Herbert Kyles, CFP®

The through-line

Across everything Herbert Kyles says, one argument keeps reappearing: technology is only worth having if it ends at a human being making a better decision. He draws a sharp line between the two things people blur together, arguing that "AI and technology are different" and that "[a]rtificial intelligence is really more of like what powers a platform" while "technology is, you know, the use of a software or hardware to deliver it" [1]. The machinery gathers, researches and assembles; the person decides. What makes his firm stand out, he says, is that "the human hasn't been replaced and will not be" [1], and he is precise about which human he means: "This human is not the adviser. This human is the client" [1].

The second half of the through-line is about incentives. He treats the arrangement between adviser and client as the thing that determines whether the technology gets used well or badly, and he returns repeatedly to a single test, whether the person across the table is being served or sold to. He describes what drew him to Farther as infrastructure that lets the adviser "sit truly on the same side of the table as the client" [2], and he frames the resulting economics as "the best conflict of interest you can have" [2]. Over time his emphasis has widened from the mechanics of planning and AI [1] to the whole family system, generational transfer and the logistics of aging [2], but the underlying position has not moved.

On what AI should and should not be allowed to decide

Kyles is enthusiastic about AI while being unusually specific about its limits. He is dismissive of the label as a marketing hook, joking that you can build "toysrs.com," but call it "toysai.com" and "apparently it's the most valuable and craziest thing ever" [1]. His actual use case is narrower and more defensible: using artificial intelligence "to gather data more quickly to piece it all together" and "to research information that is already published, is already vetted, is already accurate, and find the answer more quickly and tie it together," with technology then delivering the result "more efficiently, more eloquently, um, and quite honestly more desiraably" [1].

Where he draws the line is decisions. He explicitly rejects the pitch that "AI will make a portfolio decision for you" or "AI will tell us how much life insurance you need," because "that really should rest upon a human" [2]. What AI should do instead is comb documents no person realistically reads. His example is a homeowner's policy: digging through twenty-five to fifty pages of the binder to confirm an inflation rider is present, so that a home worth $800,000 but costing $1.1 million to replace is actually covered [2]. "You want to use technology to find things that people couldn't really find before unless they read all the fine print from start to finish and they weren't reading all the fine print from start to finish" [2]. He is equally clear that responsibility sits with the user, and that responsible use includes not monetising clients, noting that developing technology in house means "we're not going out and selling data to to make money" [2]. His analogy for the general principle is chocolate: fine in moderation, a problem in excess [1].

On planning that moves at the client's pace

Kyles measures the value of technology by how quickly a client can get an answer to "what if." He describes building plans early in his career with pen and paper, drawing boxes for each account, lines showing where money would flow and how much income it would produce. Clients loved the one-page result, but behind it sat "a tremendous amount of frustration and math," and the plan "wasn't very malleable" [1]. If a client asked what 5% annualized returns or $500 more per month of expenses would look like, the honest answer was that he would come back next week, and in his view "[t]hat really clouds decision making and causes a lot more emotional stress" while the client lies awake wondering what happens at $750 instead [1].

Used well, the modern platform is deliberately unglamorous in his telling: "just one nicer looking Excel spreadsheet that allows you to say, 'Okay, change from 500 to 750. This is what it looks like'" [1]. The point is not speed for its own sake but removing the hurdle in front of a decision while letting the client set the timing. If they want to decide now, the plan updates; if they want two or three weeks, he opens the platform so they can move the widgets themselves [1]. The same logic drives the conversation with a client afraid of running out of money and having to leave the beach at eighty to go back to work: the answer is to model a major market downturn, lower than expected returns and cash management, show what each looks like, and then ask how it feels [1].

