Utilizing Tech

The episode contains some genuinely useful framings - the distinction between probabilistic/creative vs. deterministic/rule-based AI processes, the S-curve maturity analogy, and the customer service as horizontal entry point - but much of the runtime is spent on soft reframing and repetition of obvious points (companies are early, things move fast, trust is a concern). Nick Patience provides real structure for thinking about agentic AI tiers and vendor dynamics, but these insights are interspersed with considerable throat-clearing and abstract musing that adds limited value to an operator already tracking the space.

“if your problem involves a load of structured data, um, such as your customer records, your employee records and things like that, and that's where the automation is coming from, then that's probably going to end up in a fair amount of deterministic processes. Um, if the problem you're trying to solve involves a load of unstructured data... then you're going to end up with more probabilistic kind of um, challenges”

“I think it's similar to um, every kind of AI trend we've seen from back in the predictive days. You start with the things that are horizontal so they're not vertical specific... usually that is um, the first kind of opportunity”

While the structured/unstructured data heuristic and the MLOps-to-AgenticOps progression are sensible frameworks, they are not contrarian or surprising to anyone who has worked through prior AI cycles. The commentary on platform shifts, vendor consolidation, and regulatory maturation largely recycles familiar analyst talking points. The guest avoids truly challenging assumptions (e.g., whether current LLM-based agents will ever work at scale for mission-critical tasks, or if the deterministic/probabilistic split is even the right conceptual model).

“every company has some sort of customer service um challenge ahead of them... usually that is um, the first kind of opportunity”

“you're going to get some winners out of that I suspect. Um, um, that will either survive or get, or have a good exit”

Nick Patience is a legitimate industry analyst with 25 years of background in AI/ML research and founded 451 Research, giving him real credibility and tenure in the space. However, he is primarily an analyst and researcher rather than a practitioner who has built and scaled agentic systems in production. His insights are informed but not rooted in hands-on operational experience deploying these systems at enterprise scale, which limits how actionable his guidance can be for B2B operators facing real implementation decisions.

“I'm the vice, um, president and AI Platforms practice lead at Futurum Research... I've been looking at AI for over 25 years. I started another analyst company called 451 Research back in 2000”

“I go to a lot of um, technology vendor conferences”

The episode lacks concrete data, named customers, performance metrics, timelines, or dollar figures. Nick mentions studying help.com sites and finding a lack of true autonomous agents, but provides no quantitative findings. References to Salesforce, Oracle, Microsoft, Workday, and ServiceNow building agentic tools are vague assertions without evidence of capability, timeline, or customer traction. The discussion remains largely theoretical and abstracted.

“we did actually a project where we looked at how many of those help services are agentic, um, or how many of them are not agentic... And it's amazing how um, the lack of true agents, autonomous agents um exist in those kind of environments”

“Salesforce, Oracle, Microsoft, Workday, ServiceNow, all these companies are obviously building out their own agentic um tools, platforms, applications”

Stephen Foskett asks reasonable opening questions and attempts to push on AGI risk and trust, but largely allows Nick to meander through long, winding answers without sharp follow-ups or productive pushback. When Nick makes vague claims (e.g., about AGI risk or vendor dynamics), the host does not press for specifics or evidence. Frederik Van Haren raises one good question on the trust/complexity paradox but is then largely sidelined. The conversation reads more as a forum for Nick's analyst perspective than a probing investigation of contested or uncertain claims.

“What is your perspective on the timeline, especially around agentic AI?”

“Do we have a trust problem with agentic AI?”