In Season 5, Episode 25, Karl and Jon are joined by Artem Hrechanychenko, an AWS Community Builder and expert in DevOps and Kubernetes. Together, they discuss CloudFormation Express Mode, Amazon RDS Log Analysis with Qiro and MCP, safely releasing frontier AI models to customers, AWS GPU instance price increases, and AWS adding more data centre capacity than any other company in 2025. Along the way, the conversation somehow shifts from government control over AI models to government control of World Cup football!

08:46 - CloudFormation Express Mode

The first article discusses AWS CloudFormation Express Mode, which accelerates infrastructure deployments by up to four times. The hosts and guest discuss the potential benefits and challenges of using this mode, particularly in scenarios where rapid deployment and minimal waiting times are critical. They also compare it to other infrastructure-as-code tools like Terraform.

16:53 - Amazon RDS Log Analysis with Qiro and MCP

This article explores the use of Amazon RDS log analysis with Qiro and Machine Learning Control Plane (MCP). The discussion focuses on the practical applications of this tool for developers and SREs, particularly in environments where access to production databases is restricted. The hosts debate the effectiveness of this approach versus using built-in AWS tools like CloudWatch Logs.

24:20 - Safely Releasing Frontier Models to Customers

The third article examines the ethical and policy considerations surrounding the release of advanced AI models. The conversation touches on the recent incident with Anthropic's Fable 5 model, government regulations, and the broader implications of AI governance. The hosts and guest discuss the balance between innovation and safety in AI development.

35:06 - AWS GPU Instance Price Increase

This article reports on AWS's decision to raise prices for GPU instances by 20%. The discussion revolves around the implications of this price increase for AI infrastructure costs and the potential for higher costs to be passed on to end-users. The hosts explore the reasons behind the price hike and its impact on the AI ecosystem.

41:08 - AWS Adds More Data Center Capacity Than Any Other Company in 2025

The final article highlights AWS's significant expansion of data center capacity in 2025, claiming to have added more capacity than any other company. The hosts analyze the environmental and economic impacts of this expansion, considering both the positive aspects of job creation and the negative implications for energy consumption and sustainability.

In Season 5, Episode 26, Karl and Jon are joined by AWS Hero Johannes Koch to discuss recent developments across the AWS ecosystem, including Amazon's satellite broadband initiative, the retirement of Amazon Mechanical Turk, updates to coding tool Kiro, a new MCP server for accessing public data, and the announcement of a free AWS sandbox account. Along the way, Johannes also discovers that bloopers don't get edited out!

03:20 - Amazon Leo: Starlink Competitor

Amazon Leo, Amazon's answer to Starlink, aims to provide satellite broadband services. While Starlink has a significant lead with thousands of satellites already in orbit, Amazon is preparing to launch its own constellation. The discussion highlights the technical challenges and potential markets for satellite broadband, including enterprise use cases and connectivity for remote areas.

13:45 - Amazon Mechanical Turk Retirement

Amazon has decided to stop accepting new customers for its Mechanical Turk service, which allowed users to complete small tasks for payment. The service, which has been around since 2005, has faced competition from other freelance platforms and AI advancements. The decision raises questions about Amazon's strategy and the future of human-in-the-loop tasks in AI development.

19:48 - Kiro: Amazon’s Coding Tool

Kiro, Amazon's coding assistant, has recently made other coding models available, including GPT models. This flexibility allows users to choose from various models directly within Kiro, setting it apart from other coding tools. The discussion explores Kiro’s positioning within AWS, its potential for future monetization, and how it compares to other AWS development tools.

36:42 - New MCP Server for Public Data

AWS has launched a new MCP server to access Rhoda, the Registry of Open Data, which contains over 1,122 public datasets. This development aims to make public data more accessible but also brings concerns about the increasing number of MCP servers and the potential for sprawl. The conversation touches on the value of knowing how to query these servers and the future of data access in the AWS ecosystem.

43:40 - Free Sandbox Account

AWS now offers a free sandbox account that users can run for up to 8 hours without incurring any costs. This feature is aimed at making it easier to run workshops and experiments without the hassle of managing accounts and cleaning up resources. While the initiative is praised for its potential, there are discussions about its limitations, such as the single-use nature and the 8-hour time constraint.

In Season 5, Episode 24, Karl and Jon are joined by Gorkem Kestane, a software developer and former Amazon Community Builder specializing in machine learning and cloud technologies. Together, they discuss the introduction of AWS Lambda microVMs and what they bring to the cloud computing landscape. They also explore the growing adoption of the Model Context Protocol (MCP) in DevOps and its impact on integrating AI systems with external tools, review AWS's new certification renewal process that emphasizes continuous learning over traditional exams, analyze the potential of Amazon QuickSight and the challenges it faces in becoming a mainstream desktop application, and debate the EU's stance on cloud gatekeepers, including proposed changes to egress fees and interoperability standards. To top it all off, Karl's wife calls him in the middle of the recording, leaving Jon and Gorkem to debate whether he should have taken that all...

03:50 - Model Context Protocol (MCP)

MCP is an open standard protocol designed to facilitate communication between AI systems and various tools, APIs, and data sources. It aims to eliminate the need for custom integrations by providing a standardized way for agents to interact with enterprise applications. While MCPs are becoming more common, organizations must ensure robust security measures are in place to prevent unauthorized access to sensitive systems.

17:50 - AWS Certification Renewals

AWS has introduced a new method for renewing certifications without taking an exam. This continuous learning model involves completing AWS Skill Builder training and lab activities, making it easier for professionals to maintain their certifications without the need for periodic exams. While this method is beneficial for those who already understand the material, it is not yet available for all certifications.

28:39 - Amazon Quick and Its Strategic Importance

Amazon Quick, born out of QuickSight, is an AI assistant that aims to streamline workflows and automate routine tasks within enterprises. Despite its powerful capabilities, the setup process is complex and costly, involving multiple steps and third-party integrations. The tool's potential lies in its ability to connect enterprise data sources and applications, though it faces challenges in user adoption and integration simplicity.

42:01 - EU's Designation of AWS and Azure as Cloud Gatekeepers

The European Union is considering designating AWS and Azure as "cloud gatekeepers" under the Digital Markets Act (DMA) due to their significant market impact and the fees they charge for data egress. The EU aims to reduce vendor lock-in by forcing these providers to eliminate egress fees and promote interoperability. However, the feasibility of these regulations remains uncertain, as major cloud providers are unlikely to rewrite their platforms to meet open standards.