Will AI Ever Cure Cancer?

Plus, Robotic Horses, Light-Powered AI, and Humanoid Robots Building Hyundai Cars

Author

Steve Brown
April 10, 2025

In partnership with

Subscribe to Synthetic

The week’s most interesting and relevant AI news and analysis

This Week in AI

Automaker Hyundai has committed to deploying tens of thousands of Boston Dynamics robots in the coming years. Boston Dynamics is a subsidiary of the Hyundai Group and was purchased in 2021 for $880 million. Today, Hyundai uses dog-like Spot robots to perform predictive maintenance and industrial inspection tasks. The next wave of deployments will be with humanoid Atlas robots across its factories. 🤖

Synthetic’s Take: Boston Dynamics is a leader in the humanoid robotics sector. Beyond Hyundai, the company has significant partnerships with Nvidia, Google DeepMind, and perhaps most interestingly….Toyota. Expect more news of humanoid deployments in manufacturing and logistics in the coming months.

Fortune explores the multibillion-dollar race to bring the first AI-discovered drug to market. Leaders include Recursion and Alphabet’s Isomorphic Labs. AI can analyze vast amounts of biological data to identify potential drug candidates, shortening research times and accelerating the discovery of cures. However, AI-generated drugs must still pass through rigorous clinical trials before approval. 💊

Quick Hits

Video: 1X NEO Robot Doing Household Chores

We have two videos for you this week: one is real, the other is a future concept, and both are of robots.

In this first, short (30-second) video, robotics company 1X shows their NEO Gamma robot gardening, loading the dishwasher, and tidying up the house. Robots may be in your home sooner than you think.

In our second video, check out Kawasaki’s concept for Corleo, a hydrogen-powered synthetic horse. Yes, you read that correctly.

AI Insights

❝

“Today’s CEOs are the final generation of executives leading exclusively human workforces. Going forward, we’ll need to learn how to manage human workers and digital labor to work together to deliver efficiency and productivity. Integrating AI agents into daily operations will become a leadership skill that separates companies that thrive from those that fall behind.”

Marc Benioff, Chairman and CEO of Salesforce, writing in the Wall Street Journal below

In a Wall Street Journal opinion piece, Salesforce CEO Marc Benioff shares his perspective on the future of work and how agentic AI will revolutionize work to create a more prosperous, sustainable world. He argues that leaders must focus on trust, accountability, and our common humanity to ensure that agentic AI becomes a force for good.

In a bold move, Shopify CEO Tobi Lutke told employees that he wants them to prove they need people rather than AI to expand operations. He also encouraged employees to embrace AI and learn how to use it for the best results. Further, he shared that Shopify will now evaluate workers’ ability to prompt and use AI in their performance reviews. He shared his thoughts in an internal memo.

Synthetic’s Take: Expect to hear more stories like this as leaders encourage workers to embrace AI, learn how to use it, and accelerate their performance. Performance is not just about productivity and efficiency. AI agents will also boost workers’ creativity, intuition, knowledge, and decision-making abilities. Leaders will look at AI to expand operations and elevate customer experience without growing headcount.

AI Tech and Innovation

Meta announced Llama 4 ‘herd’ of open-source models

Meta has launched new open-source models as part of its Llama 4 herd. Two models, Llama 4 Scout and Llama 4 Maverick, are available now, while Llama 4 Behemoth is still in training. Meta claims to lead benchmarks in many areas. Here’s what you need to know:

  • The new Llama 4 models deploy a technique known as a mixture of experts, which combines many smaller models into one. Only the model most relevant for responding to queries is active at any given time. This allows developers to expand their model size while limiting computational costs.

  • Llama 4 Scout has 109 billion parameters, of which only 17 billion are active at any time due to being arranged as a network of 16 experts. It’s optimized to run on a single Nvidia H100 GPU and has an impressive 10 million token context window, which means it can ingest huge documents and videos as prompts.

  • Llama 4 Maverick has 400 billion parameters spanning 128 experts, so only 17 billion are active. Meta claims Maverick outperforms OpenAI GPT-4o and Gemini 2.0 Flash.

