magnesit @magnesit.bsky.social · 20/08/2026People who oppose or hinder mRNA research are literally traitors against humanity, and today showed just that again 000
magnesit @magnesit.bsky.social · 18/08/2026The display is lower-resolution than my Pixel 7a. It's brighter, but the 7a was already bright enough. I love the design of the Pixel 10. It's much better than this blocky atrocity from before. All of that for a higher price. It's enshittified. Been 3 years and almost no technological development. 000
magnesit @magnesit.bsky.social · 18/08/2026The CPU is faster, but heck, you don't need that in everyday life. The 7a was more than enough from a performance standpoint already. 12 GB of RAM is nice though. The camera is not that much better. Sure, I can zoom 20x now, but why? I want storage and battery life, not... that. 100
magnesit @magnesit.bsky.social · 18/08/2026The Pixel 10 is a surprisingly shitty smartphone. I bought it, coming from my Pixel 7a, and it's genuinely not an upgrade. Can absolutely not recommend. Battery life is, at best, mid for a newly bought phone with almost no apps installed, and 128 GB of storage is genuinely tight for the price. 100
magnesit @magnesit.bsky.social · 09/07/2026Tested it for a bit. Alignment is very pro-Elon, but not as disastrous as I thought. Still problematic. 000
magnesit @magnesit.bsky.social · 09/07/2026Hmm. Reading the numbers, Grok 4.5 seems to be an absurdly good model. And relatively cheap. The high price you pay isn't in dollars, though... Credit where credit is due 100
magnesit @magnesit.bsky.social · 09/07/2026For my German readers, SPD and CDU/CSU representatives voted for Chat Control. It passed. Do with that info whatever you want. 000
magnesit @magnesit.bsky.social · 09/07/2026What the fuck, EU. Losing respect at a record pace. You know you'll never be able to get this through in practice. What the fuck. 100
magnesit @magnesit.bsky.social · 24/06/2026Listening to Ribs by Lorde when a cool breath of air streams through your window while laying on the bed in the heat is corny but peak as fuck and you ain't gonna change my mind 010
magnesit @magnesit.bsky.social · 22/06/2026Die Kunden sind nicht nur Leute, die Therapiearbeit für lau online benötigen. Insbesondere agentische Arbeit im Bereich Softwareentwicklung sind viel zu komplex für ein System wie ELIZA. Ich bin mir sicher, dass eine ganze Menge Nutzer:innen schon sagen würden, dass ELIZA deutlich schlechter ist. 200
magnesit @magnesit.bsky.social · 22/06/2026Also sorry, aber irgendwann ist auch mal gut. Ich bin absolut gegen Datenzentren an jeder Ecke, aber was du hier sagst, ist vollkommen falsch. Man darf Lügen nicht mit Lügen bekämpfen. Ich arbeite seit Jahren an maschinellem Lernen und möchte dir sagen, unterschätze bitte moderne LLMs nicht. 130
magnesit @magnesit.bsky.social · 17/06/2026Can Veritasium talk about a different topic than Prime Numbers for more than 2 consecutive months ffs 😭 000
magnesit @magnesit.bsky.social · 16/06/2026Mistral can do the funniest thing now... Pleeeease let them release Le Chaton Fat 000
magnesit @magnesit.bsky.social · 06/06/2026Gotta love these GitHub accounts that fork every repository in existence, do nothing to it and let leave, refusing to elaborate Absolute cinema 000
magnesit @magnesit.bsky.social · 04/06/2026doch, schon krieg, nur kommt der krieg schneller vor die Haustür 120
magnesit @magnesit.bsky.social · 04/06/2026lol it's happening, GitHub copilot is pulling the rug so hard 010
magnesit @magnesit.bsky.social · 19/05/2026No response. Interesting, right? I'm so sick of these three-word cool aah comments about obviously false claims. It's not possible to have a decent conversation like this. 010
magnesit @magnesit.bsky.social · 18/05/2026Okay, then tell me how to do it more efficiently. I'm all ears. 240
magnesit @magnesit.bsky.social · 18/05/2026I work in IT for more than a decade, let me tell you, it's still the most efficient way of serving websites to millions of people. You cannot hold the load of 42 million Bluesky Users with a potato computer. Data Center ≠ AI. 140
magnesit @magnesit.bsky.social · 18/05/2026While I'm all in for preserving our drinking water instead of generating slop AI videos, we certainly will need a small amount of data centers to keep things like this very platform running. Which needs data centers to work. 140
magnesit @magnesit.bsky.social · 16/05/2026The best Chinese LLM, as per the Artificial Analysis Intelligence Index, is open, while all adjacent American frontier LLMs remain closed. Nonsensical claim. 010
magnesit @magnesit.bsky.social · 16/05/2026Only Qwen Max and some Hunyuan models have been notably closed, and that's nothing new, they have been doing that for years now. The gap between the open Chinese models and the closed Chinese models is minimal. 100
magnesit @magnesit.bsky.social · 16/05/2026This is hilariously false. Chinese AI labs have released the following state-of-the-art LLMs in the last months: - Qwen 3.5 and Qwen 3.6 - Deepseek V4 Flash and Pro - Kimi K2.6 - Minimax M2.7 America has only released one big one, Gemma 4. 100
