Skip to content

The Wonderful, Scary Next Stage of the Long Tech Evolution 

When you’re nearly as old as Methuselah, you start to see some patterns over time.  I remember when software was first introduced into the financial planning space, first as a spreadsheet called Visicalc, later Lotus 123, then Excel, and some financial planning calculators were built on top of them that looked a lot like the planning programs of today.  

What was the reaction?  A lot of hand-wringing, initially, because many advisors questioned their value working the good ‘ol HP 12C making retirement planning calculations for their clients when this newfangled thing called software came along.  Is this going to replace us?

The other reaction was an eager embrace by salespeople who wanted to borrow the credibility of these allegedly-impartial calculation engines.  I remember a Profiles planning program demonstration in the mid-1980s, where the person who showed it to me gleefully input a variety of client situations, and the program reliably spit out a recommendation that the client buy a whole lot of life insurance.  Since it was coming from a (theoretically impartial) computer, the conclusion became all that much more convincing.

There was a similar crisis of confidence when the online calculators began to appear on this new thing called the Internet, and the interesting thing about them was that they contained some of the same biases.  Mutual fund companies, particularly those that paid commissions, would recommend their own products, and all of them recommended savings rates beyond what would actually be needed.  (More money in our funds!)  Morningstar, alas, joined this biased analysis trend; when I was helping them set up their advisor website back around 2000, some of the salespeople gleefully reported that they had landed Merrill Lynch as a big customer.  All Merrill had asked for was that whatever portfolio the software recommended absolutely must include an allocation to one or more of Merrill’s in-house funds.

I think most of us younger than Methuselah remember the online portfolio management tools, called robos, and there was another round of widespread dismay at the idea that these automated websites, that charged just 25 basis points, would undercut and replace financial planners.  I interviewed the people who started Wealthfront and Betterment for an article about what their programs did and did not do, and they told me right out front that their goal was to replace those expensive, conflicted financial planners.

The common theme of these innovations is that they did not, last time I checked, replace financial planners as the primary providers of financial advice.  Instead, they became tools that made financial planners more efficient.  But they also changed the profession.  With the advent of PLANMAN, IFDS and the Leonard System (and even Profiles), advisors began delivering large, comprehensive financial plans to their clients, and there was a lot more modeling of net worth and its growth over time, which gradually led to more comprehensive advice.  Turns out you could do a lot more with a customized spreadsheet than you could with a hand-held calculator.

The online tools forced financial planners to offer more customized, ongoing advice, which turned the delivery of a financial plan into ongoing financial planning advice.  The robos forced financial planners to move off of asset management as their primary value proposition, and we are still in that evolution today, which I think will eventually move the profession off of the AUM revenue model toward what every other profession charges: fees that pay for advice.

Where am I going with this?  At the recent T3 conference, a number of fintech firms were cautiously incorporating artificial intelligence into their software, and a few were predicting that this would change the way financial planners managed their businesses and delivered advice.  Of course, I had been hearing something like that for roughly a decade without seeing any tangible evidence of it; now, suddenly, the evidence of actual (albeit cautious) incorporation of AI into software was visible pretty much wherever you looked.

It’s somewhat encouraging that, with this next innovation in our space, nobody seems to be wringing their hands or predicting that the profession is doomed to extinction.  Maybe we’ve learned that lesson.

But what we don’t yet know is how AI is going to change the profession the way every other significant tech innovation has done.  At T3, I saw AI elements incorporated into financial planning, into marketing, into practice management and business tools.  The easy prediction is that advisors will be able to do more with less as their computers get smarter.  They’ll know more about clients, they’ll be able to delve more deeply, more easily, into their circumstances.  Time that is currently being spent on back office chores and tedious data entry will be redeployed to coaching clients on life goals and understanding the unique challenges of niche specialty clients.  Back office staff will be repurposed to client-facing (client relationship management) duties, and the costs of doing business will go down.

But a more fundamental change might be happening below the surface.  One of the use cases that was not talked about in the sessions, but did come up in the conversations at the T3 booths, was how AI can transform the user experience of the increasingly complex software tools in the advisor ecosystem.  

Over the years, the most constant complaint I would hear from software vendors was that their advisor users were only accessing half or less of the capabilities their software was offering.  This has had the pernicious effect of making them cautious about adding new features, even features that advisory firms were asking for, because they were nervous about overloading their advisor customer with more complexity.

We are moving toward a time, still in its early stages, when the software will ‘notice’ when the user is doing something by hand that could be done more easily through the program’s feature set.  Or when advisors can ask the software how to make a certain thing happen—and if the AI bot embedded in the software doesn’t have an answer, that question will be aggregated with thousands of others, and sent back to the development team, which will be able to prioritize new features based on a rank ordering of these questions.  Fintech firms will be more willing to add these new features because they will be more easily understood and accessed.

Back in the late 1990s, I read an article in the trade press which envisioned the future of advisor fintech.  It was purely science fiction at the time, but it started with an advisor walking into the office and telling the computer which programs to turn on, and asking which client meetings were up for the day, and having a centralized voice-activated program initiate task sequences entirely through the software stack that would prepare all the information needed for those meetings.

Wow!  Wouldn’t that be incredible?  Well, if the conversations I had with T3 vendors is any indication, we’re moving in exactly that direction.  

Of course, there is always something to worry about.  The long fintech evolution from the hand-held calculator to financial planning software to increasingly sophisticated tech stacks has driven a trend where the users have less and less understanding of how the calculations and assessments are made ‘under the hood.’  Today’s financial technology has become increasingly black-box-ish as it becomes more complicated and sophisticated.  When we have AI interfaces where we simply ask for solutions instead of make them happen with our hands and our minds, when vendors will finally be freed to pile on the features and calculations inside the boxes, the trend will have completed itself, and the built-in assumptions and reasoning inside our fintech solutions will have become entirely opaque.  

AI won’t replace financial planners.  It will make their lives easier and enhance the services they provide to their clients, as every tech advancement has always done in the past.  I’m looking forward to seeing how the ability to think and anticipate will be built into the programs we use today, and I’m also planning to question whether advisors understand the silicon-based thinking that they’re relying on to make these wonderful things happen.