Matt BirchlerModify

2025-08-22 05:37:11

Let’s see what all the hype is about.

2025-08-20 08:58:29

Craig Grannell writing for Wired: Apple Finally Destroyed Steve Jobs’ Vision of the iPad. Good

For years, Apple treated the idea of windows on the iPad as sacrilege. But with iPadOS 26 installed, today’s iPads are doing macOS cosplay, becoming touchscreen Macs in all but name. And here’s the thing: It’s actually pretty good.

I bet if you travelled back in time with iOS 26 and showed someone from 2015, an iPad fanboy would tell you that’s a terrible idea and it was clearly made by someone who doesn’t “get” the iPad. “Freeform windows? A menu bar? Mouse support? Desktop mode? A file system? These are relics of the past, man, if you want these things just get a Mac. And what about that Magic Keyboard that turns it into a laptop? Apple tested that with users and they hated it!”

I’ve documented my things that keep me on a Mac, but all of those gaps are functional, not fundamental things the iPad can’t do one day. I’ve also documented how you really need to pixel peep to find the UI differences between apps on the Mac and iPad these days.

Again, there are still gaps, but iPadOS 26 is filled to gills with features that older iPad fans would have told you were Mac things the iPad didn’t need. And you know what, the iPad is a better platform for more people because of it. I'm just saying that while we've been bickering about why all these things are "unique to the Mac" or vice versa, Apple's just been out here turning the iPad into a Mac anyway.

2025-08-20 08:50:09

Ethan Ding: Tokens Are Getting More Expensive

every ai company knows usage-based pricing would save them. they also know it would kill them. while you're being responsible with $0.01/1k tokens, your vc-funded competitor offers unlimited for $20/month.

i keep seeing founders point to the "models will be 10x cheaper next year!" like it’s a life raft. sure. and your users will expect 20x more from them. the goalpost is sprinting away from you.

the 10x cost reduction is real, but only for models that might as well be running on a commodore 64.

And:

so this is the first faulty pillar of the “costs will drop” strategy: demand exists for "the best language model," period. and the best model always costs about the same, because that's what the edge of inference costs today.

This link post is a good opportunity to state a few things that have been on my mind lately.

1. Token prices have come down considerably in the past year or two, however, we're also generating a lot more tokens, so the actual costs aren't dropping like I hoped they would. In fact, "cost per token" isn't the most useful metric for measuring LLM costs these days. For example, Grok 4 is cheaper that Claude Sonnet 4, but it's way, way cheaper to actually use Sonnet because Grok uses at ton of tokens to do anything.
2. Thinking models, which have kicked off since DeepSeek R1 came out earlier this year, have driven much of the increase in performance of newer models, but they also by their nature generate a lot more tokens to deliver an answer.
3. Despite the "no one wants these" complaints I sometimes see, user engagement is continuing to rise, and power users are asking their LLM-powered tools to do more than they used to.
4. The cost of no-longer-cutting-edge models drops once their replacements are out, but very few people keep using them. I think this speaks to the simple fact that today's LLMs are are barely good enough and we will always use the latest and greatest just to get something a bit better. I do wonder if we're finally hitting a "good enough" spot for frontier models, as GPT-4o is still preferred by many over GPT-5 and I know several developers who still use Gemini 2.0 Flash because it's hella cheap and totally sufficient for most non-code tasks.
5. High quality, local processing can't get here soon enough. These costs are unsustainable, and while pay-per-use is the only real sustainable pricing plan for a lot of these LLM apps, consumers have shown time and time again that they would much rather pay a fixed monthly fee, even if that means they sometimes overpay. I don't see how LLM-powered tools succeed in the consumer space with that model.

2025-08-20 05:00:48

2025-08-17 22:30:37

2025-08-16 04:13:04

They went hard on this one.