Tag Archives: AI

Vibing Vs Buying

In this episode, we get into current market news, including pending IPOs, hardware investment, and IBM’s AI-related announcement with Red Hat, and what this may mean for open weights models and hybrid AI. We also talk about hyper-compressed and ultra-quantized models, specialized hardware, and how memory, compute, and power limits are shaping what can run on device and at the edge. We examine the enterprise software implications and the economics of AI pipelines, including token budgeting, guardrails, human supervision, and the need for validation, governance, and operational control.

Transcript: otter.ai/u/z-PbDS6v8w_4n2Eeey…?utm_source=copy_url

AI Harnesses and Hosting

This week we have an exciting conversation about harnesses, neoclouds, open weights models, and the current AI infrastructure landscape. We discuss bare metal automation, agent-friendly APIs, CPU compute, GPU economics, and why some providers are moving up the stack toward models and APIs.
We also look at the harness as an execution, optimization, and governance layer, including tools, sub-agents, routing, and safety controls. The conversation covers smaller and fine-tuned models, continuous validation, model drift, governance, security, compliance, and AI sovereignty, as well as sandboxing and air-gapped execution for agentic systems.

Transcript: https://otter.ai/u/COlnbT2AmRCPpH2K7qCHiDE0d38?utm_source=copy_url

Mirantis IREN Acquisition

In this episode we discuss the acquisition of Mirantis by the neocloud company IREN, and use it to examine the current market for infrastructure and AI deployment. We look at Mirantis’ consulting business, Kubernetes distribution, and its earlier acquisition of parts of Docker, including the registry and enterprise components.
We also consider whether trusted registries could play a role in distributing agents and tools for AI systems, and we compare this idea to package management. We get into an interesting discussion on neoclouds and managed agent providers as part of the operational layer for running AI workloads, including GPU capacity, model hosting, and repeatable deployment.

Vibe Digital Rebar Install [TechOps]

In this episode, we continue our exploration of vibe coding, this time we build a prompt for installing Digital Rebar. We test and refine the prompt across multiple runs, focusing on readiness checks, firewall ports, SSH keys, license file handling, and keeping the model within the provided instructions.We also look at where the install keeps failing, including NAT gateway setup, DNS configuration, key selection, and bootstrap completion, and use those failures to make the instructions more explicit. Finally, we discuss using prompts as install instructions for an agent system and how the same approach could apply to other complex setups like as OpenShift.

AI UX Building

In this episode of Cloud2030, we explore AI’s transformative impact on user experience (UX) and the relevance of stochastic and deterministic systems in designing effective AI interfaces. We give our predictions about the decline of traditional form-based UX in favor of conversational interfaces that allow users to interact with AI more naturally.
We go into how AI can enhance user experiences by dynamically gathering information, while also addressing security concerns related to sensitive data. It’s a great conversation, hope you enjoy!

Transcript: otter.ai/u/_KLPdnZ6UvfApikAAt…?utm_source=copy_url

Back After a Break

In this episode, we discuss the rising cost of using AI and how usage-based pricing, model changes, and capacity limits are affecting daily work as AI moves from experimentation into operational use. We also talk about multi-model workflows, hybrid infrastructure, and examples of using hosted models alongside open models locally for tasks such as writing and named entity resolution. We get into the need for enterprises to run their own AI infrastructure, including questions around GPU pooling, routing, reservation, data sovereignty, and service levels.

MCP Agents and Context

In this episode, we continue our journey even deeper into how agentic vibe coding and other AI-based automation. This time we focus on Model Control Protocol (MCP) and its application in our bare metal automation solution, Digital Rebar. We examine deterministic versus stochastic AI approaches and the importance of reliable system integration without competing with other agentic systems. We highlight MCP’s role in streamlining interactions across data sources, with a focus on practical applications in finance and infrastructure resilience. The episode ends with a preview of future conversations on user experience transformation in infrastructure operations. Enjoy!

Transcript here: otter.ai/u/LmtQ9QAc79izN0PacE…=transcript&tab=chat

Vibe Coding Mapping [TechOps]

In this episode, we continue our Vibe Coding experiment. Now that we’ve figured out how to interface with MaaS, this time we wrestle with mapping and how different systems interact with each other. We’re joined by Greg Althaus, RackN CTO, who reviews the project and asks some really great questions. We talk about our decision to restart the experiment, taking the lessons we’ve learned to the newer software available. Enjoy!

Transcript: https://otter.ai/u/NohM8_D7DfbUu9iZc1MSpxi00kg?utm_source=copy_url 

Vibe Coding for Ops [TechOps]

In this episode, we do some live vibe coding– using AI to write code. We share tips and tricks on having the best vibe coding experience and avoiding some common pitfalls. You’ll get to hear what we do, how we discover what the steps are, just how easy it is to interact with the system, to set up a basic environment. We also start to explore the limitations of vibe coding. We encourage you to listen along and try on your own!

Transcript: otter.ai/u/CqKdtWZWYb3AdPtcb-…?utm_source=copy_url