I haven't used Mistral much until today but am impressed. I normally use Gemma 3 27B locally, but after regenerating some responses with Mistral 3 14B, the output quality is very similar despite generating much faster on my hardware.
The vision aspect also worked fine, and actually was slightly better on the same inputs versus qwen3 VL 8B.
All in all impressive small dense model, looking forward to using it more.
In my case phones were just starting to become commonplace, the Razr was the coolest phone to have, we had iPods but not iPhones, etc. Most instructors didn't want to see any phones and would threaten to take them away, so we became skilled at using T9 under the desk or in sweatshirt pockets, etc.
Block Dude! I also spent quite a bit of time writing functions and tools on my TI-84+, probably the closest thing I'll have to "growing up writing BASIC" since I missed that bus.
Imagine the worldly gains of allowing such an amazing technology to permeate society. Ah, well, that's against the interests of the shareholders. It's better to lock shit down and earn a dollar than precipitate betterment for human kind. The dollar! All hail!
Seconding this simply for data purposes. iOS 26 was the worst release I've ever dealt with coming from an iPhone 4S user to present. Goddamn there were so many obvious bugs and flaws. The .1 release fixed some of them but my keyboard still randomly shifts to the left by a few pixels every time it opens.
Take me back to the days where things were governed by UX and not revenue.
Interesting examples, thank you for sharing. My interactions with GPT 5.1 have shown knowledge of past interactions but I don't explicitly prohibit it as you have done through settings.
I would like to hear others' opinions as well. I guess for me it's a simple mechanism shared with many other social sites. A series of downvotes without explanation demonstrates the zeitgeist disapproves. Maybe the reason why is obvious, maybe it's not and thus deserves elaboration for the sake of everyone reviewing.
That'd be a stronger argument if there was an actual up/down vote count visible. Right now, there is not one - for users other than yourself, all you have is "positive/not" indicator.
So if you see someone's grayed-out comment, it could be a single disagreeing user, or hundred of them - you have no way to know.
Comments here only go to -4 before becoming dead and hidden by default. So if you're seeing a greyed out comment, at most four people found it disagreeable enough to downvote, and no one felt like upvoting it. Or there's been a lot of downs and ups but it balances out to, say, -2.
As our species looks pretty ill fitted in its niche these days, the zeitgeist is usually about preserving paradoxes at all costs to keep status quo. And that's reflected here pretty strongly. Much of what's come out of SV appeared at one time beneficial, and now it looks like the industry has shifted to dystopian. The zeitgeist that enforces that may be lost.
I think the TUV is mostly about safety equipment (light, breaks) that don't pass the two year check (they probably also test the alignment and other stuff), and your link might be about engine/transmission/power train defects.
Yes, I think that's an important part of the experience. One can edit some text, some CSS/JS if desired (optional), and immediately see the effect. Publish it and now your changes are global.
I have been reading through this guide a bit at a time on my phone simply because it's so well written and interesting. I don't even own a Mac Pro, it's just fascinating to read through all the challenges and workarounds to keep the platform modern. Kudos to the author and all the contributors who made it possible.
I haven't used Mistral much until today but am impressed. I normally use Gemma 3 27B locally, but after regenerating some responses with Mistral 3 14B, the output quality is very similar despite generating much faster on my hardware.
The vision aspect also worked fine, and actually was slightly better on the same inputs versus qwen3 VL 8B.
All in all impressive small dense model, looking forward to using it more.
reply