magnesit @magnesit.bsky.social · 16/05/2026Sorry but can anyone confirm this? They released more models than ever last month??? 010
magnesit @magnesit.bsky.social · 12/05/2026You don't have to take every rhetorical device literally, it was an example 120
magnesit @magnesit.bsky.social · 12/05/2026It's a self-sustaining cycle of rage bait. Many accounts spam AI generated images, and then many other accounts spam how much they hate these images in the comments. So weird 000
magnesit @magnesit.bsky.social · 12/05/2026How does this keep happening? Bluesky is a platform full of very loud anti-AI people, but still, it's the single platform that drowns in slop and and everybody keeps liking these images. Again, what do you want, man? How does this keep happening? 100
magnesit @magnesit.bsky.social · 12/05/2026Serving video, one of the key things social media does, to thousands upon thousands of people surely does use an enormous amount of compute as well. Data centers had a huge negative environmental impact long before AI came in. The problem is much bigger than boomers generating slop unfortunately. 100
magnesit @magnesit.bsky.social · 08/05/2026It's crazy how much Android differs from the "default" Linux desktop distribution many have in mind when talking about Linux. So many exploits are just outright not applicable because extensive hardening has been done by the AOSP team alone. Very reassuring. 040
magnesit @magnesit.bsky.social · 05/05/2026Did you read the post? They're trying to act now. Give it a bit of time, even if it may be tight, it's better than nothing! :) 100
magnesit @magnesit.bsky.social · 02/05/2026OpenRouter has literally spent time adopting OpenAI GPT-4o-mini Transcribe and Whisper Large V3 Turbo this (very young) May, as if there's nothing else to do. As you can here, all of that makes me just a little bit salty. 000
magnesit @magnesit.bsky.social · 02/05/2026When OpenAI launches "GPT-5.5-Codex-mini (xhigh Special Edition)", support is there within 1/10th of a second, but as soon as a Mistral model is remotely okay-ish, it's professionally ignored. Leaves a very subtle and weird taste. Again, I hope this is just a genuine coincidence. 100
magnesit @magnesit.bsky.social · 02/05/2026What's going on? Adoption of this model is surprisingly slow. I don't understand why several services pretend like Medium 3.5 does not exist, but Small 4 does. Historically, adoption of new models was much faster. Let's give it time. It's annoying though - this is REAL revenue Mistral is losing. 100
magnesit @magnesit.bsky.social · 02/05/2026However, they pretend like it's not there. You have to specifically search for it to see it there. I believe that this is not done with malicious intent, but it is weird nonetheless. Another example: OpenRouter. Not providing the new model. Support for Small 4 was there quickly though. 100
magnesit @magnesit.bsky.social · 02/05/2026Something's odd about the launch of Mistral Medium 3.5, and it's not their fault. Artificial Analysis paid close to $1k to benchmark the model but still defaults to showing Mistral Small 4, a much worse model, on their front page. Medium 3.5 scores way better (Deepseek V3.2 level). 100
magnesit @magnesit.bsky.social · 29/04/2026The new pricing of Medium 3.5 suggests Mistral is in need of money, probably because of failed training runs or low demand. They were able to provide Devstral 2 (123B dense) at a much lower cost than Medium 3.5 (128B dense), even though it should effectively be almost identically heavy on hardware. 000
magnesit @magnesit.bsky.social · 29/04/2026Mistral did a good write-up about it here: mistral.ai/news/vibe-re... 5/5mistral.aiRemote agents in Vibe. Powered by Mistral Medium 3.5. | Mistral AIIntroducing Mistral Medium 3.5, remote coding agents in Vibe, plus new Work mode in Le Chat for complex tasks. 000
magnesit @magnesit.bsky.social · 29/04/2026model Mistral has released in a long time. Now it's a good time to wait for independent benchmarks to come in and listen to the user experience of others. I've not been able to try it out myself for now, but definitely feel like it's going to be fire! Unfortunately hard to run locally. 4/5 100
magnesit @magnesit.bsky.social · 29/04/2026Benchmarks looking promising, but not SOTA. There's still room for improvement, but for now, it seems like they're doing okay. They've also told us that Medium 3.5 will be the new model powering Le Chat from now on. It's priced at a whopping 7.5$/M output tokens, the most expensive 3/5 100
magnesit @magnesit.bsky.social · 29/04/2026and strong agentic capabilities to it. This might actually be it. This might be a ticket back to the frontier-iiiish (doing a lot of heavy lifting here). It's probably slow and expensive to run at first, but if more inference providers pick it up, hey, why not? 2/5 100
