Happy Sunday morning, everyone. Oh, how I am overjoyed that Sunday has arrived — it’s been a week for the ages. You know how some life events just stick with you? Either for their life lessons or for their impacts on your future? This was one of those weeks. Perhaps most importantly for me this week was the thought of handling these types of career and family events with a growing faith. There is no better tool in the toolkit for handling life than having a faith in a gracious God.
Occasionally, I wonder if life events make you a better person, though. My impression is that there’s this default idea that all hardship makes you a stronger, better person. And we’re not really allowed to question that, either. But what about those events where you get burned and vow to never get burned again?
More and more, I feel like certain life events make you not only harder, but also a seemingly less compassionate person. A worse person, by what, I feel, are human standards. A less trusting person. A less empathetic person. I don’t feel these life events always make someone a better person, by default.
Because of the week I will be a little shorter than normal this week.
How I Tend to Use AI
Another week, another quick commentary on a Matt Birchler post. (A quick aside, head over to Birchtree.me for an app-like website redesign. I love it.) I promised last week I’d dive in a little deeper on how I’m using AI. Here are three quick ways I’m using AI and how I’m using them more and more each day.
ChatGPT — This has become my main daily chatbot. I have the desktop app installed on my Mac and use it increasingly for more advanced “Google searches” — questions I feel might be answered across multiple webpages or questions which may require a little more nuance than a single-viewpoint webpage might provide.
But more importantly, I’m finding myself using custom GPTs between 10 and 20 times a day right now. There are custom GPTs built to answer questions on Canadian tax or accounting practices, which have become invaluable for answering general inquiry client emails during tax season. They’re also great for ensuring I have a complete list of documentation for a specific deduction or claim. This has made very short work of emails that used to take me 20 or 30 minutes to research and write.
The custom GPTs do get questions wrong, though. I was showing my business partner the Canadian tax custom GPT, and he typed in two or three queries, and each came back with a mistake (not a good first impression at all!). One pretty in-depth query I asked about stop-loss rules resulted in an exactly wrong answer, which I only knew was wrong because I took a lesson on the topic about a month ago. This has certainly entrenched the belief that you need to know what you’re doing first before using these AI tools to do your job; that AI tools are not going to replace a studied brain tomorrow. But to have this capability to communicate with a client at this speed is truly invaluable.
NotebookLM — While I’m learning that ChatGPT can handle documents, I have found NotebookLM to be better suited for handling multiple sources around a particular topic. There are two ways I find myself using NotebookLM:
- For limiting the scope of what I want the AI bot to have access to. By this, I mean that I want the bot to look at only the source documents I provide it, not at the general internet at large. This is good for proofreading papers, research memos, and more.
- For consistent additions over a longer period of time. I really liked Federico Viticci’s idea where he consistently updates one of his notebooks in NotebookLM to always know what he’s written about for his website. I’ve tried this for keeping track of links I’ve shared here for The Sunday Edition.
NotebookLM was the first real AI tool I started to use regularly, and I use it most often for proofreading papers and memos and verifying logical holes in my writing relative to the source documents I’m writing about. Most of my general queries and quick email compositions come out of ChatGPT, while the heavier, lengthier pieces find their way through NotebookLM.
Update: After some research last night, I expect to use ChatGPT for this more going forward.
Superhuman AI — I don’t get 200+ emails a day the way many folks are in the tech industry. I’m an accountant — I receive 30-40 actionable emails a day. At this time of year, each email I receive is actionable for about 30 minutes, so a list of 75 emails is effectively 37.5 hours of work. Anything I can do to cut down on that time requirement is going to save money in the long run.
Superhuman’s AI — at least at this point because it’s only slated to get better — for quick canned responses that I can click two buttons and respond is really helpful. My favourite part about Superhuman AI is how you can set some parameters in the settings by describing who you are, what line of work you’re in, and more. I’ve found the generated email responses to be more in my voice than any other AI product I’ve used so far as a result.
I do believe that accounting is particularly susceptible to the impacts of AI. The profession is old, with readily available information online and, frankly, not too many new ideas. The key to being a great accountant is to know how the rules apply to a situation and to advise a client on how best to operate within those rules. There’s room for creativity, but not the kind many folks think of when they think of an “artist”.
So, I think we should embrace it. Individuals without the knowledge can, but shouldn’t, hammer their queries into an AI bot and trust everything that comes out. Professionals, though, should supplement their knowledge and utilize the efficiencies found in AI to improve the number of clients they can service and improve the depth of the information they analyze and provide.
Minimum Levels of Stress
It wouldn’t be a complete week without a Morgan Housel article:
If that sounds like wishful thinking to you, let me propose a reason why: Part of the reason today’s world is so petty and angry is because life is currently pretty good for a lot of people.
There are no domestic wars.
Unemployment is low.
Household wealth is at an all-time high.
Innovation is astounding.
It’s far from perfect, and even an optimist could list hundreds of problems and injustices. A pessimist could do worse.
But let me put it this way: As the world improves, our threshold for complaining drops.
In the absence of big problems, people shift their worries to smaller ones. In the absence of small problems, they focus on petty or even imaginary ones.
Most people – and definitely society as a whole – seem to have a minimum level of stress. They will never be fully at ease because after solving every problem the gaze of their anxiety shifts to the next problem, no matter how trivial it is relative to previous ones.
I read a tweet thread at one point (I can’t seem to find it right now) about a person born in 1905. That person would have likely grown up in squalor. They would have been 10 or so years old when World War I broke out. They would have been 18 years old in the Roaring 20s, only for their lives to be ripped out from underneath them right when they got going in the 1930s. They may have had to fight in World War II, then seen the Korean War, the Vietnam War, and lived through the Cold War. There were also world pandemics in this timeframe, such as the Spanish flu in 1918. And wealth? Come on. There wasn’t anything left after food was on the table.
Let’s stop pretending we live in terrible times. There are struggles, yes. But to pretend like we have it bad runs the risk of extreme arrogance.
The Second Cup
- My dream desk | vitsoe, ikea, new layout — (Georgene Loh)
– Unique setup. I think I’d get tired of moving my display each time I wanted to use it. - Andor | Season 2 Trailer | Streaming April 22 on Disney+ — (Star Wars)
– Andor is one of the few bright spots in Disney’s Star Wars. I’m hoping they stick to Star Wars in season 2. - Apple’s BEST keyboard… reimagined — (mtt)
– The absolute creativity to make this happen is mind-bending. How cool is this layout, too? Look at that weird “return” key. - Where’s the iMac love? — (trevorwelsh)
– Wonderful photography. There’s something about bright sunlight on wood grains in a photo. - The Day You Became A Better Writer — (Dilbert.Blog)
– My favourite quote: “Don’t write, “He was very happy” when you can write “He was happy.” You think the word “very” adds something. It doesn’t. Prune your sentences.”
Happy Sunday. I hope you have a wonderful week ahead.