316: Microsoft’s New AI Agent Has Trust Issues (With Software)

Welcome to episode 316 of The Cloud Pod, where the forecast is always cloudy! This week we’ve got earnings (with sound effects, obviously) as well as news from DeepSeek, DocumentDB, DigitalOcean, and a bunch of GPU news. Justin and Matt are here to lead you through all of it, so let’s get started! 

Titles we almost went with this week:

• Lake Sentinel: The Security Data Monster Nobody Asked For
• Certificate Authority Issues: When Your Free Lunch Gets a Security Audit
• Slash and Learn: Gemini Gets Command-ing
• DigitalOcean Drops Anchor in AI Waters with Gradient Platform
• The Three Stages of Azure Grief: Development, Preview, and Launch
• E for Enormous: Azure’s New VM Sizes Are Anything But Virtual
• SRE You Later: Azure’s AI Agent Takes Over Your On-Call Duties
• Site Reliability Engineer? More Like AI Reliability Engineer
• Azure Disks Get Elastic Waistbands
• Agent Smith Would Be Proud: Google’s Multi-Agent Matrix Gets Real
• C4 Yourself: Google Explodes Into GA with Intel’s Latest Silicon
• The Cost is Right: GCP Edition
• Penny for Your Cloud Thoughts: Google’s Budget-Friendly Update
• DocumentDB Goes on a Diet: Now Available in Serverless Size
• MongoDB Compatibility Gets the AWS Serverless Treatment
• No Server? No Problem: DocumentDB Joins the Serverless Party
• Stream Big or Go Home: Lambda’s 10x Payload Boost
• Lambda Response Streaming: Because Size Matters
• GPT Goes Open Source Shopping
• GPT’s Open Source Awakening
• When Your Antivirus Needs an Antivirus: Enter Project Ire
• The Opus Among Us: Anthropic’s Coding Assistant Gets an Upgrade
• Serverless is becoming serverful in streaming responses

General News 

02:08 It’s Earnings Time! (INSERT AWESOME SOUND EFFECTS HERE) 

02:16 Alphabet beats earnings expectations, raises spending forecast

• Google Cloud revenue hit $13.62 billion, up 32% year-over-year, with OpenAI now using Google’s infrastructure for ChatGPT, signaling growing enterprise confidence in Google’s AI infrastructure capabilities.
• Alphabet is raising its 2025 capital expenditure forecast from $75 billion to $85 billion, driven by cloud and AI demand, with plans to increase spending further in 2026 as it competes for AI workloads.
• AI Overviews now serves 2 billion monthly users across 200+ countries, while the Gemini app reached 450 million monthly active users, demonstrating Google’s scale in deploying AI services globally.
• The $10 billion increase in planned capital spending reflects the infrastructure arms race among cloud providers to capture AI workloads, which require significant compute and specialized hardware investments.
• Google’s cloud growth rate of 32% outpaces its overall revenue growth of 14%, indicating the strategic importance of cloud services as traditional search and advertising face increased AI competition.

03:55 Justin – “I don’t know what it takes to actually run one of these large models at like ultimate scale that like a ChatGPT needs or Anthropic, but I have to imagine it’s just thousands and thousands of GPUs just working nonstop.”

04:31