55 66 88

Please don't get mad or annoyed, but we're graduating our most recent pre-release 55 66 88 into a new product. It's launching on Kickstarter on Tuesday, and we'd love your eyeballs on the campaign before we go live. And of course your support Tuesday once we do.

55 66 88 preview on Kickstarter

We've never made a custom aluminum extrusion, and we're excited to see how that goes. Our plan is to work with an extruder in New Hampshire, which hopefully (fingers crossed) means we can go there and take some video as they squish hot aluminum through a custom mold into our cool profile.

Astrology.

Sometimes Che-Wei makes fun of how I use google to resolve day to day disputes. He thinks I'm not an objective Googler. "You can find anything on the Internet to support your claim. If you want to find the answer you are looking for, you'll find it."

Whatever, I think my googling is just fine. Arms cross foot stomp. I obviously can't argue with him. And I know there are millions of other people out there who use the internet like I do, like a human.

That said, I can't help thinking about AI in these terms. And how it reinforces our propensity to seek meaning, irrespective of truth.

This past rainy Saturday I came across a headline that made me laugh on a 2-page spread on AI in MIT Technology review. "Exquisite bullshitters" it read. The spread gave a quick history of humans being duped by personified intelligence, beginning with a description of early Automatas pausing to "think", which was really a cleverly choreographed pause to allow cams to shift position.

Che-Wei, is a brave and adventurous adopter of new tools. Me, not so much. But it's been fun to sideline, late night on the couch, peek over at his laptop and see a chat GPT window open. I perk to it like a 5 year old hearing the rustling of a snack wrapper. Whatcha doing with it now?

One time Che-Wei asked Chat GPT about CW&T. And since we're embarking on new efforts to run more of a proper business (more on that another time), he asked the AI to run a SWOT analysis on our company. I took a look at it and bleh, yeah, sure a lot of stuff in there was probably true. But it could have been written about a gagillion other companies.

This is how I mostly feel when I've used AI, it's like reading a horoscope. Not necessarily true, or helpful, but it can be a guide towards noticing something you didn't before. Or wildly exciting when we form connections with what we believe to be true.

The most interesting time I've seen Che-Wei use AI, was while 3D modeling. He was working on a thing, and needed to figure out the length of a flexure made out of titanium. Usually, he'd make an educated guess based on past experience (which he has a lot of). I don't remember the actual terms, but he asked Chat GPT to take into account the thickness + width of the material and the amount it needs bend. He wanted to know how long the piece has to be to bend without permanently deforming.

After some back and forth, Chat GPT came back with a range. CW tested the the dimensions in a FEA simulation to make sure it actually worked. It checked out. And yes, he also made the part out of titanium and it actually works.

We have no visibility into how Chat GPT arrived at that conclusion, and that matters for so many reasons. But what matters right here, is that it was a tool to get unstuck, to give that extra nudge of confidence to move forward.

Phone furniture.

We were brainstorming fun video ideas for our Kickstarter and I really needed to see a bunch of phones from the late 90s / early 00s lounging on a red velvet couch with gold trim. Although too zoomed in, stable diffusion really did the trick.