Get the Ricochet app
Ricochet will work fine in a browser on your mobile devices, but for the best experience please download our app.
When it comes to predictions of the impact of AI, I've never held to any extreme position. But the more I use it, the more disturbing it gets. While I was an early user of ChatGPT, I really started to use AI in my business (financial advisor) in December. The use cases I'm interested in are how you use it to get your work done.
A couple of examples:
In January, we sent our clients a "1099 letter," which lists the tax forms they should expect to receive, the related account, and the triggering event. Last year, I sent a generic letter, but this year I was able to use my AI tool (Hazel) to produce a report with data that can be copied into an email. Normally, this would take hours, if not days, to gather the data, review it, and probably enter it into at least one other software program, for dozens or hundreds of households. I produced the report in one prompt.
We did another report on IRA contributions for 2025: how much the client contributed, a year-end projection of their contributions, their limit based on their age, and how much further they have to go before hitting the limit. Again, customized down to the account in one prompt. We spend more time sending the emails.
For tax planning, I can upload a pay stub to Hazel and get back a projected tax return. I also feed it tax returns to give me a summary and analysis of the key data, along with suggestions for planning discussions.
Some realizations:
Structured data appears to be unnecessary. We use CRMs, or even spreadsheets, to collect, organize, review, and analyze our data on clients/customers. AI, though, doesn't need the data formatted in a particular way.
Related to that, how you use your data will become your main preoccupation. My focus has shifted to capturing data and then thinking more about how I want to use it: new ideas for solutions; finding referral opportunities with my clients; preparing for upcoming meetings. I can also use Hazel as a kind of chief of staff, helping me create an agenda and stay focused on the right priorities.
Based on the examples above, AI is enabling personalization at scale. Tasks that would be impractical can now be done in moments.
Those who have experience and expertise are probably getting the most out of the technology. I think of it as a junior colleague, someone you train and supervise on the job. Of course, it's even better because they never show up to the office late, complain, or get tired.
From my conversations with colleagues in the industry, adoption will be lumpy and slow. My prediction: early adopters will gain an advantage by staying on the frontier of the technology's capabilities. Everyone else will use it alongside their other tech, which is the dumb way to use it. The innovative firms and individuals will reorient their organizations around the technology, the way we've all done with digital tech like email, Zoom, and other software.
What does this mean for the future?
I have no idea. I'm not ready to say work will go away. But I do think most people are closer to clueless about how disruptive the technology is. They still operate under their first impression of it as an amusing piece of software that hallucinates.
My advice?
Check out @iWe's post with a great article by Matt Shumer. He recommends using it now and getting familiar with the technology. That strikes me as prudent. In content creation, AI has been disruptive, but I also see new opportunities for creators. Hopefully, that will be the case elsewhere: the creation of more supply and demand, more jobs (especially ones that never existed), and more opportunities.
Discussion (27)
CONTRIBUTOR • CHARTER MEMBER
To read this bio, please join Ricochet.
Great post, Bereket. Thanks!
BTW, I think it's been a while. Stop by more often!
To read this bio, please join Ricochet.
CHARTER MEMBER
To read this bio, please join Ricochet.
One way it helps in our business is that clients (who heretofore struggled to describe what they wanted from us) can send a visual approximation of what they're looking for in terms of design. That helps in terms of keeping clients happy, but also saves time in not having to produce additional concepts that are rejected.
Another way it helps is when cropped images need to be expanded to fit certain spaces. AI tools allow you to expand an image without losing resolution. Before you might have to scrap the image and use something less ideal because it could be cropped to fit the space.
I had a client who used AI to create an outline for a whitepaper. My work was to write based on the outline, but I found it repetitive and vague when it came to key concepts, and so still spent hours at research before starting to write, though likely a couple hours fewer. The outline did contain links to sources, though some didn't seem to exist and others were behind paywalls. The owners of that content might be concerned.
All of this jibes with what Bereket says above, but I don't believe it conforms to Matt Shumer's views. There are others who describe his piece as "hype" and some took the time to present what the current state of AI is.
As always, a tool is most useful when the wielder knows what to do with it.
CHARTER MEMBER
To read this bio, please join Ricochet.
What he said.
CHARTER MEMBER
To read this bio, please join Ricochet.
Yup.
Always a pleasure, Bereket.
You're seeing five of 27 comments. Sign in, or join for just $5 per month, to read them all and add your own.