I delivered a speech at church today where I identified "Christmas" as being "Christ" and "Mass", both meanings somewhat lost in pronunciation -- similar to "breakfast" where we say "brek" and "fist" instead of "break" and "fast". Sometimes it helps to break things down to preserve original meaning!
I'm starting to settle on an idea I had a long time ago where I interviewed an international violinist on controlling nerves. This is a massively important topic for any top performers and I feel like it should be talked about at that level so aspiring artists can learn earlier.
Didn't realize how much good sneakers cost! I remember when they were like $50... now they're three times that! Yet, I still bought a pair because I'll use them for years and am getting older. Sometimes it's worth the price for comfort/protection!
Agonized over iPhone 17 vs Air for months. My beloved 13 mini is dying a slow death. Thought the Air would be perfect. Kept asking GPT repeatedly to help surface the pros/cons.
The ultimate cons of Air: no USB-C Display Port, too big of a screen, $200 too high. Thanks for the assist, GPT!
Worked with AI on a talk for church. Told AI explicitly to cite sources and provide quotations in an outline without extrapolation. I need to own the words in the end, so it's only allowed to help me research. This ensures I stay cognitively sharp.
Drove 6 hours today. For many, this is just an average day. For me, it was like 12 times my usual commute.
No wonder people drive fast.
Met a man today who had lots of interesting personal stories to tell, particularly around running his small business. He gave me a specific point of advice that gave me some food for thought: fractional CTOs charge by the hour; businessmen generate passive income.
This has basically turned into a public diary. Not sure anyone will follow at this point. Not sure that's the point anymore. Honestly, reading my posts back is helping me become more efficient and self-aware. So I'll just keep doing it for now -- if anything, it's good practice!
My throat is finally starting to show signs of improvement after 15 grueling days of coughing and wheezing. I stopped using Niquil last night and got through my first unassisted sleep in more than a week. Funny how bodies heal. Glad they do.
I started learning Mandarin using Spanish as my base language, and with AI. So far it's going well -- it can handle basic grammar pretty well, although it often switches to Mandarin explanations instead of Spanish by accident. Some fine-tuning required to get it to stay consistent.
I worked with GPT and Claude to draft up my first internal version of the YY Method. Very satisfying to see how much is being incorporated into it.
Now I have to slash it down to nice, cut the fat, and rewrite entirely in my voice. Simple, right?
My Spanish is kinda good. I can guess a lot of words from French. But My spelling? Atrocious. Impeccable in English, usually. Quite accurate in French. But I never took classes in Spanish.
AI is helping me straighten this out, faster than any tutor. For free, too, in measured doses. Super cool.
I learned today that getting from B1 to B2 in CEFR standards for a language is more about taking risks than being 100% correct -- counterintuitive to the school system. I'll keep doing it. Current targets for solid B2 are Mandarin and Spanish.
Wife got called to jury duty, so I've been doing double duty with the kids while working. Reminds me how much she does -- definitely appreciate her more!
Playing around with learning frameworks in my handwritten journal. Latest idea revolves around language acquisition using multiple AIs to triage corrections. This way, I focus on fixing known, falsifiable errors en masse. French, Spanish, Mandarin, German, and Russian are currently in play.
I rewrote one of my business profiles as a high bar net for unicorn companies in crisis, following a personal framework I've been intentionally living for years. All major life changes up to this point are thanks to this framework.
Making the bet that this one will work, too.
Knocked flat by a virus this weekend. Benadryl + Zyrtec to the rescue.
Firefight against an invisible building aflame, always home court disadvantage.
Sometimes using AI feels like this.
Other times, the roles are inverted.
Depends on how it's used -- and where our head's at.
Published my first raw essay on Substack, written without any AI assistance whatsoever. Nerve-wracking, but I'm finally doing it. Training wheels off.
benchanviolin.substack.com/p/scars-visi...
Big leap of faith: speaking 4 languages out loud for 40 seconds each as a baseline. Excited to see my progress over time!
40spolyglot.substack.com/p/french-5
40spolyglot.substack.com/p/spanish-5
40spolyglot.substack.com/p/mandarin-5
40spolyglot.substack.com/p/german-5
Day 2 of AI detox. Wrote an essay about paying attention. Felt good to just let loose.
Detox from AI as writing assistant begins today. Thirty days to recover my natural prose. A very intense, interesting, and somewhat terrifying experiment -- almost lost my voice.
Yet, there will be a Round 2 afterwards with AI. But I'll be better prepared, fresh. Editor only; I retain final say.
Major system meltdown tonight during a very busy set of events. Couldn't figure out the issue until I used GPT to triage the symptoms. Managed to grab a server output that pinpointed the issue. Taught myself how to edit a system file and reboot safely. Notified the main dev team to autopsy later.
A new dev received production server access today and asked if I could review a PR. With halloween tonight, my kids need my time. I gave a green light to push and test the changes directly. Fingers crossed!
Debated a bit with the team on some KPI definitions. My code was suspected but I cleared the air quickly, and each of us took turns explaining our positions. It's nice when there's mutual friction -- in the end, we settled on a blended approach that meets all of our needs. Teamwork!
A colleague and I were trying to figure out how to add tags to all products. The system offered "tagged with" but we needed "not tagged with".
We thought about writing an API to do it... then I realized we could just tag all products, then filter to a tag on the ones we didn't want.
API avoided!
Dealt with a difficult situation today. One vendor caused a bug in another vendor's system that broke a product sync. It was firing twice and they couldn't figure out where the second sync was coming from -- they were sure it wasn't their fault.
It was. Overconfidence early can backfire sometimes!
Settling on the new project, finally:
Teaching Americans to write Chinese characters by thinking "radically" differently. Starting with my kids.
Headed back to YouTube where I gained my original fame (BenChanViolin ~16M views). Doubles as hobby time reinforcing my own Mandarin.
Can't wait!
Finally learned about GPG today.
Here's my fingerprint: 63FC60BA7F66F645CFBF9C8482BBD72944B8168E
I'm getting ready to start a new legacy project for my kids that I hope will gain traction slowly and deliberately with high signal people. This is my post to myself time stamping the official start, and a way to keep myself accountable and motivated.
If my kids use it, I've already won.
There is a lot of HATE around AI right now, especially GPT-5.
I don't think I'm ready to debate. I'll just stay in observation mode, for now.