Sponsored by
Making Hydraulics Obsolete
Every excavator, forklift, and crane on the planet runs on hydraulic fluid. It leaks. It fails. It burns through 60% of the energy you put into it. That's been true for a hundred years.
RISE Robotics built Beltdraulics™ to fix all of that. Their patented actuator swaps out hydraulic cylinders for a fluid-free electric system that runs up to 3X faster and cuts operating costs by 50%. No oil. Full digital control. Built-in sensors that hydraulic systems can't touch.
The U.S. military is already a customer. MIT-founded. $9.3M in revenue. 20+ patents protecting the core technology. Dylan Jovine of ‘Behind the Markets’ said RISE “has all the little ingredients to be one of those really big winners.” His readers have been backing it ever since.
You can invest today through the community round on Wefunder.
Today:
Google Introduces Gemini 3.5 to Handle Complex, Multi-Step Tasks
Anthropic Commits $45 Billion to Rent SpaceX Computing Centers
OpenAI Prepares for Public Stock Debut Following Recent Legal Victory
Nous Research Adds Grouped Tool Commands to Hermes AI Assistants
Anthropic Consults Community Leaders to Improve AI Moral Guidelines
Google has officially launched the Gemini 3.5 model family at I/O 2026, marking a heavy pivot from standard conversational AI to "agentic workflows"—AI that can autonomously plan, code, and execute multi-step tasks.
Important Details:
Gemini 3.5 Flash is Live: The flagship model of the release, 3.5 Flash, reportedly outperforms Gemini 3.1 Pro and runs 4x faster than competing frontier models. It is built specifically to handle complex, long-horizon tasks and multi-turn tool calling.
New Agent Platforms: Google introduced Antigravity, a brand-new "agent-first" development platform that allows developers to spin up isolated, persistent Linux environments where AI agents can execute code and collaborate.
Meet Gemini Spark & Omni: On the consumer side, Google is rolling out Gemini Spark, a 24/7 proactive personal AI agent capable of managing your digital life in the background. They also unveiled Gemini Omni, a powerful new model that seamlessly blends text, images, and video to create cinematic, editable media.
In a staggering display of how capital-intensive the AI race has become, Anthropic has agreed to pay Elon Musk’s SpaceX nearly $45 billion over three years for cloud computing resources to train and run its Claude AI models.
Important Details:
The Deal Structure: Disclosed in SpaceX's recent public IPO filings, Anthropic will pay SpaceX $1.25 billion per month through May 2029.
The Infrastructure: The deal grants Anthropic access to more than 300 megawatts of computing capacity housed at SpaceX’s massive "Colossus" data centers in Memphis.
Musk's "Dual Monetization": The filings highlight a clever strategy by Musk. While building his own massive compute clusters for his AI ambitions (xAI/Grok), SpaceX is subsidizing the massive hardware costs by renting out excess compute to competitors. Interestingly, Anthropic's single contract ($15B/year) almost rivals SpaceX's entire satellite/rocket revenue from the previous year.
After securing a decisive legal victory against co-founder Elon Musk, OpenAI is reportedly preparing to confidentially file its IPO paperwork within the coming days. The ChatGPT maker is eyeing a public debut as early as September 2026 in what could be the biggest tech listing of the decade.
Important Details:
Sky-High Valuation: OpenAI recently closed a staggering $122 billion private financing round in March, valuing the company at $852 billion.
The Financial Picture: While OpenAI's revenue continues to grow (hitting an annualized run rate of $20 billion in 2025), the company is facing scrutiny over its astronomical spending. OpenAI has reportedly committed up to $1.4 trillion over the next 7 to 8 years to secure data centers and compute capacity.
The Motivation: Working with Goldman Sachs and Morgan Stanley, CEO Sam Altman appears eager to tap into the public markets to fund these massive capital commitments. A swift IPO also allows OpenAI to lock in capital while AI valuations remain at historic highs, especially as competition from Anthropic and Google heats up.
🧠RESEARCH
AI models can often answer questions about documents correctly while relying on the wrong text, which is dangerous in fields like medicine or law. This paper introduces a new testing method to ensure models properly link their answers to accurate, verified evidence, making these AI systems much safer and trustworthy.
Teaching AI to read very long texts requires massive computing power. The authors propose Lighthouse Attention, a method used only during training. It temporarily shrinks the text, letting the AI learn much faster. After training, the model easily switches back to reading entire texts, lowering costs while maintaining high performance.
Many powerful AI assistants (programs that solve problems or use tools) are locked behind closed doors. This paper introduces Orchard, a free framework for building these independent AI agents. It provides reusable training environments and includes three highly capable models designed for coding, web browsing, and acting as digital assistants.
📲SOCIAL MEDIA
🗞️MORE NEWS
Nous Research: Hermes Skill Bundles Nous Research updated its Hermes assistant with "Skill Bundles," a feature that groups different software tools together into one package. Instead of turning on each tool individually, users can now type a single command to launch a complete workflow (a series of connected tasks). This saves time and makes setting up complicated digital chores much easier.
Anthropic: Widening the Conversation on AI The technology company Anthropic is working with religious and cultural leaders to teach its artificial intelligence (computer systems that mimic human thinking) how to behave ethically, meaning knowing right from wrong. Instead of just focusing on technical computer code, developers want to build good character by giving the system strong human values. They are even testing a feature where the software can pause and review its moral guidelines before making big decisions.
OpenAI: Guaranteed Capacity OpenAI is now offering businesses a steady, guaranteed supply of computing power (the machine strength needed to run complex software) in exchange for one- to three-year contracts. This ensures that a company's digital tools will not unexpectedly slow down or crash during busy times. By offering discounts for long-term loyalty, OpenAI hopes to build more reliable and lasting partnerships with large businesses.
AI Startup Revenues A new report shows that just two companies, OpenAI and Anthropic, now collect nearly ninety percent of all the revenue (total money earned) across the entire artificial intelligence startup industry. This massive market dominance, or complete control over the industry, makes it incredibly hard for smaller competitors to survive and grow. Instead of creating completely new technology, most new businesses are forced to build their products as small add-ons to these two giant systems.
Mark Zuckerberg on Meta Layoffs Meta recently fired about eight thousand employees, which reduces its overall workforce by ten percent. In a company message, Chief Executive Mark Zuckerberg explained that these cuts were needed to afford massive new investments in advanced computer infrastructure (the physical hardware that runs software). To reassure the remaining staff, he also promised that there would be no more widespread job cuts for the rest of the year.
