Dabbling in more AI as a non-techie
asdadas
Happy Monday folks,
Another fun week in the tech world is in the books, and as I dig deeper into AI, it gets more interesting and exciting.
Let’s jump in:
Story: building custom AI algorithms as a non-techie
This week I got to dig my teeth into some of the AI we’re working on at Visto.ai (for those who don’t know, that’s my immigration tech company). In short, we make it quick and easy for Canadian immigration professionals to prep applications for their clients.
One of our focuses for 2024 is adding in more AI, and while we have software engineers working on the tech itself, as the resident lawyer, I am still heavily involved in testing the end product and discussing how we design it, improve it, etc.
Won’t go into full detail on what the end product is itself yet - maybe in a few weeks closer to launch - but for now I can share that we’re doing some cool work building our own Canadian immigration algorithm.
And as a tech noobie, I gotta say, it’s quite fun and also harder than expected.
Mainly because I think we’ve been pretty spoiled over the years with all of this no-code and out-of-the-box type software that let’s anyone and their mother build websites, platforms, apps and more, sometimes without even needing to code.
When it came to AI I (mistakenly) thought it would be as easy as uploading some text files to ChatGPT and letting Sam Altman and co. do the rest.
Boy was I wrong.
Now, if you’re just looking for a basic model and not too concerned over nuances and accuracy then you could probably get away with that. In fact, I’ve done some tinkering with the custom GPT builder and Open AI assistant builder. It’s pretty cool stuff and they make it quite easy.
But alas, we are working on an in-depth, extremely reliable, legal-centric algorithm that can be trusted by lawyers and other professionals in Canada (who, by the way, are pretty picky folks and require extreme accuracy).
Probably the most interesting part so far has been trying to design how we store and categorize large sets of data. For example, imagine you had 500 different sets of data/information and you needed to train an algorithm on how to sort through it and find the right answer from a question asked - quickly and reliably.
How would you sort that data?
I know this is very high-level since I’m not going into detail on the product, but as a non-technical guy, it’s been fun to explore and help solve. We’re getting close to being ready to beta test with some of our clients too, which is exciting.
Stay tuned, it could be a big product launch for us…
Key takeaway: “if a problem can be solved with money, you don’t have a problem, you have an expense”
Now here’s something else we realized, and something I re-learned (as is the case with many business lessons).
We were spending a lot of time trying to design the perfect algorithm, but realized that, as with most things, there are people out there far smarter than you who can probably quickly provide you with a good solution.
So we quickly booked a consultation with an AI expert who is now doing some coaching work to help us work through any more challenges we have.
Most of us think what we’re working on is super special and unique, when in reality, there are probably tons of people out there who have the knowledge and experience to walk you through any difficulties far quicker than you could get through on your own.
As they say: “if a problem can be solved with money, you don’t have a problem, you have an expense.”
So not only are we continuing to work hard to get this new feature built, we’ve brought on an expert to help us shortcut the process with their added expertise.
If you’re struggling with something in your business, or starting your business, ask yourself: is there someone or something I can pay for to help me shortcut this problem?
I’m not saying every problem can and should be solved by money, or maybe you’re not in the financial position to do so, but don’t forget it’s an option.
Content for the week:
Vlog: I’m continuing with my cadence of a weekly Substack post and also a vlod/podcast too. You probably got the notification for the podcast episode yesterday (which is just the audio of my vlog), but if you want to watch it on video, check it out here:
Book(s): Still not done Onward: How Starbucks Fought for Its Life without Losing Its Soul, although I’m getting close. I’d say about 70% of the way through.
Continues to be very entertaining, especially in learning what goes on behind the scenes at a huge company, and I is definitely a book I’d recommend reading. Due to the topic, it would especially be valuable to someone trying to turn a company around. Hopefully full review next week.
Tiktok: my most viewed Tiktok vid from last week was this clip about how we launch new tech features to our users at Visto.
Thanks as always for following along on my journey and have an awesome week,
- Josh Schachnow
Canadian immigration lawyer, CEO at Visto.ai
If you’re looking for more of my help, you can check out:
My immigration tech platform, Visto, that helps Canadian immigration firms prep and file applications much quicker.
Make sure you’re subscribed on Youtube where I post the weekly vlog and am also posting shorter snippets of my tips/advice as Youtube Shorts!