On translating finance for two different people at the same table

A recurring preoccupation is that households contain people with very different appetites for detail, and a plan has to satisfy both at once. His standing example is Jack the plumber and Karen the software engineer: Jack is excellent at plumbing and useless at the books, Karen runs everything and knows every detail, and the job is getting Jack to a clean yes or no while giving Karen enough to dig through, monitor and stay engaged [2]. His solution is a single screen delivered over fifteen or twenty minutes, with one click through to a full printout for the person who wants the underlying numbers, questions taken from both, and notes recorded and sent afterwards [2].

He extends this into a broader claim about financial literacy, and he does not blame clients for lacking it. Confusion is "extremely standard," and "it's it's honestly partially our industry's fault," given that opening an account at a bank or local financial institution is often an arm and a leg of headache [2]. He also thinks the alternative to advice is not self-reliance but folklore, noting that everyone has an Uncle Jimmy with a view on which crypto token to buy [2]. Education is where he places the emphasis in generational wealth as well: passing money down without passing down understanding is how people blow it [2]. The same conviction shows up outside finance, in his defence of music in schools, a position he says he will die on because "it inspires creative, logical and critical thinking that I think we can all agree is sort of missing in society to today" [1].

On one screen for the whole financial life

Asked to summarise Farther, Kyles describes letting an individual, a family or an institution, including not for profits, see their financial life in one place: tax, estate planning, investment management and their specific financial goals, on a digital page rather than a paper one [2]. The purpose of putting it all on one screen is communication, so clients understand why and how something is being done and stay invested in the process [2]. His own practice at Farther, which he calls pragmatic wealth planning, is built around packaging tax, estate and insurance together, including making sure home, auto and renters coverage is right rather than "not just selling you life and and long-term care insurance and calling it a day" [2].

The goal behind the consolidation is the removal of stress. "[P]ersonal finances are personal and it shouldn't be something that's stressful where you have to take on the second job of being a portfolio manager at 7 p.m. analyzing your portfolio after work," he says [2]. He makes the same case on cost and structure: built correctly, a practice can tell a client it will use a product without locking them in for five years, or move them into a lower cost portfolio with better returns because unnecessary expenses have been stripped out [1].

On tax complexity and knowing what to ask

Kyles treats the tax code as a knowledge problem more than a technical one. He points to qualified small business stock as the case in point: for a qualifying, generally technology or innovation focused small business, a founder or early employee who files the paperwork correctly within thirty days can exempt the first $15 million from federal tax on capital gains, and this is how a great many tech entrepreneurs built their wealth [2]. It is not new. It is "buried in that 88,000 plus pages of of IRS tax code, but it's been around since 1993," and people are only now hearing about it, partly because of the bill passed recently [2].

His conclusion is that search engines do not solve this. Google will tell you about a strategy you can name, but "if you don't know even what to type or what to hear about, how are you supposed to even know that this is an option, let alone figure out how to coordinate it, um how to how to file it, the auditing behind it, the compliance" [2]. That gap between availability and awareness is the space he thinks an adviser occupies.

On aging parents, scattered families and buying back time

Kyles frames demographics as a technology problem as much as a financial one. His picture is the baby boomer parents in southern Indiana with one child in California and another in Massachusetts, and a driver's license quietly being sunsetted, where ride sharing through Uber is what actually solves getting to the grocery store [2]. The financial version of the same problem arrives when the parents are eighty-five or eighty-six, the burden of their affairs falls on children who already have their own families, careers, soccer games, gymnastics and doctor's appointments, and nobody wants to spend the day on schwab.com checking whether the account went up [2].

His answer is to move the legwork onto the advisory team: updating the estate plan, handling the paperwork, using ancillary services to get the home into a revocable trust, checking the powers of attorney, and sending a notary to the house rather than making an eighty-five year old take a fifteen dollar Uber to a bank that turns out not to have a notary on site [2]. Done properly, and reported back so the family and their other professionals stay informed, this "can not only buy back a ton of time for all generations of the family, it can also provide substantial amount of peace of mind" [2]. He adds that the technology lets him assemble the right professionals across the whole country rather than only the best available locally, and says he cannot really quantify the value of that [2].

