Welcome back to Daily Zaps, your regularly-scheduled dose of AI news ⚡️ 

Here’s what we got for ya today:

• 🤑 $100 million AI job offer

• 🦾 Hexagon taps NVIDIA to build humanoid

• 👯‍♂️ Gemini 2.5 Flash-Lite

• 🧾 OpenAI is undercutting Microsoft on AI pricing

Let’s get right into it!

Sign Up | Advertise | Tools

CAREERS

$100 million AI job offer

Meta has aggressively pursued top AI talent, offering OpenAI employees signing bonuses up to $100 million and even larger compensation packages, according to OpenAI CEO Sam Altman, who noted that none of their best staff have accepted the offers. Meta, which reportedly views OpenAI as a key rival, has faced setbacks with its own AI model releases and is investing heavily to close the gap—most notably by acquiring a 49% stake in Scale AI for $14.3 billion and recruiting its CEO, Alexandr Wang, to lead a “superintelligence” lab.

Despite criticism from Altman that Meta’s copycat tactics and lavish offers could harm innovation culture, others argue that Meta’s contributions to open-source AI—especially through its Llama models—and recent high-profile hires like Jack Rae from DeepMind show it remains a formidable force in AI development.

Direct link to Meta poaching section

BIG TECH

Hexagon taps NVIDIA to build humanoid

Hexagon’s robotics division has unveiled AEON, a humanoid robot developed in partnership with Nvidia, designed for deployment in industries like automotive, aerospace, manufacturing, and logistics. Revealed at the Hexagon LIVE Global conference, AEON features advanced manipulation, inspection, and reality capture capabilities, including autonomous 3D scanning and defect detection.

Built using Nvidia’s full-stack robotics platform—including AI supercomputers, Omniverse simulation tools, and Jetson Orin edge computers—AEON leverages multimodal sensor fusion and real-time AI to operate independently. The collaboration also uses Nvidia's Isaac platform to enhance AEON’s reasoning and generate synthetic motion data. AEON’s data feeds into Hexagon’s Digital Reality platform for creating high-fidelity digital twins using OpenUSD and Omniverse, pushing the frontier of physical AI solutions.

BIG TECH

Gemini 2.5 Flash-Lite

Google has officially released its Gemini 2.5 Pro and 2.5 Flash models for general availability, signaling they are now stable and production-ready for real-world use—already adopted by companies like Snap and Spline. Alongside this milestone, Google introduced Gemini 2.5 Flash-Lite, a new preview model optimized for speed and cost-efficiency.

Designed for high-volume, latency-sensitive tasks like translation and classification, Flash-Lite outperforms the older 2.0 Flash-Lite across benchmarks in coding, math, science, and reasoning. Despite its “Lite” label, it retains core Gemini 2.5 features such as multimodal input, a 1 million-token context window, and integration with tools like Google Search and code execution, making it a powerful yet efficient option for developers.

STARTUPS

OpenAI is undercutting Microsoft on AI pricing

OpenAI has begun offering 10–20% discounts on ChatGPT Enterprise subscriptions to companies that commit to purchasing additional AI products or API usage, a move aimed at boosting enterprise adoption but causing friction with Microsoft, its key partner and competitor. While OpenAI’s strategy reflects a desire to pass on cost savings from improved model efficiency and attract long-term clients, it has led to Microsoft losing deals when unable to match the price cuts.

OpenAI projects significant enterprise revenue growth—up to $14.7 billion by 2030—and is gaining traction with major firms like Notion and Salesforce due to access to engineers and roadmap visibility. Meanwhile, Microsoft sales teams have struggled to compete on pricing, even as they tout Azure’s compliance and security features. The pricing war highlights growing competition in the enterprise AI market and the strategic tensions within the OpenAI-Microsoft partnership.