  • Both models were distilled from Llama 4 Behemoth, a two trillion-parameter model still in training. Meta claims Behemoth outperforms GPT-4.5, Claude Sonnet 3.7, and Gemini 2.0 Pro on several benchmarks.

  • These models will be Meta’s last without reasoning abilities. Meta has promised that a version of Llama 4 with reasoning is coming, though no timeframe has been offered.

  • Llama 4 is already available to use in Google’s Cloud.

Synthetic’s Take: Meta continues to hold a leadership position in the open-source AI model race, but this announcement smells of “don’t forget about me.” It’s rumored Meta was caught off guard, and Llama 4 was delayed when DeepSeek was released and outperformed what Meta had cooking in the labs. The fact that Meta released Llama 4 details on a Saturday and its knowledge cut-off is August 2024 might tell us something. Benchmarks are somewhat cherry-picked, and Meta is attempting to sell the future with their Behemoth claims. Every AI research lab has models in training, but Meta chose to compare a future model with competitors’ existing offerings.

This week, Google announced its seventh-generation AI chip, a Tensor Processing Unit (TPU) codenamed “Ironwood,” at its Google Cloud Next conference. The chip is optimized for AI inference rather than training, accelerating generative AI and agentic AI applications. Up to 9,216 liquid-cooled chips can be linked together in a single pod with 42.5 ExaFLOPS of computing power. Google claims the new chip delivers twice the energy efficiency of its sixth-generation TPU and 3600x the performance of its original externally available TPU chip—more details from Google here.

Startup Q.ANT has developed a photonic chip that uses light instead of electrical signals to power an artificial neural network. They claim their analog photonic chip could deliver a 30-fold increase in energy efficiency and a 50-fold boost to computing speed compared to conventional silicon-based chips.

Q.ANT projections for photonic energy efficiency versus today’s silicon-based AI accelerators

Synthetic’s Take: AI is rightly criticized for its extreme energy consumption. New approaches, including analog and photonic chips, could accelerate AI deployment and performance without soaking up substantial energy resources that further damage the planet.

Toolkit for the Future

Rapidly build and deploy specialized AI workers with an intelligence infrastructure that gives everyone—from developers to non-technical teams—the power to build with AI. MindStudio already powers over 100,000 AI workers at companies such as Adobe, Google, IBM, and Meta. The easy-to-use, no-code, graphic workflow makes it easy for anyone to build an AI coworker. A free level is available.

Try Artisan’s All-in-one Outbound Sales Platform & AI BDR

Ava automates your entire outbound demand generation so you can get leads delivered to your inbox on autopilot. She operates within the Artisan platform, which consolidates every tool you need for outbound:

  • 300M+ High-Quality B2B Prospects, including E-Commerce and Local Business Leads

  • Automated Lead Enrichment With 10+ Data Sources

  • Full Email Deliverability Management

  • Multi-Channel Outreach Across Email & LinkedIn

  • Human-Level Personalization

Generate ad banners, text, images, and video that outperform those of your competitors with better conversion and click-through rates. Get AI-powered creative insights and analyze your competitors’ campaigns to stay ahead of the curve. Free to try.

This all-in-one AI-powered tool helps you stay on top of your spending and investments and offers personalized guidance and insights on managing your credit and personal finances. There’s a conversational chat interface, and new features are added each month.

Turn customer feedback into action with powerful AI-powered surveying and analytics tools. These tools allow you to measure experience across the customer journey, capture key metrics, convert feedback into growth opportunities, enhance customer satisfaction, and turn customers into loyal advocates. Trusted by brands including American Express, SAP, and Accor, and free to try.

Help your sales team stay focused on customers during meetings. Laxis captures attendees’ comments verbatim and flags items for follow-up. It auto-generates meeting summaries and follow-up emails in seconds, quickly identifying customer requirements, pain points, and action items. Free to try.

Extract data from websites and place it in a spreadsheet that fills itself and notifies you of any changes. Pre-built robots make it a breeze to get started. Extract whatever you need: job listings, property details, TikTok videos, hotel info, or competitor pricing. It's free to try, and no coding is required.