On retirement as something other than stopping

He resists a single template for retirement. Retirement looks different for every household, and often "someone's idea of retirement is actually still working, but just doing something that they they truly enjoy," whether volunteering for no pay, coaching for less pay, taking a break to re-evaluate and spend time with family, or starting a mission driven not for profit or a business they are passionate about [1]. He connects this back to technology through professional networks: because he had been following and had met an employee advocate named Dan Goodman on LinkedIn, he could hand a client facing a deteriorating employment situation a calendar link on the spot and calm her down immediately, a connection he says would have been impossible in 2017 when Goodman would not have been marketing there [1].

On the car payment and earning the business

The ethical core of his thinking comes from advice his godfather gave him: that he is technically in sales, and "the toughest thing you're going to face is that you're going to have a car payment due in the back of your mind and a client in front of you," so you have to separate the two, because "you've got to act in their best interest" [1]. Sometimes the right answer is no, or not yet, and he accepts that the business is not always won even when it has been earned. His view is that "when you focus on doing the right thing and you have a repeatable process that is focused on the person in front of you, it it just it always ends up working out," even if the ride there is a roller coaster [1]. He cites a colleague at Farther, Anthony Danaher, who steered a prospect toward the right investment decision at a time when working together was not right, and three or four years later that person came back and became his biggest client [1].

He grounds the same principle in compensation. Because the fee comes from the client rather than from mutual fund providers offering steak dinners for volume, the incentive is straightforward: "if my team and I do a good job for you, you will continue being a client. If we don't do a good job, you'll probably go somewhere else. Right? So, we've got to earn your business every day" [2]. He is explicit that the alternative is being told to sell a portfolio because a publicly traded parent needs to hit quarterly metrics for shareholders [2]. His own choice of firm reflects the same seriousness: he interviewed at 156 institutions before deciding, on the view that where an adviser sits genuinely matters [1].

On why people resist better tools

Kyles has a working theory of adoption in four stages: something gets created, you hear about it and do not want to use it because you are comfortable knowing where to click, you are then forced into it by a mandatory update that moves the profile button from right to left, and roughly thirty days later you say you do not know how you ever lived without it [2]. He includes himself in that pattern, and says moving to Farther was a real adjustment because it meant using genuine technology rather than the kind where you log on to eighteen different screens to accomplish one thing after buying another license that was supposed to save money and did not [2].

Wealth management, in his reading, is still in the hesitant phase. Clients are shy of AI advancement, deep fakes and cyber security risk, but they are adopting anyway [2]. He also thinks the do-it-yourself instinct is stronger than it should be, noting how often he hears "I can manage it myself" when there is a reason people go to school for this and do it professionally [1].

Takeaways

• Distinguish the layers: AI is what powers a platform and makes it faster, while technology is the software or hardware that delivers the result to the client [1].
• Never delegate judgement to a model. Portfolio and insurance decisions "should rest upon a human"; AI's job is finding the inflation rider buried in fifty pages of a policy binder [2].
• The test of a plan is whether a client can ask "what if expenses rise by $750" and get an answer in the room rather than next week, because delay clouds decisions and creates emotional stress [1].
• Build the presentation for two people at once: a clear yes or no for the spouse who wants the headline, and a full printout one click away for the spouse who wants the detail [2].
• Qualified small business stock can exempt the first $15 million from federal capital gains tax if filed correctly within 30 days, and it has existed since 1993 inside 88,000 plus pages of tax code [2].
• Advisory value for aging clients is often logistical: sending a notary to the house, getting the home into a revocable trust, checking powers of attorney, and buying back time for every generation of the family [2].
• The right conflict of interest is being paid by the client, so the business has to be earned every day, rather than by fund providers offering steak dinners [2].
• Separate your own car payment from the client's need; saying no or not yet costs deals in the short run and works out over time [1].