By far, my favorite pattern of programming with an AI is the Ping Pong Pattern. I have it write several tests based on a description. I use that to TDD some methods, and boom, we're done.

Putting my money where my mouth is. 90% of the research I did for our new HVAC unit and mini split was done in ChatGPT. It was a blend of information and consideration questions. And this is why we spent so much time building out categorization in Knowatoa.

Going to fill in for Mike on the Twinning Strategy pod. You can see us live today at 3pm on YouTube or LinkedIn. .

Traveled back in time for spring break

Why does it sound like noise when my kids play it versus music when I play the same thing. Am I old now?.

The podcast tour begins. Thanks for hosting @brendanwb!.

There's all this debate on whether AI generated code should be used or not. I look at it as another tool. I'm giving our junior marketer access to our backroom to retrieve some metrics. The report doesn't have to be 100% bug proof or elegant, it just has work.

Monthly occurrence of a one character update, causing me to wonder why something isn't working for a couple of days. I've have the unit tests for this all stubbed and mocked out. Maybe I should have reached for VCR.

Just updated this on the pricing page. It's not a feature in the traditional sense of the SaaS but our customers have been happy with it. We're in a shipping cycle so expect a number of new features in the next couple of days/weeks.

3. Any research involving the traditional search journey goes to ChatGPT. This mostly involves things that I would never click on "I'm feeling lucky" for. Because I'm still trying to zero in exactly what I'm searching for. So again the interactive chat UX is superior here.

2. Anything that involves reading documentation goes to ChatGPT (Curl, Stripe, Rails, etc). The chat UX just parses out all the noise and gets me exactly what I want faster.

There's an argument of whether AI search will be the death of Google Search. Looking at my search behavior over the last 24 hours I see a several patterns:. 1. Instant searches are still won by Google (images, spelling, who a person is).

Haven't had time to fill out my March Madness bracket and it's getting down to the wire. So I told ChatGPT everything I knew about winning basketball and had it use deep research to make the picks for me. Will report back in a couple of days to share how we did.

The good: I’ve been successfully pushing all mosquito tasks to the end of the day and getting more flow work done. The bad: I was so deep in the flow this morning I didn’t see any notifications and showed up 10 min late to a support call.

My auto indent wasn't working in Neovim this morning. Rather than yak shaving, I pointed ChatGPT's deep research to my public dotfiles to see if it could figure it out while I stayed in my flow. Basically delegating to a junior dev.

You want them reading case studies with graphs pointing up and to the right. Or datasheets describing their market and why your services fit their needs. You want to answer every possible question that is paralyzing them from making a decision.

As much as we roll our eyes at the phrase “sales enablement content” that’s what your prospects need to see. Plus, AIOs are going to crush you at top of the funnel so follow the money.

If you’re selling a prospect on AI search optimization you’re not going to show them a research paper on how the different AI’s work. Their eyes are going to glass over as their tummies digest Jimmy John’s. 🥱.

We've been advising people to focus on bottom of the funnel content. What does this actually mean?.

In some of our Knowatoa meetings we’re hearing about Reddit posts impacting AI search results for their company. Even posts with very little upvotes and comments.