runtime_terror

Let us hope it was mentions about how idiotic of a book it is...

M5 pro Mac mini with 64gb is $3k

Must have missed that cause I didn't see that text anywhere

Using Tesla as a source is hilarious. They don't even report accidents with proper FSD attributions to the government July 2026 was teslas worst month on record: https://electrek.co/2026/09/15/tesla-four-fatal-driver-assis...

https://www.iseecars.com/most-dangerous-cars-study

https://www.informedforlife.com/tesla-safety

https://static.nhtsa.gov/odi/inv/2022/INCR-EA22002-14496.pdf

https://www.tesladeaths.com

It's not being jaded, it's being realistic based on past trends and the current admin being 100% ok with padding corporate pockets instead of the working class

Into the back of a parked semi truck

I believe you're conflating systematic racism with racism more broadly. Anyone can be racist.

Gotta love the kraken destroying the city in the background of the Hong Kong lovers motorcycle scene despite nothing in the prompt about monsters. Filmmakers are still safe.

Mostly using OpenRouter hosted GLM Flash Latest

"They said it refills itself" and you believe their word because it's been so reliable?

Classically conflating weather with climate

Says the person whose obviously never been to a city with functional public transit

Gets me about a day or two lol

Anecdotal but I was doing regular adversarial reviews of new work using about 5 models (mostly the open Chinese ones like DS, GLM, Qwen, Kimi, etc) and GLM routinely took much longer for similar results and the same instructions.

It would just loop on things and go down rabbit holes more. DS flash has been so consistently good I've kinda settled on it for 90% of my work.

I still use adversarial review for bigger changes where I want to make sure any issues are caught early, but I genuinely feel as tho if no new models ever get made I'd be fine with just using DS forever.

I think is more just traditional fraud

This is the correct take

Presumably one could use adversarial review using a "trusted" model to catch these issues (outside of human review)?

Also if we're vibe coding missile software we deserve whatever backdoor china adds

But judges are humans who can be blamed and held to account. LLMs seem incapable of taking on blame or receive any repercussions

All these dudes saying it costs the price of the hardware I guarantee don't even use keynote they just want to whine

Yes, and I'm not saying they're good, just that they're not some third party trying to sell you weiner pills or some gambling site.

I mean, yes of course it does, are you serious? If you say a crypto ad or a kalshi ad wouldn't you be more pissed off then Apple telling you about a similar product that is in their ecosystem?

Cool concept!

One big issue: not having a price and acreage range selectors means you see a lot of 0.2acre lots and $8mm parcels which I'm assuming most of us wouldn't be looking for

Pretty sure most people commenting here didn't even look at the post to see it's Apple promoting their own related product to their customers, not a third-party ad

You've got to be kidding? Microsoft has literal third party ads in the OS...

Until it hallucinates something no one bothers to verify

I didn't realize I had to spell it out but I was implying strictly a choice between model properties/capabilities

Perhaps on certain benchmarks and for certain work, but anecdotally I've not been able to see a difference between it and Opus on a lot of dev work (web, Go, iOS/AppleTV native, scripting, general tasks)

If you're not using the Superpowers set of skills, give it a shot. It's been working really well for me on a variety of tasks.

I'm using DSv4.1 in OpenChamber (eg OpenCode) using the Superpowers skills and a lot of custom AGENTS.md instructions to iron out the kinks and I genuinely cannot see a difference between it and Opus and I've been building native iOS and AppleTV apps, Go servers, Typescript, Cloudflare workers, Svelte/Astro, etc.

It's a super capable model all around from my experience.

IME GLM is inferior compared to Deepseek v4.1 Flash, highly recommend running it through some real work