I published my first full framework for applying Deterministic Code in the Loop. The use case is cloud migration, but the pattern is bigger than migration. For the past year, the industry has talked a lot about human-in-the-loop AI. That is useful, but it does not scale as the primary control model for enterprise automation. A human cannot inspect every recommendation, every code change, every migration step, every policy decision, and every exception across a large application estate. So the question becomes: what sits between probabilistic AI and human review? My answer is deterministic code in the loop. In this framework, LLMs do not own the migration process. They are called only when developer-like adaptation is required. The control plane owns the process. - Playbooks constrain the work. - Deterministic code executes known transformations. - Agent harnesses provide controlled execution. - Validators define done. - The landing zone validates fit. - Traceability records authority. - Humans handle exceptions and authority boundaries. That distinction matters. Human-in-the-loop is not wrong. It is just insufficient by itself. At enterprise scale, humans should not be the first line of defense against every AI mistake. They should sit at the authority boundaries: low-confidence decisions, exceptions, policy conflicts, failed validators, and material risk. Everything else should be constrained, tested, validated, and recorded by deterministic systems. Cloud migration makes the pattern visible because the risk is obvious. You cannot simply let an agent assess, plan, refactor, validate, and migrate workloads without knowing where decision authority lives. This is why I frame the problem as a migration control plane, not an AI migration factory. The core loop: - LLM proposes. - Playbooks constrain. - Deterministic code enforces. - Agent harness executes within bounds. - Landing zone validates. - Traceability records authority. - Humans resolve exceptions. That is the practical model for using AI in enterprise automation without letting decision authority drift into the model. Read the framework here: thectoadvisor.com/blog/2026/…

Nice official video on building an AI cluster on using Macs over Thunderbolt 5 WWDC26: Explore distributed inference and training with MLX | Apple youtu.be/CzgK02zsRg4?si=EMD8… via @YouTube

What I love about this video is that the creator built a rough Age of Empires clone for roughly $65 worth of Fable 5 credits. He could buy Age of Empires for $20 to $60, with no additional development overhead. The upside is that he could get a fully custom experience, and it may be worth the time and cost. However, that's a different outcome from wanting a Strategy/RPG experience.

The beauty of this post is that it's a meme for a real conversation. Can I build a more customized solution for my workflow has never been the problem. The question is whether the lift is worthwhile, given that it comes at a cost well beyond that of the build cost.

Tokens are the wrong way to measure the value of AI consumption - Period. What's the right measurement? I don't know. I just know that people way smarter than me on capacity management are spending one year's worth of AI budgets in one quarter. And the consistent feedback is that tokens aren't an indication of value.

I was in the middle of watching this fascinating comparison of Opus 4.8 and Fable 5 when I got the alert that govment shutdown Fable 5. The video clearly showed how Fable is a much more capable model for these coding tasks. This challenges my notion that older models are good enough for the work I do today. However, if I were running a business that depended on these models being available, this government action would give me pause. youtu.be/TzJCly4YgDQ?si=m6G9…

I cancelled my $10/mo Calendly subscription and vibe coded my own with Fable for $12,000

I have a budget alert set for $50/Month for my @googlecloud account. I'm serving up a bit on this $50/month. advisorbench.net virtual.thectoadvisor.com layer2c.com cloud.layer2c.com advisorbench.net/simulator cto-scanner.web.app/ I just got an alert I'm 50% through my budget. I may need to cut some of this down or reset my budget :)

The car analogy I think of is a little different: in F1 Mercedez, McLaren, Williams and Alpine share the same Mercedez engine. The gap in performance between them shows how much the car around that engine determines the result.

Which AI company should be the highest valued $TSLA $SPCX?

I’m done with my draft assessment of @Canonical OpenStack against my Fourth Cloud Framework. I sent the assessment to Canonical to check accuracy. Openstack has been the top requested addition to the research instrument.