Yeah shit, I don't know which is worse. My plan is just dying before I reach that stage.

Opus is obviously the better model, but Cursor's "harness" is doing so much heavy lifting in terms of just magically supplying the broader context the model needs to understand the ramifications of its edits.

Artistic oddities aside (why are the 8-bit sprites 16-bit, why do the charcoal drawings have colour, why does the art of specifically the Gen 1 Pokemon look so off.), 271 is Lombre, not Lotad.

I've recently come to the conclusion that even spending $100/mo on AI is basically peanuts for what we get out of it, and moved a lot of what I do to Opus in Cursor (despite the cost) over and above Claude Code because, despite costing 5x the price, it's more thorough.

I've seen CompSci guys especially (I'm EEE background, we have our own problems but this ain't one of them) launch conceptual complexity into the stratosphere just so that they could avoid writing two separate functions that do similar things.

- Write a correct, pretty implementation

- Beat Claude Code with a stick for 20 minutes until it generated a fragile, unmaintainable mess that still happened to produce the same result but in 300ms rather than 2500ms. (In this step, explicitly prompting it to test rather than just philosophising gets you really far)

- Pull across the concepts and timesaves from Claude's mess into the pretty code.

Seriously, these new models are actually really good at reasoning about performance and knowing alternative solutions or libraries that you might have only just discovered yourself.

In the case of all of them, they may make some questionably ethical business decisions but at the same time do genuinely care about the craft they're in, pushing boundaries and making quality products.

That's not too say they're not helicoptered too, but Chinese parents are a whole other level.

Think strapping a 3-4 year old into a high chair and handfeeding them, or scheduling every waking moment of a primary school kid's life.

There has long been a culture of deliberately making the installation of certain types of free and libre software needlessly complex and using deviancy-coded language simply because it makes the in-group feel cool and elite.

This whole idea of "sideloading" and related terminology being Google FUD only came about in the past couple of years. For the decade before it was people on xda-developers deliberately throwing words like that around because they wanted to prove they were true 1337 h4xx0rz.

(Actually, anyone else from BMD here? Was that the product that the Industrial Designers won second place in the design awards for, losing out to the accessible playground?)

Used to have lunch regularly with one of the owners too. Need to check in with him again!

At least back in 2019, BMD made a lot of money selling professional licences for DaVinci Resolve. I don't know exact figures but that part of the business was healthily profitable of its own accord. Very, very healthily profitable!

Most parts of the business were profitable standalone, AFAIK. Their model didn't revolve around loss leaders, burning VC money or anything like that; just selling good products at fair prices and making bank.

I think a big part of it was a fairly lean culture (whole company was bootstrapped and grown sustainably), and specifically in the case of DaVinci they bought out an existing business that had already done a lot of the development and marketing work for absolute peanuts.

Very smart team doing good work.