The conversation between Brendan Ryan, Partner at Algorithmic Investment Models (AIM), and Gerald Gugger, Senior Financial Consultant at Manning & Napier, starts where most AI discussions should but rarely do — with candour about the hype and pragmatism about what actually works on the ground1.
Gugger is not a tech evangelist. He is a financial advisor who came up through portfolio management, spent years straddling the institutional and intermediary divide, and now works directly with high-net-worth individuals. That background — varied, analytically rigorous, commercially grounded — shapes his view of AI: not as a revolution to be embraced uncritically, but as a set of tools to be tested, validated, and deployed where they demonstrably work.
The Structural Shift Advisors Cannot Ignore
Before AI enters the picture, Gugger situates the broader transformation underway in wealth management. The old model of distributing single-solution products to intermediaries is fading. In its place, a barbell architecture is emerging. "More and more portfolios are being built with sort of a core passive slug and then very tactical niche strategies that are doing something different," he says. "Adding value, providing risk management downside protection, targeting an asset class that is otherwise getting missed."
This structural shift matters as context for the AI adoption conversation. The advisor's role is no longer product distribution — it is orchestration. "Whether you're advising million dollar grandma or a billion dollar pension fund," Gugger observes, the client is "still looking for an orchestrator, someone who can see the big picture, know where to allocate, to understand what's the right sort of portfolio structure for me and my goals and then guiding them to get there."
Ryan frames the stakes clearly: "All of this technology is benefiting the winners. The biggest, most successful advisors can now manage more money than they could before, more easily. And so the benefits of all these improvements accrue to them." The implication for the rest of the advisory community is pointed. "I think it is really important that people embrace this technology or it's going to be to the benefit of their competitors."
Shared Skepticism, Real Consumer Surplus
Both Ryan and Gugger are skeptical of the capital currently being deployed into AI infrastructure. "I too am skeptical of all the dollars that are getting poured into some of this stuff," Gugger says flatly. "I don't know where the ROI is going to come from. I find it hard to believe that $20 subscriptions are going to justify trillions and trillions of spending right now."
But the distinction Gugger draws next is important. Whatever the investment calculus for hyperscalers, the consumer surplus being created for end users — advisors included — is already extraordinary. "The amount of consumer surplus that is being created by the ability for somebody to go out and interact with a chatbot and get any answer to any question basically right now, written in the way that they want, the tone that they want — upload your context or provide colour on your situation — is amazing."
He illustrates this with his daughter's extended ChatGPT conversation about electricity that evolved into a 30-minute exploration of how the human heart works. "The ability for it to adapt to its speaker and be able to provide educational value and insight in this specific way that worked for her is amazing." The parallel to advisory is direct: the same adaptive framing that helps a child understand biology can help an advisor synthesize regulatory complexity in an unfamiliar jurisdiction.
Context Engineering: The Real Competitive Differentiator
The most technically substantive portion of the conversation concerns what Gugger calls "context engineering" — a concept he borrows from software development. The core insight: the quality of AI output is almost entirely a function of the quality of the context fed into it. "When it has no context on who you are, what you're doing, how it's interacting with you — you'll be amazed at how the output... it doesn't even know how to reply to you really. It doesn't know what to say back because none of that's been fed back into it."
Software engineers have developed systematic pipelines — documentation, project specs, task definitions — that they feed into AI systems before generating a single line of code. Gugger sees direct application to wealth management, where client context is equally rich and equally consequential. "Think about high net worth situations — family dynamics, the number of children or parents, how many homes, what's the income situation, Social Security, how many different accounts, what's the tax situation, what is in a taxable account that can't be moved or can't be sold, and how is tax law likely going to evolve." That complexity does not argue against AI. It argues for disciplined context-building before deploying it.
The Workflow That Works Now
The most immediately actionable part of the conversation covers what advisors can do today. Gugger's framework for AI in the advisory workflow has three ascending layers.
The first is dictation and note-taking. "Even in the absolute most punt situation — you forget to use it, you haven't used it — hey, I just walked out of an hour and a half meeting. I need to get my to-dos down. What happened? What did we talk about?" Gugger describes hitting record and rambling for 10 minutes post-meeting. "That AI agent is going to — even if you're not using the most cutting edge models — be able to parse that, put it in proper context, clean it up, bucket it out. It's already right there, ready to go, absolutely right now."
The second layer is CRM-integrated meeting prep. Tools like Zox, which Manning & Napier has been trialling, can integrate with Salesforce to pull client history, prior notes, and outstanding action items ahead of client meetings. "Its ability to then pull more information in and be able to start dictating before the meeting is pretty much being built out now and happening today."
The third layer — proactive book management — is where Gugger sees the industry heading. "It's so obvious to me that the next step is going to be not only proactively doing this for individual meetings, but proactively helping you manage your book entirely."
On prompting mechanics, Gugger offers a practical tip most advisors will find immediately useful: typing "think" into ChatGPT activates heavier reasoning models. "If you say 'think hard' it uses more, and if you say 'think very hard' it uses more. And then the big one is called ultra think." Small changes in prompt structure, he emphasizes, produce meaningfully better outputs.
Compliance, Trust, and the Human Premium
On compliance — the friction point most frequently cited by advisors — Gugger's position is clear: purpose-built platforms are non-negotiable. "Typically we're not putting client information into just ChatGPT.com — we're using a purpose-built financial advisory system with the right regulations to make sure no conversation is recorded or even transcribed." Tools like Zox are not general-purpose chatbots. They are advisory AI platforms built with data governance at their core.
And on the question of whether AI ultimately threatens the advisor relationship, both Ryan and Gugger are aligned in their view. Ryan argues that in a world saturated with synthetic content, "the trust in actual personal relationship is going to increase." Gugger agrees. "I see these tools as being massive efficiency gainers, quality enhancers — you're just going to be able to do more, more quickly, reduce the operational burden — but it's not going to change the inherent relationship at play or the dynamics of how this industry operates."
3 Key Takeaways for Advisors and Investors
1. Start with low-lift, high-ROI workflows.
AI note-taking and post-meeting dictation are available, compliant (via purpose-built platforms), and immediately impactful. Advisors who are not using these tools are absorbing an operational burden their competitors are already eliminating.
2. Context engineering is the skill that separates effective AI users from frustrated ones.
The quality of AI output is entirely dependent on the richness and precision of the context provided. Advisors who invest in building structured client context — CRM data hygiene, pre-meeting briefing pipelines — will extract exponentially more value than those who treat AI as a simple query tool.
3. AI amplifies the advisor relationship; it does not replace it.
The complexity of high-net-worth client situations — multi-account structures, cross-border taxation, family dynamics, evolving tax law — exceeds current AI capability for holistic autonomous management. The near-term value lies in using AI to handle operational and informational tasks so advisors can invest more deeply in the judgment-intensive, trust-building work that remains irreducibly human.
Footnote:
1 Advisor Turntable Podcast. Ryan, Brendan. "Easy Ways Advisors Can Start Using AI Today – aim." 19 Feb 2026
2 Photo by JESHOOTS.COM on Unsplash
