 🌱 AI Data-Centres’ Growing Carbon Footprint Alarms Analysts

  • 🐉 Nvidia Plans Low-Cost Blackwell Chip for China After U.S. Curbs

Hangzhou’s World Robot Competition debuted a live-streamed “mecha fighting” series where 1.3-metre Unitree G1 humanoids traded jabs, hooks and acrobatic spin-kicks. Spectators cheered as the lightweight bots popped back up from knock-downs, demonstrating torque-controlled joints and vision-based balance algorithms. Organisers tout the event as a “train-through-competition” lab that will accelerate embodied-AI research; a full-size robot championship is booked for December in Shenzhen. Industry analysts say the spectacle could spark a new esports-meets-robotics genre and drive demand for modular actuator kits.

With Nvidia shares up 88 % YTD, traders are combing supply-chain data ahead of Wednesday’s print. Street consensus expects record data-centre revenue on Blackwell pre-orders and a fresh dividend raise. Options markets imply a 12 % post-earnings swing—bigger than last quarter’s—but bulls argue that hyperscaler cap-ex guides still lag actual GPU lead-times, setting up upside. Some desks warn of margin pressure from the first wave of Blackwell wafers at TSMC’s CoWoS lines, but others note early H100 discounting has been minimal. Portfolio managers tell Yahoo Finance that any miss could spark a sector-wide rethink on AI infrastructure multiples.

Google is enhancing its AMIE AI to interpret medical images using the Gemini 2.0 Flash model, enabling it to process visual data like X-rays and MRIs alongside text for improved diagnostic accuracy. In simulated tests with 105 medical scenarios, AMIE outperformed human doctors in image interpretation, diagnostic accuracy, and empathy, as rated by patient actors. It is now being tested in real clinical settings with Beth Israel Deaconess Medical Center, aiming for telehealth integration. This advancement could enhance patient outcomes and support medical professionals, though real-world complexities and ethical considerations, such as ensuring accuracy in diverse scenarios, remain under scrutin

A PBS NewsHour investigation estimates that U.S. AI-centric data-centres could consume as much electricity in 2030 as the entire state of Florida today. Grid planners warn that delays in transmission upgrades and new nuclear approvals mean more gas-fired peakers unless efficiency improves. Researchers urge vendors to shift training to off-peak renewables and to adopt liquid-cooling, which can cut power-use effectiveness by 30 %. Lawmakers in Oregon and Virginia are drafting bills that would tie tax incentives to verifiable emissions targets. Industry groups counter that AI accelerates green innovation and pledge a voluntary reporting framework by year-end.

Sources say Nvidia will ship a scaled-back “B20” variant of its Blackwell architecture to Chinese OEMs next quarter, priced at under $10k—about 40 % below the restricted H20. The chip reportedly trims inter-GPU NVLink lanes and halves FP8 throughput to meet U.S. export ceilings while preserving enough tensor muscle for local LLM training. Analysts expect Huawei and ByteDance to test samples in July, though success hinges on Beijing’s evolving tech-import rules. Some U.S. lawmakers blasted the plan, arguing it undercuts Washington’s strategic aims, yet Nvidia’s Asia sales teams say demand is “overwhelming.”

🛠️ AI tools updates

Google is rolling out Gemini Live’s new screen- and camera-sharing modes to Pixel 9 and Galaxy S25 owners, making it possible to point your lens at a broken coffee machine, stream the view to Gemini, and get annotated instructions in real time. The update, visible in version 5.3 of the Gemini app, works over on-device Gemini 2.5 Flash for latency under 80 ms and falls back to cloud Gemini Pro when bandwidth allows. Google says the feature will reach all Android users on the Gemini Advanced plan within 48 hours and will arrive on iOS “this summer,” adding multimodal context to its rivalry with OpenAI’s GPT-4o voice mode.

💵 Venture Capital updates

Google launched the AI Futures Fund to invest in startups using DeepMind’s AI tools, offering early model access, expert collaboration, Google Cloud credits, and potential direct investment. Applications opened on May 12 at AI Futures Fund. Supporting startups like Viggle and Toonsutra, this initiative builds on Google’s prior AI investments, though competition for funding and the need for scalable AI solutions may challenge smaller startups.

🫡 Meme of the day

⭐️ Generative AI image of the day